You are on page 1of 176

Lesson 1 Introduction Outline Statistics Descriptive versus inferential statistics Population versus Sample Statistic versus Parameter Simple

Notation Summation Notation Statistics What are statistics? What do you thing of when you think of statistics? Can you think of some examples where you have seen statistics used? You might think about where in the real world you see statistics being used, or think about how statistics in used in your major. Statistics are divided into two main areas: descriptive and inferential statistics. Descriptive statistics- These are numbers that are used to consolidate a large amount of information. Any average, for example, is a descriptive statistic. So, batting averages, average daily rainfall, or average daily temperature are good examples of descriptive statistics. Inferential statistics- inferential statistics are used when we want to draw conclusions. For example when we want to determine if some treatment is better than another, or if there are differences in how two groups perform. A good book definition is using samples to draw inferences about populations. More on this once we define samples and populations. Population- Any set of people or objects with something in common. Anything could be a population. We could have a population of college students. We might be interested in the population of the elderly. Other examples include: single parent families, people with depression, or burn victims. For anything we might be interested in studying we could define a population. Very often we would like to test something about a population. For example, we might want to test whether a new drug might be effective for a specific group. It is impossible most of the time to give everyone a new treatment to determine if it worked or not. Instead we commonly give it to a group of people from the population to see if it is effective. This subset of the population is called a sample. When we measure something in a population it is called a parameter. When we measure something in a sample it is called a statistic. For example, if I got the average age of parents in single-family homes, the measure would be called a parameter. If I measured

the age of a sample of these same individuals it would be called a statistic. Thus, a population is to a parameter as a sample is to a statistic. This distinction between samples and population is important because this course is about inferential statistics. With inferential statistics we want to draw inferences about populations from samples. Thus, this course is mainly concerned with the rules or logic of how a relatively small sample from a large population could be tested, and the results of those tests can be inferred to be true for everyone in the population. For example, if we want to test whether Bayer asprin is better than Tylonol at relieving pain, we could not give these drugs to everyone in the population. It’s not practical since the general population is so large. Instead we might give it to a couple of hundred people and see which one works better with them. With inferential statistics we can infer that what was true for a few hundred people is also true for a very large population of hundreds of thousands of people. When we write symbols about populations and samples they differ too. With populations we will use Greek letters to symbolize parameters. When we symbolize a measure from a sample (a statistic) we will use the letters you are familiar with (Roman letters). Thus, if I measure the average age of a population I’d indicate the value with the Greek letter “mu” (µ =24). While if I were to measure the same value for a subset of the population or a sample then I would indicate the value with a roman letter ( X =24). Simple Notation You might thing about descriptive statistics as the vocabulary of the "language" of statistics. If this is true then summation notation can be thought of as the alphabet of that language. Notation and summation notation is just a short hand way of representing information we have collected and mathematical operation we want to perform. For example, if I collect data on a variable, say the amount of time (in minutes) several people spent waiting at a bus stop, I can represent that group of numbers with the variable X. The variable X represents all of the data that I collected. Amount of Time X 5.0 11.1 8.9 3.5 12.3 15.6 With subscripts I can also represent an individual data point within the variable set we have labeled X. For example the third data point, 8.9, is the X3 data point. The fifth data point X5 is the number 12.3. Very often when we want to represent ALL of the data

points in a variable set we will use X by itself, but we may also add the subscript i. Whenever you the subscript i, you can assume that we are referring to all the numbers for the variable X. Thus, Xi is all of the numbers in the data set or: 5,11.1,8.9,3.5,12.3,15.6. There are other common symbols we will use besides X. Sometimes we will have two data sets to deal with and refer to one distribution as X and the other distribution as Y. It is also necessary for many formulas to know how many data points are in a data set. The symbol for the number of data points in a set is N. For the data set above the number of data points or N = 6. In addition, we will use the average or mean value a good deal. We will indicate the mean, as noted above, differently for the population (µ) than for the sample ( X ). Summation Notation Another common symbol we will use is the summation sign ( ∑ ). This symbol does not represent anything about our data itself, but instead is an operation we must perform. Whenever you see this symbol it means to add up whatever appears to the right of the sign. Thus, ∑X or ∑Xi tells us to add up all of the data points in our data set. For our example above it would be: 5 + 11.1 + 8.9 + 3.5 + 12.3 + 15.6 = 56.4. You will see the summation sign with other mathematical operations as well. For example ∑X2 tells us to add all the squared X values. Thus, for our example: ∑X2 = 52 + 11.12 + 8.92 + 3.52 + 12.32 + 15.62 -or25 + 123.21 + 79.21 + 12.25 + 151.29 + 243.36 = 634.32. A few more examples of summation notation are in order since the summation sign will be central to the formulas we write. The following examples should give you a better idea about how the summation sign is used. Be sure you recall the order of operations needed to solve mathematical expressions. You will find a review on the web page or you can click here: http://faculty.uncfsu.edu/dwallace/sorder.html For the examples below we will use a new distribution. X = 1 2 3 4 Y=5678

ΣX 2 ≠ (ΣX )

2

what is important is to get a feel for how the summation sign works in equations. and then the addition. Thus: 12 + 2 2 + 3 2 + 4 2 ≠ (1 + 2 + 3 + 4) 1 + 4 + 9 + 16 ≠ (10)2 30 ≠ 100 ΣX 2 ≠ (ΣX ) 2 2 For the next expression we show. Again. that the law of distribution applies to the summation sign as well.For this expression we are saying that the sum of the squared X’s is not equal to the sum of the X’s squared. Notice here we want to perform the operation in parentheses first. Σ( X + Y ) = ΣX + ΣY (1+5)+(2+6)+(3+7)+(4+8) = (1+2+3+4)+(5+6+7+8) 6 + 8 + 10 + 12 = 10 + 26 36 = 36 . like in algebra. and then the exponents.

on the other hand. Whenever we measure a variable. Another dimension on which variables might differ is that they may be either continuous or discreet. but are instead numbers used to differentiate objects. Each scale presented is in order of increasing order. This is a broad definition that includes most everything we will be interested in for an experiment. can assume only a few possible values on the scale used to measure it. You could have an object that weighed 1 pound or 1. Real world examples of these variables are common. Thus. time measurement. it could be a measurement (quantitative) difference or a categorical (qualitative) difference. All are possible measures. weight.25. The following scales are delivered in order of increasing complexity.” Scales of Measure – whenever we measure a variable it has to be on some type of scale.5 or 1. The numbers .25 televisions in your home. These are also called qualitative variables because there is some quality that distinguishes these objects. their reactions times. Discreet variables. Categorical variables are measures of differences in type rather than amount. Nominal scales – These are not really scales as all. Examples include things like height or weight. Divisions of measures are usually not valid. or color. gender.” A discreet variable is a measure of “how many. as well as 1. Thus. A continuous variable is a variable that can take on any value on the scale used to measure it. ordinal. It could be the age or gender of participants. You either have a television or you don’t. These examples are also called quantitative because they measure some quantity. Another way to keep this difference in mind is that with a continuous variable is a measure of “how much.5 pounds or 1. interval.5 televisions or 1. height. Divisions of these values are not valid. or number of children in a household. You should know both terms for each type. Examples include age. if I measure the number of television sets in your home it could be 1 or 2 or 3.Lesson 2 Scales of Measure Outline Variables -measurement versus categorical -continuous versus discreet -independent and dependent Scales of measure -nominal. It is something we can measure. Any division on any unit on the scale produces a valid possible measure.25 pounds. Measurement variables are things to which we can assign a number. ratio Variables A variable is anything we measure. a measure of 1 or 2 is valid. So. Examples include anything categorize such as race. you could not have 1. or anything we might be interested in.

Interval scales contain an ordinal scale (objects are in order). but have the added feature that the distance between scale units is always the same. Ordinal Scales – Ordinal scales use numbers to put objects in order. Interval Scales. A good example is class rank. In fact scientist invented the Kelvin temperature scale so that they would have a measure of temperature on a ratio scale. social security numbers. Ratio Scales – Ratio scales contain an interval scale (equal intervals between units on the scale). Notice that it is not valid to have a measure below zero on any of these scales. Class rank would not qualify because we don’t know how much better one unit is than another. Something could not weigh a negative amount.are just labels. but we don’t know how much better four is than five. in order to make ratio statements such as something is twice or half of another then it must be a variable on a ratio scale. . the channels on your television. and sports team jersey’s are all good examples of nominal variables. Examples include temperature (in Fahrenheit or Celsius). but with interval there is the same distance from one unit to the next anywhere we are on the scale. but have the added feature that there is a true zero point on the scale. No other information other than more or less is available from the scale. So. or altitude. Examples include height or weight or measures of amount of time. This zero point is necessary for ratio statements to have meaning. These scales are much more common than interval scales because if a scale usually has a zero point. Someone ranked at four had a higher GPA than someone ranked as five. or any type of ranking. Again. For temperature you know that the difference in ten degrees is the same no matter how hot or cold it might be.

especially if it causes confusion. you can reverse the process as well. It might also be helpful in some examples to go from a frequency distribution back to original data set. if I record the number of trips out of town (X) a sample of FSU students makes. four people took two trips out of town. four people took one trip out of town. X 0 1 2 3 4 5 6 7 f 5 4 4 5 4 3 2 1 From the graph we can see that five people took no trips out of town. The f-values are just a count of how many. . and the corresponding value for “f” indicates how many people in the sample gave us that particular response. Below is a simple frequency distribution where the X column represents the number of trips. we can record how many of each value (f) there are for the entire x-distribution. With frequency distributions we will simply record the frequency or how many values fall at a particular point on the scale. So. and so on. For example.Lesson 3 Data Displays Outline Frequency Distributions Grouped Frequency Distributions -class interval and frequency -cumulative frequency -relative percent -cumulative relative percent -interpretations Histograms/Bar Graphs Frequency Distributions We often form frequency distributions as a way to abbreviate the values we are dealing with in a distribution. I might end up with the following data: 0 2 5 3 2 4 3 1 0 2 6 0 4 7 0 1 2 4 3 5 4 3 1 6 1 0 5 3 Instead of having a jumbled set of numbers. It is important not to confuse the f-value and the x-value.

Notice that if I tried to count how many values fall at any single point on the scale my frequencies will all be one.95 10. Since there is a gab in each class interval.In the following example I start with a frequency distribution and go backward to find all the original values in the distribution.87 7. but when we measure a variable it is often on a continuous scale. with . X 0 1 2 3 4 f 2 3 4 3 2 What is the most frequent score? The answer is two because we will have four ‘twos’ in our distribution: 0 0 1 1 1 2 2 2 2 3 3 3 4 4 Grouped Frequency Distributions The above examples used discreet measures. In turn. In order to build the frequency distribution we will group several values on the scale together and count any of measurements we observe in that range for the frequency. In the end.92 6.69 9.25 8. we will be actually counting any values that would get rounded down or up into a particular interval.56 9. we want to construct a display that has between 5 and 15 intervals. Thus: Class Interval 0-2 3-5 6-8 9-11 12-14 Once we have the class interval. For example.59 12.58 8.25 3.53 14. This will be the range of value on the scale we include for each interval.95 4.45 14.34 13. 3.86 We will begin by forming the class interval.99 11.61 5. but for this course I will always indicate how big the interval should be.67 10.24 10. we will count how many values fall within the range of each interval. if we measure the running time of rats in a maze we might obtain the following data. there will be few values we measure that are at the exact same point on the scale.12 7. For example. There are many rules we could use to determine the size of the interval.34 11.

5 2. For the third interval we have five values.26 would be rounded down into the 6-8 class interval. at the first interval we have zero frequency. so cumulatively we have three. That includes the five for the third interval.5 11.5 Class Interval 0-2 3-5 6-8 9-11 12-14 f 0 3 5 7 5 Notice that my real limits cover half the distance of the gap between each class interval.5 is half the distance.5-14. If a value falls exactly on one of the real limits we could randomly choose its group. Cumulative Frequency Once we have formed the basic grouped frequency distribution above.5-11.5 11.5 8. Real Limits -. Most of the time this value will be 0. real limits have no gap. So.5-8. The value 8. plus the three from the previous intervals. Notice that when we reach the last interval we have all the values in the distribution represented.5-8. the bottom cumulative frequency is N or the total number of values in the distribution (20 here).5-2. so cumulatively we have 8.5 5. f 0 3 8 15 20 So. The first of these is the cumulative frequency column. we can add more columns for more detailed information. We will include a column to indicate the real limits of the class interval. With this column we will keep a running count of the frequency column as we move down the class interval. These are the limits of the interval. but the class intervals do. We continue this process until the last interval. including any rounded values.5-14.5 since most scales will have one unit values and 0.5 8.5 Class Interval 0-2 3-5 6-8 9-11 12-14 f 0 3 5 7 5 Cum.5 2. . Real Limits -.69 would be rounded up into the 9-11 class interval.5-5.5-5.5-11. So.5-2.5 5. so cumulatively we have zero values.the above data the value 8. For the second interval we have three.

5 2. what percentile is associated with a score of 8.5? We would use the cumulative relative percent that corresponds to 8.5-5.5 5.5-8.5. Thus: Real Limits -.5-5. what score corresponds with the 75th percentile? The answer is 11. f 0 3 8 15 20 Rel % 0 15 25 35 25 Cumulative Relative Percent For a final column we will keep a running count of the relative percent column in the same way we did with the cumulative frequency.5-2. the score 8. or at or below a certain .5 11. Other interpretations from the table can be made as well. Since we will be rounding values down into a certain interval based on the real limits. So.5 because any values of 11.5 8.5 5. You often get percentile information when you get your SAT or ACT test scores back. For example. we might be interested in how many people fall at a particular interval.5 or less are within the bottom 75% of the distribution. which is 40%. but we could also obtain the same numbers by computing the percentage for each cumulative frequency as well. Percentile information is found in the cumulative relative percentage column.5-14.5 11.5 corresponds with the bottom 40% or 40th percentile of the distribution. % 0 15 40 75 100 Notice that we can keep a running count of the relative percent column.5 8.5-14. we will express the frequency (column) as a percentage of the total. Interpretations The data display gives a good deal of information about where values in the sample fall. Rel. That is.5-11.5 2. Similarly. Each value in that column tells us the percentage of the distribution at that point or less on the scale. Move the decimal over two places (or multiply by 100) to change the proportion into a percent.Relative Percent Another column will tell us the proportion of total values that fall at each interval. Keep in mind we are counting relative percentages now as we move down the display.5-8.5-11.5-2. Real Limits -. f 0 3 8 15 20 Rel % 0 15 25 35 25 Cum. then we will indicate where on the scale a certain percentile is based on its corresponding upper real limit. One good piece of information is about percentiles. This will give us the proportion of values for that particular interval.5 Class Interval 0-2 3-5 6-8 9-11 12-14 f 0 3 5 7 5 Cum. To convert the frequency to a percentage take the frequency (f) and divide by the number of values (N). For example. A percentile is the percentage at or below a certain score.5 Class Interval 0-2 3-5 6-8 9-11 12-14 f 0 3 5 7 5 Cum.

5-2. The bars touch like this when we are dealing with continuous data rather than discreet data. we would use a bar graph if we wanted to create a data display.5-11. or eight.5 11. and the lines do not touch. and independents in a sample. When the scale measures discreet values we call it a bar graph.5-14. Rel.interval. if I measured the number of democrats. We will use this value to graph the display.5 5. we will use the midpoint of each interval to indicate different points on the scale. % 0 15 40 75 100 Note that the bars are touching. How many scored 8.5 8. republicans.5 2. Real Limits -.5-5. but notice I have created a new column that indicates the center or midpoint of each interval. Histograms/Bar Graphs We can also take the frequency information in our frequency or grouped frequency distribution and form a graph. How many scored between 3 and 5? The answer is a found in the frequency column. or three. We will continue with our previous example.5-8. . For example. On the x-axis we will place values from our scale. f 0 3 8 15 20 Rel % 0 15 25 35 25 Cum.5 or less? The answer for this question is in the cumulative frequency column. and on the y-axis we will plot the frequency for each point on the scale. For grouped frequency distributions.5 Class Interval 0-2 3-5 6-8 9-11 12-14 MP 1 4 7 10 13 f 0 3 5 7 5 Cum. In the graph we will form a simple x-y axis.

600 500 Frequency 400 300 200 100 0 Dem Rep Party Ind .

I will draw these shapes to illustrate different point throughout the semester. If we trace the outline of the edges of the frequency bars you can idea about the shape. Keep in mind what you are looking at is a line indicating the frequency or how many values in a distribution lie at a particular point on the scale: just like a histogram.Lesson 4 Measures of Central Tendency Outline Measures of a distribution’s shape -modality and skewness -the normal distribution Measures of central tendency -mean. . and mode Skewness and Central Tendency Measures of Shape With frequency distribution you can an idea of a distribution’s shape. median. From this point on.

2. A single major peak is unimodal. 10. . If the tail is to the right it is a positive skew.Modality – measures the number of major peaks in a distribution. On the other hand a negatively skewed distribution might be: 1. 10. 3. 2. Two major peaks is a bi-modal distribution. 9. and is not skewed. 10. The outliers are on the high end of the scale. 3. and is the most common modality. 1. a positively skewed distribution would be: 1. 2. 3. Here the outliers are on the low end of the scale. Skewness – measures the symmetry of a distribution. 2. Skewness is indicated by the “tail” or trailing frequencies of the distribution. 9. 10. 11. If the distribution is not symmetric. 9. A symmetric distribution is most common. 9. You could also have multi-modal distributions. and one side does not reflect the other. then it is skewed. 11. 11. If the tail is to the left then it is a negatively skewed distribution. For example.

we end up with mostly “normally” shaped distributions.Q. is the arithmetic center of the distribution. That’s because most of the scores in a distribution will tend to cluster about the center. With I.The normal distribution is one that is unimodal and symmetric. Think about measures of height. Measures of Central Tendency Knowing where the center of a distribution is tells us a lot about a distribution. You can find the mean by adding all the scores ( ΣX ) together and dividing by the number of values you added together (N). Regression toward the mean is an idea that states values will tend to cluster around the mean with few values toward the trailing ends or “tails” of the distribution. There are very few people that are extremely tall or extremely short. Measures of central tendency give us a number that describes where the center lies (and most scores as well). the same pattern will repeat. As a result. Most things we can measure in nature exhibit a normal distribution of values. most things we measure will tend to have a normal shape. or most anything we can measure. but most tend to cluster around the average. Since most things we measure have more values close to the mean. measures of weight. or average score. X 1 2 3 4 5 ΣX = 15 N=5 For the Population: µ = ∑ X = 15 = 3 N 5 For the Sample: x = ∑X= n 15 =3 5 . scores. Mean – The mean.

It is the value in the middle when the values of the distribution are arranged sequentially. This concept will be important when we consider standard deviation because we will need to look at differences between values in our distribution and the mean. The distribution: . 4. A deviation is just a difference. If there are several outliers in a distribution it will often result in skewed shape to the distribution. 5 has an average of 3. 25 has a mean of 7. Most students are familiar with these measures of central tendency. 3. or a skewed distribution. Thus. 4. The new distribution 1. 3) The third property is that it is strongly influenced by outliers. 2. An outlier is an extreme score. 3. say by substituting 25 for the 5 in the above distribution. the mean is a poor measure of the center when we have outliers. Thus. when we say the sum of the deviations about the mean must always equal zero is just a way of saying that there are just as many differences between values above the mean and the center as there are differences between values below the mean and the center. 1) The first property of the mean is that it is the most reliable and most often used measure of central tendency. then the mean changes a great deal. The deviations are the differences between the score and the mean. if there is an outlier. Other statistics will differ in how they are computed for the sample versus the population. The last two properties need further explanation. no matter what original values we use.Note that we calculate the mean the same way for both the sample and the population we symbolize them differently. 2) The second property of the mean is that it need not be an actual score in the distribution. for a simple distribution 1. Let’s use population symbols and say µ = 3. Median – The median is the physical center of the distribution. Thus a distribution of values 1. It is a score that lies apart from most of the rest of the distribution. Seven is not really close to most of the other values in the distribution. However. So. X 1 2 3 4 5 µ=3 X. but there are several properties that may be new to you.µ 1-3 = -2 2-3 = -1 3-3 = 0 4-3 = 1 5-3 = 2 Σ( X − µ ) = 0 Now if we add these deviations we will always get zero. A deviation from the mean is the difference between a score and the mean. 5 the average is 3. 2. 4) The fourth property is that the sum of the deviations about the mean must always equal zero. 4. Outliers tend to pull central tendency measures with them. Three does a good job of describing where most of the scores in this distribution lie. 3. 2.

4. 1. 7 there is only one mode (4). in the distribution: 1. 2. 5. 4. 2. 7 has a median value of 4 because there are five values above this point and five values below this point in the distribution (1. 5. 5. But. The opposite is true for positively skewed distributions. Here the outliers are on the high end of the scale and will pull the mean in that direction a great deal. 5. 4. 3. 5. 3. 3. 3. 2. 1. 3. 6 then the two values in the center are 3 and 4 (1. 5. For example in the distribution: 1. 2. 6.1. Let’s consider this information further. 3. 4. but the mode will be unaffected. If you have an even set of numbers then there will be two values that are at the center. 6). 5. It is possible the median might move slightly in the direction of the skew or outliers in the distribution. 5. 3. 5. If there is only one of each value then there is no mode. 3. The mode is not affected by outliers. Since it is only determined by one point on the scale. 5. That is. 5. 2. 2. 4. 1. 4. It is simply the value that appears most often. 1. but not the mode. and you average these two values together in order to determine the median. 6 there are two modes (3 and 4). other values will have no effect. 3. 1. Mode – The mode is the most frequent value in the distribution. . with a positively skewed distribution the mean is numerically higher than the median or the mode. 2. 3. 3.5 and that is the median. 2. The median might be slightly affected as well. 6. 5. if we take out one of the numbers in the distribution so that we have 1. 5. 6. 4. 4. with a negatively skewed distribution the mean is numerically lower than the median or mode. 5. 2. 4. For example. outliers will generally not affect the median and it will not be affected as much as the mean. The median might be slightly lower due to the outlier. 4. Thus. 7). 2. Skewness and Central Tendency We have already discussed how each measure is affected by outliers or skewed distribution. 4. Thus. The average is 3. In a positively skewed distribution the outliers will be pulling the mean down the scale a great deal. The median is resistant to outliers. 4.

The second quartile is the same thing as the 50th percentile and the median. However. Interquartile Range (IQR) – The interquartile range is the range of the middle 50% of a distribution. A quartile is a quarter or 25% of the distribution. of that possibility we will use a method that will yield a consistent. IQR = Quartile 3 – Quartile 1 or IQR = 75th percentile – 25th percentile. Measures of dispersion or variability will give us information about the spread of the scores in our distribution. 9.Interquartile Range . 11 the range is 11 – 1 = 10. 3. 2) Next find the median of the bottom half of the distribution .Standard Deviation and Variance Measures of Dispersion Measure of central tendency give us good information about the scores in our distribution. yet have the same central tendency. with the interquartile range we will eliminate the bottom 25% and top 25% of the distribution. let’s learn some new definitions. Remember to subtract the two numbers to give one number for the final answer. When we compute the IQR we will want to find each of the quartiles. To compute the IQR first arrange your numbers from lowest to highest and 1) find the median. So. So. Before we compute the value. we can have very different shapes to our distribution. To actually compute some IQR’s we would need to use calculus. Thus. The IQR is the found by eliminating the values that lie between the bottom end and the first quartile (bottom 25%). Instead.Range . It’s a starting point for us to find the other quartiles. We will also eliminate the values between the third quartile and the top of the distribution. and then measure the distance between the extremes of the middle 50% of the distribution that remains. in the distribution: 1. the range as a measure of dispersion can be strongly influenced by outliers. or are the scores spread out over a large segment of the scale? Range – The range is the difference between the high and low score in a distribution. 5. Simply subtract the two numbers to find the range. Are the scores clustered close together over a small portion of the scale. The third quartile is the same as the 75th percentile. and somewhat accurate answer. One solution to this problem is to eliminate the ends of the distribution and measure the range of scores in the middle. The median is the 50th percentile and second quartile. The first quartile is the same as the 25th percentile because 25 percent of the distribution is at or below that point.Lesson 5 Measures of Dispersion Outline Measures of Dispersion . We then subtract the new low and high score of the left over middle part of the distributions. Because any outliers in our distribution must be on the ends of the distribution.

3) Find the median of the top half of the distribution just like we did for the bottom.(ignoring the top half). 4) Next subtract the upper and lower medians you found in step 2 and 3. we use all the values in our data to compute variability. and outliers will not have undue influence. 1 2 5 6 7 9 10 12 15 19 Once we find the median we can divide the distribution into two halves 1 2 5 6 7 9 10 12 15 Bottom 50% Top 50% 19 The median of the bottom 50% is 5. The value 8 is the 50th percentile or second quartile. This value is the 25th percentile or first quartile because we have taken the bottom 50% and cut it in half. In the following example the median is 8 because it is the average of the two middle numbers. The median of the top 50% is 12. Generally. this is not acceptable because of the difficulty in collecting data in the first place. IQR = 12 – 5 = 7 Bottom 25% Top 25% 1 2 5 6 7 Middle 50% 9 10 12 15 19 Standard Deviation and Variance While the interquartile range eliminates the problem of outliers it creates another problem in that you are eliminating half of your data. though we will not use this number in the computation. So. The solution to both problems is to measure variability from the center of the distribution. In this way. . This value is the 75th percentile or third quartile. Both standard deviation and variance measure how far on average scores deviate or differ from the mean.

2. In this example it is 10/5 = 2. 3. Since we are in squared units and not the same units as our scale we can take the square root of the variance in order to get the standard deviation. I’ll use the same example with the simple distribution 1.µ 1-3 = -2 2-3 = -1 3-3 = 0 4-3 = 1 5-3 = 2 Σ( X − µ ) = 0 (X – µ)2 4 1 0 1 4 Σ( X − µ ) 2 = 10 Once we add the squared deviations we have a measure of overall variability in the distribution. First we find the mean and the deviations about the mean. X 1 2 3 4 5 µ=3 X. Once we divide these squared sums we will get the average squared deviation or variance. When we square all the numbers the negative values will all become positive and we can then add the deviations without getting zero. Recall that we did the same thing when discussing properties of the mean. The standard deviation is the average deviation about the mean.µ 1-3 = -2 2-3 = -1 3-3 = 0 4-3 = 1 5-3 = 2 Σ( X − µ ) = 0 One solution is to square all of the deviations. Please refer back to this section if formulas give you problems later on in the course. What we want to do is add up these deviations and find out how far on average the scores deviate from the mean. 4. How can we get an average if the sum is always zero? X 1 2 3 4 5 µ=3 X. and will be used in almost everyone formula we learn this semester. For our example we take the square root of 2 and find 1. The sum of the average squared deviations is called the sums of squares.41 is the standard deviation.To compute standard deviation and variance we first start by finding the deviation about the mean. 5. The problem we run into is that whenever we add the deviations (in order to find the average of the deviations) they will always sum to zero. .

the one presented on the web page will be easier to use in the long run.edu/dwallace/ssandrd1. we divide by n – 1. and instead of subtracting the numerator by N. n −1 sample variance S 2 ∑ (X − X ) = 2 s = s2 sample standard deviation Please review the animated demonstration on variance and standard deviation for another example of how the population formula works. Note that σ2 is just the symbol we use for population variance and σ is the symbol we use to denote population standard deviation.uncfsu. In addition an alternative formula for these same computations is presented.The formula that contains all these operations is as follows.58).html . N population variance σ 2 ∑ (X − µ ) = 2 σ= σ 2 population standard deviation When dealing with a sample a minor change to the formula is made. Try the numbers in the above example to compute the sample variance and standard deviation (variance is 2. See http://faculty.5. Although the formula detailed here is the best for understanding the concept. Both appear in the homework packet formula section as well. standard deviation is 1.

the relationship between values is not affected. but then found the tape measure I was using was missing the first two inches (so it started at inch two instead of zero).Lesson 6 Z-Scores Outline Linear Transformation -effect of addition/subtraction -effect of multiplication/division Z-Transformation Z-Distribution -properties Using Z-scores to compare values Linear Transformation Anytime we change a distribution by using a constant we perform a linear transformation. what would I have to do to find the true heights of everyone? If you think about it you will see that I must subtract two inches from each measurement to get the true heights (because the start position was too high). This example of a linear transformation is one in which we simply shift the numbers up on the same scale. How does adding or subtracting a constant affect the mean and standard deviation? How does multiplying and dividing a constant affect the mean and standard deviation? . Notice that even though all the numbers move. For example if I measure the heights of everyone registered in this course. X 5’5” 5’7” 5’8” 5’10” X+2 5’7” 5’9” 5’10” 6’0” You will need to know how linear transformations affect the mean and standard deviation of a distribution as well.

91 Z-Transformation The z-transformation is a linear transformation. For these transformations the mean will change by the same amount as the constant. just like those we have discussed.When adding or subtracting a constant from a distribution. Here is the formula for transforming a raw X−µ z= score in a population to a Z-score: σ Notice that the distance a score lies from the mean is now relative to how much scores deviate in general from the mean in the population. We do not affect the distance between values. we change the distance between values rather than just shifting them up or down the scale. In the following example. when we use the Z-transformation we obtain a standard measure of the distance of the score from the mean. the raw score that produced the Z is exactly 1. for example. we add a constant and see the changes to the mean and standard deviation. Regardless of what the raw score values are in the population. we multiply a constant and see the changes to the mean and standard deviation. X 1 2 3 4 5 µ=3 σ = 1. Transforming a raw score to a Z-score will yield a number that expresses exactly how many deviations from the mean a score lays.41 X +5 6 7 8 9 10 µ=8 σ = 1. The standard deviation will remain unchanged. the raw score that produced the Z is exactly one standard deviation from the mean for that population. This fact is true because.5. Anytime Z=1. In the following example.41 X*5 5 10 15 20 25 µ = 15 σ = 7. That is because when we multiply numbers together. the mean will change by the same amount as the constant. . Anytime Z=1.41 The effect is a little different when we multiply or divide by a constant.5 standard deviations from the mean for that population. X 1 2 3 4 5 µ=3 σ = 1. again. but this time the standard deviation will change too. we are just shifting the distribution up or down the scale.

Let’s transform the simple distribution into a distribution of z-scores by plugging each value into the z-formula: X 1 Z-Tranformation 1− 3 z= = 1.42 µ=0 σ=1 2 3 4 5 µ=3 σ = 1.41 X−µ = σ X−µ z= = σ z= z= z= X−µ = σ X−µ = σ Z -1. What z-score will we expect to get for this score? Will it be less than one standard deviation or more than one standard deviation away from the mean? You can estimate the z-score by counting from the mean.42 -. X 1 2 3 4 5 µ=3 σ = 1.41 What z-score will we expect the value “3” to have in this example? That is. We have to go down 1.71 0 . One standard deviation is 1.41 .Think for a minute about what it means to know how many standard deviations from a mean a score lays. So. That’s a little less than one standard deviation. so it is zero standard deviations from the mean and we will get a z-score of zero for the original value of three. Consider our simple distribution example. we are two units from the mean and a little more than one standard deviation below the mean. how many standard deviations from the mean is “3”? The answer is that it is at the mean. Now consider the value “1” in the distribution.41 units.71 1. when we get to “1” on the scale.41 units from the mean before we reach one standard deviation. Counting down from the mean the value “2” is one unit from the mean.

71 standard deviations below the mean. and units on the z-distribution are themselves standard deviations. The value “2” is . it will be expressed in a standard form. so any scores below the mean in the original distribution will always have a negative z-score and any score above the mean will have a positive z-score. We can compute the mean and standard deviation of the resulting z-distribution as well. Who has more stress. These facts make sense because the mean is always zero standard deviations away from the mean.The value of “1” is 1. Using Z-scores to Compare Values Since z-scores reflect how far a score is from the mean they are a good way to standardize scores. but both are at the mean of their own distribution. Notice that the mean is at zero. We can take any distribution and express all the values as z-scores (distances from the mean). Let’s say that Joan got an x = 88 in a class that had a mean score of 72 with a standard deviation of 10 (µ = 72. . say I measure stress in the average college sophomore on a scale between 0 and 30 and find the mean is 15. For example. This standard form can be used to convert different scales to the same scale so that direct comparison of values from the two different distributions can be directly compared. Properties of the z-distribution. σ = 10). and so on. a person with a stress score of 15 from the first distribution or a person with a stress score of 150 from the second group? The value of 150 is a much larger number than 15. So. and the standard deviation will always be one. but uses a scale that ranges from 0 to 300 with a mean of 150. they will both have the same z-score of zero. Thus.42 standard deviations below the mean. Another researcher measures stress with the same population. Also notice in the above example that we had to compute the mean and standard deviation of the simple x-distribution in order to compute the zscore. no matter scale we originally used to measure the variable. Both values indicate an average amount of stress. Consider another example. The mean of the z-distribution will always be zero.

and Bob’s score is only 1 standard deviation above the mean. however. Bob Z= 92 − 87 5 = =1 5 5 . The mean for Bob’s class. was 87 with a standard deviation of 5 (µ = 87. however. if we simply compute the z-score for each.6 Joan has the higher score because she is 1. Joan Z= 88 − 72 10 16 10 = = 1.6 standard deviations above the mean. Who had the higher grade relative to their class? If you think about it for a second you will know that Joan’s score of 88 is much higher relative to the average of 72 compared to Bob’s score of 92 to the average of 87. We can easily compare the values. σ = 5).In a different class let’s say Bob got a x = 92.

Start this section by reviewing the first two topics in the above outline on the web page.Lesson 7 Z-Scores and Probability Outline Introduction Areas Under the Normal Curve Using the Z-table Converting Z-score to area -area less than z/area greater than z/area between two z-values Converting raw score to Z-score to area Converting area to Z-score to raw score Introduction/Area Under the Curve Please note that area. proportion and probability are represented the same way (as a decimal value). . The table shows the z-score in the left column.edu/dwallace/sz-score. Whenever we compute a z-score it will fall on the distribution at some point. See page A-24 through A-26 in your text. and finally there is a third column that shows the proportion or area under the curve in the tail of the distribution. The larger portion is the body. There is a table of z-scores that gives the corresponding areas or probabilities under the curve. You will need the z-table in Appendix B of your text for this discussion.uncfsu.html Using the Z-table Knowing the number of standard deviations from the mean gives us a reliable way to know how likely a score is for a population. Find the Z-score animated demonstrations or click here http://faculty. Just move the decimal place to the right two places to turn the decimal into a percentage. and then the proportion or area under the curve in the body. and the smaller portion is the tail. Some examples require that you convert the decimal value to a percentage.

1587. Now let’s consider the situation if the z-score is negative. Converting a Z-Score to an Area Finding areas below a z-score What area lies below a z-score of +1? If we look this z-score up on the table we find that the area in the body is .5) up on the table we find that the area in the body is .0668 is the proportion in the population below a z-score of -1. We will use a different column on the table. So.If the z-score is positive. so if we want to know the area for a negative value we just look up the positive z-score to find the area. The distribution is symmetric. Since the z-score is positive and we want the area below the z-score.5? You will not find negative values on the table. If we look this z-score (1. . then the body will be the area that is above that z-score. What area lies below a z-score of -1. .0668. then we will want to look at the body area. and the tail will be the area that is below that zscore. and that is why we must consider whether z is positive or negative when using the table. and the area in the tail is . .843 is the proportion in the population that have a z = 1 or less.8413 and the area in the tail is. Since the z-score is negative and we want the area below that point we will be using the tail area. So. then the body will be the area that is below that z-score. If the z-score is negative. and the tail will be the area that is above that z-score.9332.5.

What area lies above a z-score of -1. The only difference is in which column we will be using to answer the question. For example.5? Since both scores are positive and we want the area between them.9332.5. . we can find the tail area for both z-scores and subtract the two areas. then we will want to look at the tail area. only areas from the table. and the area in the tail is . You could just as easily find the two body areas for both z-scores and subtract them as well. Note that you never subtract z-scores. So. . What area lies above a z-score of +1? If we look this z-score up on the table we find that the area in the body is . you will need to look up the positive z-value for 1.1587 is the proportion in the population that have a z = 1 or more. For our table. we will look up the tail area and subtract the two table values.8413 and the area in the tail is. Finding areas between two z-scores When we have two different z-scores and want to find the area between them. what is the area between Z = 1 and Z = 1.5? Again.5. If we look this zscore up on the table we find that the area in the body is . Since the z-score is positive and we want the area above the zscore. Now let’s consider the situation if the z-score is negative.1587. So.Finding areas above a z-score The process for this type of problem is the same as what we have already learned. we first must consider if both values are on the same side of the mean.9332 is the proportion in the population above a z-score of -1. or if one value is positive and the other negative. if the values are either both positive z’s or both negative z’s. Since the z-score is negative and we want the area above that point we will be using the body area. .0668.

the area we are looking for will cross the mean. then you subtract the two areas. Once you get these values off the table. Converting a Raw Score to a Z-score and then into an Area These problems are exactly the same as the others we have been working. For example let’s look at IQ scores for the population with a mean of 100 and standard deviation of 15 (µ = 100. It is equal to z = 1.5? Since one score is positive and the other negative. σ = 15). Once you get the two areas off the table.On the other hand. if you have one positive and one negative z-score then you must use the body area for either one of the z-scores. what is the area between Z = -1 and Z = 1. What proportion of the population will have an IQ of 115 or less? I first must compute Z. but now you must first compute the zvalue using the z-formula before using the table. Use the body area for one value and the tail area for the other. Now the question becomes what proportion of scores lie above z = 1? . You must still find areas above/below/and between two z-scores. and the tail area for the other. subtract them to find the area in between. For example.

edu/dwallace/sz-score2. Once we get the z-score we will plug in the values we know and solve for X in the z-score equation. We will start with a given area or proportion. and it might help to draw the distribution in order to be sure you are using the right column. σ = 15). when you use the z-table for these problems you must look up the area in either the body or tail column and then trace it back to find the z-score.html Converting an Area to a Z-score and then into a Raw Score For these problems we will be doing the same process we have been doing. Since we want the top 10%. Always use the z-score closest to the area of interest. You then use the z-table to find the area. You can look in either column. We won’t be computing z. However. but instead finding it from the table. We are solving for X now.10 is Z = 1.uncfsu.Please review other examples of this type of problem on the web-page. Prior examples were all asking for an area or proportion. The first step to solving this type of problem is to find the Z-score.28. .10 and the body is . but everything will be done in reverse order. what score cuts off the top 10% of the distribution? Notice that these questions are always asking what the score is for a certain point. For example. Once we have that number. Find the link to Z-scores and probability or click here http://faculty.90. we will be looking for the area on the table where the tail is . IQ scores for the population with a mean of 100 and standard deviation of 15 (µ = 100. Make sure you are not using the z-column at this point. We find the z-score that leaves the tail at the top of the distribution equal to .

you can use the following formula: X = Zσ + µ Special note for values in the lower 50% of the distribution: Whenever we want to find a z-score for a value below the mean. Recall that the z-table only gives positive z-values. Please review other examples of this type of problem on the web-page. Alternatively. if I were looking for the IQ score for the bottom 10% of the distribution. we must remember to make the value negative.html .90.28 even though the table shows only the positive value. if you have trouble with algebra. then I would have look up a tail area of . The zscore we need is -1.uncfsu. then you must remember to insert the negative sign before doing the computation. Find the link to Z-scores and probability or click here http://faculty. If the value is below the mean. For example.edu/dwallace/sz-score2.we can plug in what we know into the z-formula and solve for X.10 or body area of . in the above example.

we were only looking for one side in the last problem. If I draw a card from a deck of cards what is the probability it is a heart? P ( Heart ) = 13 1 = = 0.25. Most of the events we will be interested in are possible events. A certain event has a probability of one or 100% because it will always happen. Another experiment we could use to look at probabilities is drawing cards from a standard deck. out of a total of six sides. there are 26 red and 26 black cards. ¼ is 0. Always leave your answer in decimal form. What is the probability of rolling a ‘5’ on a single die roll? Most people know it is 1/6 or .25 52 4 Note that a standard deck of cards has 52 cards with 13 hearts/13 clubs/13 diamonds/13 spades. So. Probability ( P) = # _ items _ in _ criteria Total _# _ items So. Simply divide out any fraction you have by dividing the top number by the bottom number. Since diamonds and hearts are red cards and the rest are black. instead of fractional form.167 because there is only one side that is a ‘5’ and there are six sides on the die.Lesson 8 Probability Outline Probability of an Event Probability of Single Events Probability of Multiple Events -without replacement -mutually exclusive events Conditional Probability Probability of an Event There are three classes of events: Impossible Events--------------Possible Events-----------Certain Events P=0 P = 0 to 1 P=1 The probability (P) of an impossible event is zero because there is zero chance of it happening. A simple experiment we could run to examine probabilities is to roll a six-sided die. These probabilities will always have a value between zero and one. . You can determine the probability of any even in this manner.

what is the probability that I roll a ‘5’ or a ‘6’ on a single die roll or P (5 or 6)? With single events you will see this “or” connector and you will add the two individual probabilities. So: P (5 and 6) = 1/6 + 1/6 = . So.167 + .167 = .25 = . what is the probability of rolling a ‘5’ and a ‘6’ on two die rolls. For example.062 .028 We will only be dealing with independent events in this section.5 Probability of Multiple Events With multiple events we will be interested more than one outcome be realized.25 = .Probability of Single Events An individual event is a single event. For example. then. or events that do not affect the outcome of other events.167 = .334 What is the probability I draw a Heart or Club with a single draw from a standard deck of cards? P (Heart or Club) P(Heart or Club) = P(Heart) + P(Club) = 13/52 + 13/52 = .167 * .25 + . but we are still just going to roll the die once or draw a single card from a deck. When you see this “and” connector you will multiply individual probabilities. For example. what is the probability of drawing a Heart and a Club from standard deck? P(Heart and Club) = P(Heart) * P(Club) = 13/52 * 13/52 = . With single events we are measuring the likelihood of a one thing happening. P (5 and 6) = P(5) * P(6) = 1/6 * 1/6 = . With the card experiment. we will roll the die more than once or draw more than one card from a deck.25 * . To get both a five and a six I will have to roll the die more than once. we will not look at multiple draws where one draw could affect the probability of a separate event. We might be interested in different outcomes.

For these problems frequency data (or counts) will be given in a contingency table.S. and the rules we learned above will not apply.1176) Mutually Exclusive Events Mutually exclusive events are events that cannot happen together. 15 20 . Computer No Computer 30 10 Not in U. Try to find the probability of drawing three red cards from a deck without replacement. This table will display the frequencies for different combinations of events.S. Thus. and living the U. For mutually exclusive events the probability the two events will occur together must always equal zero. being a freshman and a sophomore are mutually exclusive. For example. Notice that when figuring how many total events there were we used 52 every time because we assumed each draw was from a fresh deck. (Answer: 0. If the problem. specifies that there is no replacement then we must take this into account when figuring the probabilities.064 We might also have to subtract a value from the numerator as well as the denominator. Thus: P(Heart and Club) = P (Heart) * P (Club) = 13/52 * 13/51 = . For example.Without Replacement Although we will focus on independent events like the last example. The above examples assumed that once we drew a card from the deck that it was replaced before another draw was made. these are not independent events. For example. Conditional Probability With conditional probabilities we will consider the probability of an event given that some other event has already happened.S. You are either one or the other but not both. what is the probability of drawing a Heart and a Club from a deck without replacement? When we count how many cards are left for the Club draw. we will also consider what happens to probabilities in situations in which there is no replacement.255 = . there will be one less card in the deck because we already had to draw the Heart from the deck. however. consider the probability of having a computer or not.25 * . In U. or elsewhere.

but instead we are restricting our sample to just those that live in the U.33 . We want to know what proportion have a computer out of these 40 people living in the U.S. So. For conditional probabilities the total or denominator is the value given. there is a 0. use the original table values out of the total.6 probability that a random draw will yield a person with a computer. For a single event in a table like this one. We could make this conditional by saying what is the probability of having a computer given that we know the person is living in the U. So.S. the conditional probability is 15/45 = . and divide by the total number in the sample.S.S. or 40. and have a computer.Before we consider conditional probabilities.S. what is the probability that choosing someone from our sample will yield a person with a computer? To answer this question we will need to add up the total for each row and column in the table: In U. 15 20 35 Total 45 30 Since there are a total of 45 people in our sample with a computer out of 75 total people. 30 out of the total of 75 people or 0. Again.75 In this same example what is the probability someone does not live in the U.S.S. we will use one of the original values.S.S. let’s look at some of the types of questions we have already examined. have a computer out of 40 the probability is 30/40 = . For this example it is the total for those living in the U. For conditional probabilities we will restrict our sample to those items given to have already have happened. Instead of looking in the total column for this type of problem. the vertical line means “given” and the second probability “computer” is what is given. For a combined event. for the entire sample. For example we might know the probability of having a computer and living in the U.? With the second question we are not asking the probability of picking a person at random from the total. our new total is those with a computer or 45. For example. | computer) Where the first probability is the one we are interested in. Now find the probability that a random draw will yield someone with a computer that is living in the U.S.S. Since 30 of those in the U. Computer No Computer total 30 10 40 Not in U. There are 30 people living in the U.S. that also have a computer.4 live in the U. given they have a computer? We can write: P (not in U. or 15. use the values in the margin or the totals.S. We want to know the proportion out of these that are not in the U.S.

was different than the survival rate of those who do not receive the new treatment. such as saying one treatment is better than another. even though we are talking about a sample. Again. That is. We will want to see if a value or sample comes from a known population.Lesson 9 Hypothesis Testing Outline Logic for Hypothesis Testing Critical Value Alpha (α) -level . start by converting the value to a z-score and find out how likely the score is for the known population. Logic 1) To determine if a value is from a known population. keep in mind that we will be inferring that what we observe with our sample is true for our population.01 One-Tail versus Two-Tail Tests -critical values for both alpha levels Logic for Hypothesis Testing Anytime we want to make comparative statements. Recall that inferential statistics is the branch of statistics in which we make inferences about populations from samples. for example. The population is either the one we already know about. if I were to give a new cancer treatment to a group of patients. Although we will start using just one value from the population and eventually a sample of values in order to test hypotheses. but instead comes from some other unknown population (the treatment had an effect). 2) If the value is “likely” for the known population then it is likely that it comes from the known population (the treatment had no effect). Up to this point we have been mainly concerned with describing our distributions using descriptive statistics. 3) If the value is “unlikely” for the known population then it is probably does not come from the population we know about. What we are testing then is whether the sample patients who receive the new treatment come from the population we already know about (cancer patients without the treatment). Hypothesis testing begins the section of the course concerned with inferential statistics. I would want to know if their survival rate.05 -level . Hypothesis testing is all about populations. we do it through hypothesis testing. or some new population (created by the new treatment in this example). . we infer that the sample is just part of an entire population.

Any value that occurs 5% or more of the time for the known population is “likely” and part of the known population. or does he come from a different alien population? To answer the question.” Z= . find out how likely this z-score is for the population. Bob has an IQ score of 115? Is it likely that he comes from the known population and is human.4) A value is “unlikely” if it is less than 5% likely for the known population. We want to know if Bob is an alien. Let’s say the average human IQ (the known population) is µ = 100 σ = 15. Let’s look at a simple example. Thus. or unlikely. we will be looking in the Tail Column when deciding if the value is “unlikely. and it will change as we progress. The only way to tell them apart from humans is to give them an IQ test since they are quite a bit smarter than the average human. For now we will use it as a starting point to illustrate several concepts. Say the earth has been invaded by aliens that look just like humans. 15 15 For hypothesis testing we will always be interested in whether the value is extreme for the population. first compute the z-score. 115 − 100 15 = =1 Next. The 5% cut-off point is rather arbitrary.

Use the z-table to determine the z-score that cuts off the top 5% of scores. That is. Since we will be doing hypothesis testing from this point on.0228. So. find out how likely this z-score is for the population. and will be one of the steps that must be performed anytime we conduct a hypothesis test. 15 15 So. and then find how likely it is to get that value or one more extreme. the likelihood of observing an IQ of 115 or more is . Since the probability is not less than 5% or . What z-score will be exactly 5% likely for any population? This is the z-score we will make comparisons against. many points on subsequent exams will come from just knowing the critical value.05 we have to assume Neil comes from a different population than the one we know about (the general population of humans). the likelihood of observing an IQ of 130 or more is . The reason for this fact is that the z-score that marks the point where a value becomes unlikely does not change on the z-scale. Since the probability is less than 5% or . I don’t have to look up each particular area when I compute my z-score. we can stop once we compute the z-score without reference to the z-table. Any z-score beyond that point is less than 5% likely. Say Neil has an IQ of X = 30. 130 − 100 30 Z= = =2 Next. Instead I only have to verify that the z-score I computed is more extreme than the one that is 5% likely. there is only one z-score at the that is 5% likely. Critical Value Critical Values are a way to save time with hypothesis testing.1587.So. We don’t really have to look up the probability of getting a particular value in order to verify it is less than 5% likely. Finding the critical value is important. .05 we have to assume Bob comes from the general population of humans. Thus. Is it likely Neil comes from the general population? Again we first compute the z-score.

If Z=2. Use the Tail column and find . If alpha is 1% then the critical value will be different than the one we found above.05) for that example. We . Alpha Alpha is the probability level we set before we say a value is unlikely for a known population.05 (α = . but you must use that information to determine the critical value.01. Alpha is given in every problem. If alpha is . instead of above it.01 or 1% as well. Z = 1. Thus. What z-score will be exactly 1% likely for any population? This is the z-score we will make comparisons against when alpha is set to 1% (α = . or Z =0.99.43 you conclude that the value is likely. You could also look in the Body Column and find .So. or Z=1. and so must be part of a different population.88 you conclude that the value is unlikely.01). the numbers will not change. Any value more extreme than 1. Alpha was .1. we will also use an alpha level of . Sometimes researchers want to be very sure before they decide a value is different.64 is the critical value. If we were interested in a value that was below the mean.5. and therefore part of a different population. It assumes that a value must be less than 5% likely to be unlikely. Use the z-table to determine the z-score that cuts off the top 1% of scores just like the last example.01. Since the distribution is symmetric. If Z = 1. and all other events will be likely for the known population. and so must be part of the known population. then we would flip our decision line to the other side. The critical value we just found is only one that we will use.64 is unlikely. then a value must be less than 1% likely before it is said to be unlikely for a known population. Note that we were only working on one side of the distribution in the above problem.

we want to be looking for “unlikely” events in both directions at the same time. We still want 5% of our events to be “unlikely” and 95% of our events to be “likely” for the known population. One-Tail versus Two-Tail Tests Another factor that will affect our critical value is whether we are performing a one or a two-tail test. leaving only . So. we will split the “unlikely” block into two parts.33 when Alpha is set to the 1% level.025 in the Tail Column of the z-table.58). I will discuss how to determine if a problem is a one or a two tail test in a later lesson.will use Z = 2. Can you find the critical value on the z-table (answer: z = 2. Let’s begin with an alpha level of 5%. What z-scores will then mark the middle 95% of our distribution? You will have to look up an area of . however. This time we will want 99% of our values in the middle.01 is the same.005 or half of one percent on each side. . The process for finding the two-tail critical values when alpha is set to . Also note that when we are interested in determining if values below the mean are unlikely our critical value will be negative. Now. With two-tail tests we will look for unlikely events on both sides of the mean (above and below) at the same time. each half the total 5% area. but let’s go ahead and find the critical values for two-tailed test the same way we did the one-tail tests above. The critical values we have looked at so far were for one-tail tests because we were only looking at one “tail” of the distribution at a time (either on the positive side above the mean or the negative side below the mean).

both positive and negative.58 Notice that you have two critical values for a 2-tail test.96 α = . we have learned four critical values.58/-2. You will have only one critical value for a one-tail test (which could be negative). 1-tail 2-tail α = .64 1.01 2.So. .33 2.96/-1.05 1.

That’s what we want the research hypothesis to say. then we indirectly support the research hypotheses since it competes directly with the null. Although we would like to directly test the research hypothesis. We will be writing two hypotheses: the research (H1) and the null (H0) hypothesis. Here we are trying to show that the genetically altered plant grows at a faster rate than unaltered plants. are we looking for differences in either direction at the same time (2-tail)? The above problem is one-tail since we are looking for a growth rate higher than the average.3 σ = 0.05. We have to be explicit about the scale on which we expect to find differences.2 cm per week (X = 3. Let’s take a look at a sample problem: Suppose some species of plants grows at 2.2). Look for words that indicate a direction in the problem for one-tail test (e. Again. Are we predicting the treatment population will be greater or less than the general population (1-tail)? Or. We will discuss this fact in more detail later in the lesson. 2) State the dependent variable. If we disprove the null. The new plant grows at 3. when you write your hypotheses.g.3 (µ = 2. .3 cm per week with a standard deviation of 0. one will be a treatment population and the other will always be the general population. we must first write the hypotheses in a formal way. be sure to include three elements: 1) explicitly state the populations you wish to compare. the research hypothesis matches the research question in the problem. For now. Let’s focus on writing hypotheses. Did the genetic alteration cause the plant to grow faster than the general population? Set alpha = . The null is a competing hypothesis. In order to write the research hypothesis look at what the researcher is trying to prove. higher/lower.Lesson 10 Steps in Hypothesis Testing Outline Writing Hypotheses -research (H1) -null (H0) -in symbols Steps in Hypothesis Testing -step1: write the hypotheses -step2: find critical value -step3: conduct the test -step4: make a decision about the null -step5: write a conclusion Writing Hypotheses Before we can start testing hypotheses.3). rather than any other steps we have learned for now. we actually test the null. However. more/less. 3) State the type or direction of the effect. I take a sample plant and genetically alter it to grow faster. The research hypothesis matches what the researcher is trying to show is true in the problem.

so we include left-over elements with the null. > 2. If they end up growing slower it won’t support the research hypothesis. For this example: H0: The population of genetically altered plants grows at the same or lower rate as the general population. It’s basically the opposite of the research hypothesis. We do this because we want to make inferences about the population. Our inferences will be that the entire population the plant comes from grows at a faster rate.better/worse). Notice that we state both the treatment population and the population we will compare that to. You could vary the wording a bit. the general population. In general it states that there is not effect for our treatment or no differences in our populations. The value of 2. The current example is easy to translate into a hypothesis. Although it is represented with a “mu” in the problem.3 H0: µgen. not the single value sample I am using to test the hypothesis. we don’t the . It would be two-tailed if the problem had stated that we expected a “different” growth rate than the general population. The two-tail null would say the groups are do not differ. The null hypothesis (denoted by H0) is a competing hypothesis. We only want to show that the treatment population grows faster. but check the homework packet because the wording is not always so obvious. We will restate both the null and research hypotheses in symbols we have been using for our formulas. substitute “different” for the word “faster” in the research hypothesis.alt < 2.3 Notice that we represent the treatment population with a “mu” (µ). and we indicate the direction by saying it will grow faster.3 is the general population mean we are comparing against.alt. Different could be higher or it could be different because it is lower. Thus: H1: µgen. I’ve included the “same or lower” wording for the one-tail test because we want to cover all the possible outcomes of the test. For the research hypotheses (denoted by H1 ): H1: The population of genetically altered plants grows faster than the general population. In Symbols We can also write the hypothesis in notational form. Growth rate is the dependent variable. as long as you include the three elements. For two-tail tests.

3 Step 4: Make a decision about the Null Reject the Null or Fail to Reject the Null (retain the null) are the only two possible answers here.3 0. Since the value we computed for the z-test is more extreme than the .05.3 H0: µgen.2 cm per week (X = 3. Let’s continue with the example we have already started: Suppose some species of plants grows at 2.alt < 2. The new plant grows at 3. H0: The population of genetically altered plants grows at the same or lower rate as the general population.symbol because we know the exact value for that population. Step 1: Write the hypotheses in words and symbols H1: The population of genetically altered plants grows faster than the general population. > 2.alt.3 (µ = 2.3 Step 2: Find the critical value for the test Since alpha is .64 Step 3: Run the test Here we find out how “likely” the value is by computing the z-score. so it is important to understand how one step follows from another now.2). and it is a one-tail test because we think our treatment will produce plants that grow “faster” than the general population: Zcritical =1.3 cm per week with a standard deviation of 0. For two-tail test we simply change the direction arrows to equal/not-equal signs (an “=” sign for the null and / “ ≠ “ sign for the research hypothesis).3).3 0. Z= 3.3 σ = 0. The steps will remain the same for each subsequent statistic we learn.9 = =3 0.2 − 2. I take a sample plant and genetically alter it to grow faster. Did the genetic alteration cause the plant to grow faster than the general population? Set alpha = .05. H1: µgen. Steps in Hypothesis Testing Now we can put what we have learned together to complete a hypothesis test.

then the then there was a difference (treatment had an effect).” Although we have a conclusion in step 4. but indulge me on this one. Instead. if we disprove the null. but we actually prove that it is not from the same population. we have: Note that we are testing the Null. . because you will notice that our decision is based on the statistical test that the treatment value is not likely to have come from the same population. We never prove the research hypothesis. We infer it is a different population. though not required for the answer. you can just rewrite it from Step 1). What did our test show? If you reject the null. It may seem like a matter of semantics. Step 5: Write a conclusion For this example. write a conclusion here in plain language without any statistical jargon. we conclude: “The population of genetically altered plants grows at a different rate than the general population. Graphically. The research hypothesis is your conclusion (you can simply restate it from Step 1). This is true. we reject the Null.critical value. we indirectly support the research hypothesis. If you fail to reject the null. then the null hypothesis is your conclusion (again. even thought that is our intent. It is either proven or disproven.

The rest of this lesson is devoted to the theory behind the changes we make when moving from tests with a single x-value to tests with samples of x-values. However. Often this will be given just like the x-value in prior problems. At a minimum you should be able to recognize the rules of the Central Limit Theorem for the exam (detailed below). however. the last page is devoted to the computations you will perform for the exams. we never test a hypothesis based on one individual from a population. Again. Links to these pages are provided below. these lessons contain conceptual information for the most part. you should be concerned with understanding the conceptual meaning of this lesson. There are no computational additions for the exam other than the formula change above. Instead. It is important to review each one. Sampling Distributions Standard Error Standard Error and Z-score Hypothesis Testing . z= X −µ σx . The formula we will want to use has a minor change from the one we have been using.Lesson 11 Hypothesis Testing with a Sample of Values We have looked at the basics of hypothesis testing using the z-formula we had already learned. Compute σ x first for the denominator by dividing the standard deviation by the square root of the given sample size (n). we will want to have a sample of values to test against the population. but now you may also have to compute it from the sample. the standard error page. and the standard error with hypothesis testing page. However. where σ x = σ n Notice that there is sample mean now in the numerator instead of just a single x-value. Take notes on the sampling distribution page. The lesson continues on the web page. Once you get that number plug it in as the denominator in the z-score formula.

and say the value comes from some other population. . this discussion on errors is strictly theoretical. there is always some chance that our decision is in error. we reject the null.Lesson 12 Errors in Hypothesis Testing Outline Type I error Type II error Power Examples in the “real world” Anytime we make a decision about the null it is based on a probability. Type I Errors Whenever a value is less than 5% likely for the known population. Thus. but not impossible. Notice that we are saying the value is really from another population distribution out there that we don’t know about. When running a test. Extreme values are unlikely for the population. I only know what my decision is about the test. and not the true state of reality. So. Recall that we reject the null when it tests a value that is unlikely for the known population. Note that we will never know whether we know we have made an error or not with our hypothesis test.

some of the time the value really does come from a different unknown population. some of the time the value really does come from the known population. We reject any values in this range. then. But. and say the value comes from the known population. but the value really does come from the known population a Type I error has been committed. even though they really are part of the known population. The part of the distribution that remains under the curve for the known population but is beyond our critical value in the region of rejections is alpha (α). Note that a Type I error can only occur when we reject the null. . A Type I error. Notice that even though the value represented is beyond the critical value it still lies under the curve for the normal population. Type II Errors Whenever a value is more than 5% likely for the known population. When we reject the null.However. we retain the null. When we set alpha we are setting the probability of making a Type I error. happens when we reject the null when we really should have retained it.

Notice that it is the area below the critical value. Power Power is the probability of correctly rejecting the null hypothesis. Although we know the probability of a Type I error because we set alpha. it is the probability of rejecting the null when it is really false.Notice in this situation the value is below the critical value. but that is still part of the other unknown distribution. the value is still under the unknown population distribution. when I should really have rejected it I commit a Type II error. Power is another way of talking about Type II errors. However. The probability of making a Type II error is equal to beta and not strictly defined by alpha. An easy way to remember all these concepts might be to put them in a table. That is. when I retain the null. However. so it is important to be aware of such concepts. Thus. much like your textbook does. Again. You can ignore the “power” demonstration on the web page for that reason. we will not be computing power in this course. You can see the region of Beta (β) below. a Type II error takes in a few more factors than that. . and may in fact come from the unknown population instead. we never really know if the null is false or not in reality. Such errors have been recognized as a problem in the behavioral sciences. so we retain the null.

The test is positive if what you are looking for is found. The test shows a negative result (what you are looking for is not there). when if fact you are pregnant. when in fact there is an effect. It is negative if the test shows what you are looking for is not there. The test shows a positive result (what you looking for is there). . and a Type II error is a false negative. Thus.Examples of Errors in the “Real World” Another way to think about Type I and Type II errors is to think of them in terms of false positives and false negatives. Since the man is sober. when in fact there is none. For example. but the test is false. but the test if false. A doctor tells you that you are not pregnant. You are positive for alcohol in that case. for this example. but you were not then it would be a false positive result. The opposite situation of the above example would apply. if a doctor told you that you were pregnant. Let’s look at another example. whether the test was positive or negative. The second piece of information is whether the test is in error or not (false or true test). What type of error has been committed (if any)? For this type of problem you will get two pieces of information. So. A sober man fails a blood alcohol test. the test is positive because if you fail a blood alcohol test it is showing that there is alcohol in your system. here we have a false positive test or Type I error. A false positive is when a test is performed and shows an effect. First. A false negative is when a test is performed and shows no effect. the test is false. A Type I error is a false positive.

The t-distribution There are several conceptual differences when the statistic uses the standard deviation from the sample instead of the population. a higher value on the scale will be needed to cut off just 5% of the distribution. It will tend to be flatter and more spread out than population distribution. Most of the time you will be given this value. Thus. when used to estimate a population in this way. and 2) the critical values will vary depending on the size of the sample we are using. That is. So. When we use the sample to estimate the population it will be much smaller than the population. we don’t know the mean and standard deviation of an entire population most of the time. that vary as a function of sample size. but in the homework packet there are problems where you must compute it yourself. and as usual whether it is one-tail or two-tail. The larger the sample size the more normal in shape the distribution will be. and so are not as “normal” in shape as a larger set of values would yield. Thus. The practical results of doing a t-test is that 1) there is a difference in the formula notation. and you use a different table to find the critical value.Lesson 13 Hypothesis Testing with the t-test Statistic Outline Unknown Population Values The t-distribution -t-table Confidence Intervals Unknown Population Values When we are testing a hypothesis we usually don’t know parameters from the population. Since the distribution is more spread out. Because of this fact the distribution will not be as regular or “normal” in shape. These critical values . The t-table Critical values for the t-test will vary depending on the sample size we are using. In fact. but instead of using the standard deviation from the population we use the standard deviation from the sample. Recall that the sample standard deviation is “S” and is computed with n-1 in the denominator (see prior lesson). the t-distribution is a family of distributions (like the zdistribution). the critical value that cuts off 5% of the distribution will be different than on the z-score. all the steps you have already learned stay the same. the t-test is exactly like the z-test computationally. but when you see that the problem gives the standard deviation from the sample (S) instead of the population (σ). X−µ s The formula is: t = . you write the formula with “t” instead of “z”. is computed differently than the standard deviation from the population. where sx = n sx The standard deviation from the sample (S). and due to the alpha level.

Since our decision to reject the null means that there are two populations instead of just the one we know about. Notice that we have one and two-tail columns at the top and degrees of freedom (df) down the side. confidence intervals give us an idea about the mean of the new unknown population. See page A27 in your text.uncfsu. the t-distribution approaches the z-distribution.are in the Appendices in the back of your book. See the Confidence Interval demonstration on the web page or click here http://faculty. Confidence intervals are a way to estimate the parameters of the unknown population. . Confidence Intervals If we reject the null with our hypothesis test. Note that this is a table of critical values rather than a table of areas like the z-table. Cross index the correct column with the degrees of freedom you compute. If you look at the bottom row (at the infinity symbol) you will see all the critical values for the z-test we learned on the last exam. Degrees of freedom are a way of accounting for the sample size. For this test df = n – 1.edu/dwallace/sci. we can compute a confidence interval. that as n approaches infinity. Also note.html for the rest of the lesson.

recall. So for example. since we no longer no any population values we will use “mu” to represent both populations. where sx = n sx The numerator will now have two sample values ( X 1 − X 2 ) instead of one sample and one population. Changes in Hypotheses All hypotheses from this point on in the course will be two-tailed. or new techniques for improving depression all involve testing a population created by the treatment or drug or technique. We will compute the standard error separately for each sample and then add them together. New treatments for diseases. Now that we have two samples we will want to include the estimate of variability from both. with the independent samples t-test we will compare two sample values directly. This fact is important because when we test hypotheses we are usually testing an idea and a population that we know nothing about. H0: µdiet = µplacebo H: µ ≠ µ 1 diet placebo X−µ s . Recall the formula for the t-test we have been using: t = Formula Changes . Thus. Note that we are still making the inference about the populations from which the samples are drawn. is the standard error (the standard deviation divided by the square root of the sample size).Lesson 14 Independent Samples t-test Outline No Population Values Changes in Hypotheses Changes if Formula -standard error Pooled Standard Error -weighted averages Critical Values -df Sample Problem No Population Values With the independent samples t-test we finally reach the point where we have no population values. Think about the kinds of scientific discoveries you hear about often. new drugs. The denominator. we will have to take into account the standard deviations and sample sizes of both samples. So. s Our standard error (denominator) was: sx = Remember that the standard error n measures variability we expect to see among samples. In addition.

t= (X1 − X 2 ) where s X s X1 − X 2 1 −X = 2 2 s1 s2 + 2 n1 n 2 Pooled Standard Error Note that the formulas I present in this section differ from your text! The above formula is useful when our sample sizes are the same. Instead we have to give more weight to the larger sample. Again. it will be easier if we switch from using the standard deviation to the variance.5 = 8.5 N = 100 S = 15 Although this example is extreme.5 = 16.25 2 . However.Because of the formula we will develop. Simple average of two groups: 15 + 1. Weighted Averages Let’s say I have one sample with Another sample has: N = 20 S = 1. In this way we can eliminate the radical in the denominator. we cannot simply add the two standard errors together.5 16. If we did that we would have the average of two groups rather than the average of all 120 people. The two formulas are equivalent: s n = s2 n Since we are adding the two separate standard errors together we have: s X1 − X 2 = 2 s12 s 2 + n1 n2 Notice that we now denote the combined standard error with ( s X 1 − X 2 ). you can see that you would not want to simply average the two groups together in order to get the average of S. in situations where our sample sizes are different. it’s just a way to symbolize the final value we will divide into the numerator.

5) than the smaller group (8. and in the final step we divide by the total number of people. so: df = n1 + n 2 − 2 . but now denote the variance as pooled. When we have unequal sample sizes we will want to use a similar process to average or pool the variances from our two samples. However degrees of freedom are now computed from two samples. (X − X 2 ) t= 1 s X1 − X 2 sX 1 −X = 2 s2 p n1 + s2 p n2 Critical Values We will use the same table to find the critical values as we did with the one-sample t-test.75 100 + 20 120 120 For the weighted average we are multiplying each variance times the sample size to get a sum of all 120 people.25). Below is the formula that does just that. Once we compute that value we plug it into the same formula we used with equal sample sizes. s2 p = (n1 − 1)s12 + (n2 − 1)s 22 n1 + n2 − 2 Here s 2 p is the symbol we use for the pooled variance.5(20) 1500 + 30 1530 = = = 12. The value n-1 or degrees of freedom is used to represent the sample size. Notice that we are doing the same process we used for the weighted average above. You can see that if I have 100 people with such a large variance that the average of those people plus 20 more of them with a small standard deviation should yield a value closer to the larger group (12. We multiply the variances times the sample size and divide by the total number of people.Weighted average of 120 people: 15(100) + 1.

Sample Problem A new program of imagery training is used to improve the performance of basketball players shooting free-throw shots.33 2 S2 = 39. ≠ µno imagery H0: µimagery = µno imagery Step 2: Find the critical value for the test Since alpha is .46 n=6 X2 5 6 10 15 18 20 X 2 = 12. A second group received no special practice. Did the imagery training make a difference? Set alpha = . and it is a two-tail: tcritical = ±2.05.05. so you just have to plug and compute. and then shot 30 free throw basket shots with the number of shots made recorded. The first group did an hour imagery practice. X1 15 17 20 25 26 27 X 1 = 21. H1: µimagery.228 Step 3: Run the test Since we have equal sample sizes (n’s) for each group we can use the first (shorter) formula: (X − X 2 ) s2 s2 t= 1 where s X − X = 1 + 2 s X1 − X 2 n1 n 2 1 2 All the values are given above.46 n=6 Step 1: Write the hypotheses in words and symbols H1: The population receiving imagery practice will make a different number of baskets than the population receiving no imagery practice.66 S12 = 25. . The data are below. H0: The population receiving imagery practice will make a different number of baskets than the population receiving no imagery practice. and also shot 30 free throw basket shots.

33) 9.28 6 6 t= (21.28 Note that we could have used the longer formula here as well because it will work for equal or unequal sample sizes.46 + = 4.33 = = 2. Step 5: Write a conclusion The population of players with imagery training made a different number of baskets compared to those with no training.24 + 6. .57 = 10.82 = 3.84 3.46 39. Step 4: Make a decision about the Null Reject the Null since the value we computed in Step 3 is more extreme than the critical value in Step 2.66 − 12. we reject the idea that they are from the same population.s X1 − X 2 = 25.28 3.

With ANOVA or analysis of variance. What we are doing when we compute between and within variability is to partition the total variability into the between and within components. how do we measure how different scores are in the distribution from one another? You should know that we use variance as a measure of variability. The table is discussed in the example below.” The null will still state that there are no significant differences between any of the groups (insert as many “mu’s” as you have groups). Computation How do we measure variability in a distribution? That is. with ANOVA we are testing hypotheses that involve comparisons of two or more populations.html Between group variability and within group variability are both components of the total variability in the combined distributions. Thus. or if some or all of the groups differ. however. H0: µ1=µ2=µ3 Critical values are found using the F-table in your book.edu/dwallace/sanova. I suggest using the notation I have outlined here since it will coincide more with what we have already done. So: Between variability + within variability = total variability Hypothesis Testing Again. Use whatever method you find easiest to understand. but you might look at the text version as well.Lesson 15 ANOVA (analysis of variance) Outline Variability -between group variability -within group variability -total variability -F-ratio Computation -sums of squares (between/within/total) -degrees of freedom (between/within/total) -mean square (between/within) -F (ratio of between to within) Example Problem Note: The formulas detailed here vary a great deal from the text. the test will not specify which two. Because of this fact the research hypothesis will state simply that “at least two of the means differ. will indicate a difference between any of the groups. we will conduct a separate test to determine which specific means differ. Instead. The overall test. we compute . Variability Please read about this topic on the web page by locating the ANOVA demonstration or you can click here: http://faculty.uncfsu.

Group two is labeled with a “2”. and then divide by the sample size (n -1 or degrees of freedom for a sample). This notation indicates that you continue to find the sums of squares as you did for the first two groups for however . we will still compute the sums of squares total (all values) for completeness. +  X 2 − (∑ X k )  ∑ k nk   SSWITHIN     Notice that each segment is the same formula for sums of squares we used in the formula for variance and for the total sums of squares above. we are computing variance. Don’t let the formulas intimidate you.a ratio of variances: between to within variance. we will do the exact same thing: compute the sums of squares and divide by degrees of freedom. but notice that after that we have group “k” instead of a number. Keep in mind that all we are doing is finding the variance for our between factor and dividing that by the variance for the within factor. Recall that variance is the average square deviation of scores about the mean. What is different here is that you consider each group separately. While we will only use the between variance and within variance to compute the Fration. Total Sums of Squares (∑ XTOT )2 2 SS TOT = ∑ XTOT − Note that it is the same formula we have been using.. s = 2 ∑X 2 (∑ X ) − n−1 n 2 ← Sums _ of _ Squares ← deg rees _ of _ freedom When we compute the Mean Square (variance) in order to form the F-ratio. These two variances will be computing by finding each sums of squares and dividing those sums of squares by their respective degrees of freedom. We will compute the same value here. Within Sums of Squares  (∑ X 1 )2 2  = ∑ X1 −  n1  2    +  X 2 − (∑ X 2 )  ∑ 2 n2   2    + . Recall that when we compute variance we first find the sum of the square deviations. Sums of Squares We will use the same basic formula for sums of squares that we used with variance. It indicates that you perform the operation for ALL values in your distribution (all subjects in all groups). but as the definition suggests. So. The N TOT subscript (tot) stands for the total. the first segment with the subscript “1” means you compute the sum of squares for the first group.. it will be called the “mean square” for the computations. So.

All we are doing is matching up degrees of freedom with the Sums of squares to get the mean square (variance) Within Mean Square SS MS Beteween = Between df Between Between Mean Square SS MSWITHIN = Within df Within . + − 2 k TOT 2 nk N TOT We have the same “k” notation here. so will use degrees of freedom within along the left margin of the table. Degrees of freedom. so we will use the degrees of freedom between along the top of your table. Don’t let the formula’s intimidate you..05 and A-30 for alpha . is different for each source of variability. The denominator of the F-ratio is the within subjects factor. Between Sums of Squares SS BETWEEN (∑ X ) (∑ X ) = + 2 1 2 2 n1 n2 (∑ X ) (∑ X ) + .. you perform the same operation for each separate group in your problem. divide the values by degrees of freedom in order to get the two mean square values we need to form the F-ratio. So.01). Total Degrees of freedom N – 1 this N value is the total number of values in all groups Within Degrees of freedom df within = N − K K is the number of categories or groups. N is still the total N Between Degrees of freedom df Between = K − 1 We will also use degrees of freedom to locate the critical value on the F-table (see page A-29 for alpha . This operation is half the sums of squares we computed for the sums of squares total. “k” could be the third group. with this formula once we compute the value for each group we must subtract an operation at the final step. Again. however. we will first compute the sums of squares for each source of variance. Mean Square Now we divide each sums of squares by the respective mean square.many groups you have in the problem. or if you have four groups then you would do the same sums of squares computation for the third and fourth group. Degrees of Freedom Again. However. The numerator of the F-ratio is the between factor.

Subjects are randomly assigned to one of three treatment groups. for some problems you might have to compute the sum of x. Test for a difference at α = .6 = 34. Below are the rated fear of spiders after therapy.88 STEP 3: Compute the appropriate test-statistic. find the critical value. MS Between F= MSWithin Example A therapist wants to examine the effectiveness of 3 therapy techniques on phobias. That is. Although in this example I have given the summary values. df Between = K − 1 = 3-1 = 2 df within = N − K = 15-3 = 12 Fcritical = 3. SS TOT = ∑ X 2 TOT (∑ X ) − TOT 2 N TOT ( ) = SS TOT = 107 − 33 = 107 − 1089 = 107 − 72. and sum of squared x’s yourself.05 Therapy A 5 2 5 4 2 Σx1 = 18 Therapy B 3 3 0 2 2 Σx2 = 10 2 =26 Σx2 Therapy C 1 0 1 2 1 Σx3 = 5 2 Σx3 =7 Σx12 = 74 STEP 1: State the null and alternative hypotheses.4 15 15 2 . H1 at least one mean differs H0: µ1 = µ2 = µ3 STEP 2: Set up the criteria for making a decision.F-ratio The final step is to divide our between by within variance to see if the effect (between) is large compared to the error (within). You might do this step after Step 3 since that is where you compute the critical value.

Notice that once we get the Sums of Squares on the table.8 + 20 + 5 − 72.6 = 17. we will divide those values by the df in the next column. + − 2 2 k TOT 2 + (5) 5 2 − (33) 15 nk N TOT 2 SS BETWEEN = SS BETWEEN 324 100 25 1089 + + − 5 5 5 15 = 64.6 2 MSWITHIN SSWithin df Within 17. very often we place them in a source table (below). Putting the values in a table like this one may make it easier to think about the statistic.SSWITHIN SSWITHIN SSWITHIN   (∑ X 1 )2  (∑ X 2 )2 2 2    X − + = ∑ X1 −  ∑ 2  n1 n2       (18)2  (10)2  (5)2    + 7 −  +  26 − 74 = −       5 5 5       324   100   25   =  74 −  + 7 −   +  26 − 5   5   5   2    + .6 F= =6 1..2 SS BETWEEN (∑ X ) (∑ X ) = + 2 1 2 2 SS BETWEEN = (18) 5 n1 2 + (10) 5 n2 (∑ X ) (∑ X ) + .2 = = 1..2 = = 8. Once we get the two mean squares we divide those to get F.2 Note that anytime you compute two of the Sums of Squares you can derive the third one without computation because Between + Within = Total df within = N − K df tot = N − 1 df Between = K − 1 df tot = 15 − 1 = 14 MS Beteween = df Between = 3 − 1 = 2 MSWITHIN = df within = 15 − 3 = 12 MS Between MSWithin 8..43 12 Once we have computed all the values.43 F= MS Beteween SS Between df Between 17.2 + 6 + 2 = 17. +  X 2 − (∑ X k ) ∑ k  nk       SSWITHIN = (74 − 64.8) + (26 − 20) + (7 − 5) SSWITHIN = 9. ..

what is your conclusion? There is at least one group that is different from at least one other group.4_ df _ 2__ _12__ _14__ MS __8.2 _17.43_ F _6_ STEP 4: Evaluate the null hypothesis (based on your answers to the above steps).6_ __1. .Source Between (group) Within (error) Total SS 17. Reject the null STEP 5: Based on your evaluation of the null hypothesis.2_ _34.

Here I show the groups. Therapy A 5 2 5 4 2 X 1 = 3. Cross index the row and column to find the value you need to put in the formula above. If you fail to reject the null. and n is the number of values we are n dealing with in each group (not total n). and degrees of freedom within is down the side. To find “q” or the studentized range statistic.6 Therapy B 3 3 0 2 2 X2 = 0 Therapy C 1 0 1 2 1 X3 =1 The first step is to compute all possible differences between means: . refer to your table on page A-32 of your text. Do not conduct a post-hoc test unless you found an effect (rejected the null) in the ANOVA problem. then there are no differences to find. On the table ‘k’ or the number of groups is found along the top. but have computed the average or mean of each group. The Mean Square value is from the ANOVA you already computed. HSD = q Example This example is a continuation of the ANOVA problem we did in the last lesson.Lesson 16 Post-hoc Tests Outline Tukey’s HSD Post-hoc test -differences between means -studentized range statistic (q) -honestly significant difference (HSD) Example Tukey Problem Magnitude of the Effect -eta-square -omega-square Tukey’s HSD Post-hoc test A post-hoc test is needed after we complete an ANOVA in order to determine which groups differ from each other. The critical value in this case is the HSD (honestly significant difference) and it must be computed. We will compare this difference score to a critical value to see if the difference is significant. MSWithin Note that “q” is a table value. It is the point when a mean difference becomes honestly significantly different. For the Tukey’s post-hoc test we will first find the differences between the means of all of our groups.

4 due to the type of therapy they received.2 − (2)1.43 HSD = q = = 3. so. η2 = SS Between Notice it is just a ratio of treatment effect variability to total variability SS total For our example: η2 = 17. One problem with eta-square is that it is a biased estimate and tends to overestimate the effect. a separate computation is needed to get a true idea of the strength of the relationship.83 .2 − (3 − 1)1.77(.99 n 5 Now we will compare the difference scores we computed with the HSD value. ω2 = SS Between − (k − 1)MSWithin SS total + MSWithin ω2 = 17.86 14.77 3. then we say the difference is significant.77 . Next we compute HSD.43 35. MSWithin 1. If we know what proportion of variability of the total is due to our treatment.50 50% of the variability in rated fear of spiders is 34. Magnitude of the effect is the amount of variability that our independent variable can account for in the dependent variable.6 X 1 − X 3 = 3.83 35. Groups 1 and 2 differ Groups 1 and 3 differ Groups 2 and 3 do not differ Magnitude of the Effect It is a common misconception that the size of the F-ratio you compute directly indicates how strongly the relationship is between the independent and dependent variable. If the difference is larger than the HSD.6 X 2 − X 3 = 0 − 1 = −1 We will only be concerned with the absolute difference. Our test indicates only that there is difference in the treatments. and it is not a measure of a relationship between variables. The easiest measure of the effect is eta-square (η2).2 = . A more accurate measure of the effect is omega-square (ω2). we will have a rough idea of how strong the relationship might be.X 1 − X 2 = 3.4 + 1.286 = 3.43 17.43 17.40 34. you can ignore any negative signs.2 − 2. However.83 35.34 = = = = . It measures the proportion of our between factor variability to the total variability.6 − 1 = 2.53) = 1.6 − 0 = 3.

.

Lesson 17 Pearson’s Correlation Coefficient Outline Measures of Relationships Pearson’s Correlation Coefficient (r) -types of data -scatter plots -measure of direction -measure of strength Computation -covariation of X and Y -unique variation in X and Y -measuring variability Example Problem -steps in hypothesis testing -r2 Note that some of the formulas I use differ from your text. Both sets of formulas are in the homework packet, and you should use the formulas you feel most comfortable using. Measures of Relationships Up to this point in the course our statistical tests have focused on demonstrating differences in effects of a dependent variable by an independent variable. In this way, we could infer that by changing the independent variable we could have a direct affect on the independent variable. With the statistics we have learned we can make statements about causality. Pearson’s Correlation Coefficient (r) Types of data For the rest of the course we will be focused on demonstrating relationships between variables. Although we will know if there is a relationship between variables when we compute a correlation, we will not be able to say that one variable actually causes changes in another variable. The statistics that reveal relationships between variables are more versatile, but not as definitive as those we have already learned. Although correlation will only reveal a relationship, and not causality, we will still be using measurement data. Recall that measurement data comes from a measurement we make on some scale. The type of data the statistic uses is one way we will distinguish these types of measures, so keep it in mind for the next statistic we learn (chi-square). One feature about the data that does differ from prior statistics is that we will have two values from each subject in our sample. So, we will need both an X distribution and Y distribution to express two values we measure from the same unit in the population. For

example, if I want to examine the relationship between amount of time spent studying for an exam (X) in hours and the score that person makes on an exam (Y) we might have: X
2 3 3 4 5 6 7

Y
65 70 75 70 85 85 90

Scatter plots An easy way to get an idea about the relationship between two variables is to create a scatter plot of the relationship. With a scatter plot we will graph our values on an X, Y coordinate plane. For example, say we measure the number of hours a person studies (X) and plot that with their resulting correct answers on a trivia test. (Y). X 0 1 1 2 3 4 5 Y 0 1 2 3 5 5 6

Plot each X and Y point by drawing and X,Y axis and placing the x-variable on the xaxis, and the y-variable on the y-axis. So, when we are at 0 on the X-axis for the first person, we are at 0 on the y-axis. The next person is at 1 on the X-axis and 1 on the Yaxis. Plot each point this way to form a scatter plot.

7 Number of Correc Answers 6 5 4 3 2 1 0 0 2 4 6 Number of Hours Studying

In the resulting graph you can see that as we increase values on the x-axis, it corresponds to an increase in the y-axis. For a scatter plot like this one we say that the relationship or correlation is positive. For positive correlations, as values on the x-axis increase, values on y-increase also. So, as the number of hours of study increases, the number of correct answers on the exam increases. The opposite is true as well. If one variable goes down the other goes down as well. Both variables move in the same direction. Let’s look at the opposite type of effect. In this example the X-variable is number of alcoholic drinks consumed, and the Y-variable is number of correct answers on a simple math test.
12 10 8 6 4 2 0 0 2 4 Number of Drinks 6 8

Number of Correct Answers

the values on Y decrease. and it would be a positive relationship because the sign is positive. They also give a good idea of how strongly related two variables are to one another. number of correct answers decreases. So. we may express a -1 value indicated a strong correlation because of the number and a negative relationship because of the sign. A value of +. . We will express the strength of the relationship with a number between 0 and 1. A zero indicates no relationship. as number of drinks consumed increases. Note that the number is independent of the direction of the effect. Measures of Strength Scatter plots gave us a good idea about the measure of the direction of the relationship between two variables.03 would be a weak correlation because the number is small. the stronger the relationship. Notice in the above graphs that you could draw a straight line to represent the direction the plotted points move. Most values will be a decimal value in between the two numbers. So. As the values on X increase. and a one indicates a perfect relationship.This scatter plot represents a negative correlation. Here are some more examples of scatter plots with estimated correlation (r) values. The variables are moving in opposite directions. Number of Correct Answers 12 10 8 6 4 2 0 0 2 4 Number of Drinks 6 8 The closer the points come to a straight line.

A

B

C

Graph A represents a strong positive correlation because the plots are very close together (perhaps r = + .85). Graph B represents a weaker positive correlation (r = +.30). Graph C represents a strong negative correlation (r = -.90). Computation When we compute the correlation it will be the ratio of covariation in the X and Y variable, to the individual variability in X and the individual variability in Y. By covariation we mean the amount that X and Y vary together. So, the correlation looks at the how much the two variables vary together relative to the amount they vary individually. If the covariation is large relative to the individual variability of each variabile, then the relationship and the value of r is strong.

A simple example might be helpful to understand the concept. For this example, X is population density and Y is number babies born. Individual variability in X You can think of a lot of different reasons why population density might vary by itself. People live in more densely populated areas for many reason including job opportunities, family reasons, or climate. Individual variability in Y You can also think of a lot of reasons why birth rate may vary by itself. People may be influenced to have children because of personal reasons, war, or economic reasons.

Covariation of X and Y For this example it is easy to see why we would expect X and Y to vary together as well. No matter what the birth rate might happen to be, we would expect that more people would yield more babies being born. When we compute the correlation coefficient we don’t have to think of all the reasons for variables to vary or covary, but simply to measure the variability. How do we measure variability in a distribution? I hope you know the answer to that question by now. We measure variability with sums of squares (often expressed as variance). So, when we compute the correlation we will insert the sums of squares for X and Y in the denominator. The numerator is the covariation of X and Y. For this value we could multiply the variability in the X-variable times the variability in the Y-variable, but see the formula below for an easier computation.

r=

 ∑ X 2 −  

∑ XY − (∑ X )
n

∑ X ∑Y
2 2  ( Y)  ∑ 2   ∑ Y − n     

n

The only new component here is the sum of the products of X and Y. Since each unit in our sample has both and X and a Y value, you will multiply these two numbers together for each unit in your sample. Then add the values you multiplied together. See the example below as well. Example Problem The following example includes the changes we will need to make for hypothesis testing with the correlation coefficient, as well as an example of how to do the computations. Below are the data for six participants giving their number of years in college (X) and their subsequent yearly income (Y). Income here is in thousands of dollars, but this fact does not require any changes in our computations. Test whether there is a relationship with Alpha = .05. # of Years of College Income X Y 0 15 1 15 3 20 4 25 4 30 6 35 ΣX = 18 ΣY = 140 ΣX 2

X2 0 1 9 16 16 36 = 78

Y2 225 225 400 625 900 1225 2 ΣY = 3600

XY 0 15 60 100 120 210 ΣXY = 505

Notice that I have included the computation for obtaining the summary values for you for completeness. Be sure you know how to obtain all the summed values, as they will not always be given on the exam. Step 1: State the Hypotheses in Words and Symbols H1 The correlation between years of education and income is equal to zero in the population. H0: The correlation between years of education and income not equal to zero in the population. As usual the null states that there is no effect or no relationship, and the research hypothesis states that there is an effect. When we write them in symbols we will use the Greek letter “rho” (ρ) to indicate the correlation in the population. Thus: H1 ρ ≠ 0 H0: ρ = 0 Step 2: Find the Critical Value Again, we will use a table to find the critical value in Appendix A of your book. Locate the table, and find the degrees of freedom for the appropriate test to find the critical value. For this test df = n – 2, where n is the number of pairs of scores we have. Df = 6 – 2 = 4 rcritical = + 0.811 Step 3: Run the Statistical Test

r=

 ∑ X 2 −  

∑ XY − (∑ X )
n

∑ X ∑Y
2 2  ( Y)  ∑ 2  ∑ Y −  n     

n

r=

(18)(140) 6 2  18   140 2  78 −  3600 −  6  6   505 −

33) = 85 7999. the more the subsequent income. we reject the null. we will compute the percentage of variability in Y. r2 Often times we will square the r-value we compute in order to get a measure of the size of the effect.92 = 85 = .95 89. The more years of school. Just like with eta-square in ANOVA. so 90% of the variability in income is accounted for by education.67] 85 (24)(333. that is accounted for by X. Step 5: Write a Conclusion There is a relationship between years spent in college and income. For the current example r2 = .r= 505 − 2520 6 324   19600   78 − 3600 −    6  6    r= [78 − 54][3600 − 3266.44 505 − 420 r= Step 4: Make a Decision about the Null Reject the null Since the value we computed in Step 3 is larger than the critical value in Step 2. .90.

We will refer to X as the predictor variable. our equation for the line will be slightly different: ˆ = bX + a Even though we use “b” here. but instead be the value we would predict for Y. we will express the line we come up with in an equation. though our Y-value will not be the same value we have in our data. You may recognize Y = mx + b. Thus. Even though we are not inferring a causal relationship. We will use a similar equation to express a straight line. but instead of measuring the relationship we will make predictions based on the relationship. and Y as the criterion variable. and plug them into our equation. We can then plug in an X-value and get out a predicted Y ˆ ). That is. As with similar topics in geometry. that is exactly what we want to do. we can nevertheless predict one variable if we have information about the other.Lesson 18 Regression Outline Equation of a Line -Slope -Y-Intercept -regression line Making Predictions Error in Prediction -residual variation -standard error of the estimate Regression is very similar to correlation. we will attempt to use X in order to predict what Y should be. value ( Y Slope Slope is the unit change in Y for each single unit change in X. Since our data are scattered about. Equation of a Line Recall that when we were first looking at scatter plots and we drew a line through the dots in order to indicate the direction of the effect. Thus. since we will multiply the slope with the X-value we want to make predictions about. it is still the slope of the line. With regression. We will compute the best fitting line for the data that we have. and “a” is the Y ˆ instead of just Y to indicate that this is a predicted value for y-intercept. but we will construct a line that simultaneously comes as close to each data point as possible. We also use Y way based on the regression line rather then an actual Y-value. it is unlikely that the value our line would predict for Y is actually a point in our data set. we will compute these regression coefficients (b and a). In order to make predictions. In a scatter plot a single line will not hit every data point. the predicted Y will change by the amount of the slope for each single unit of X. .

Let’s look at an example problem. For this problem we will continue with the example we used with correlation. The denominator is the sums of squares for X.Slope computation is very similar to correlation.72 76. First compute slope and y-intercept. we will compute the equation of the line that predicts income from number of years in college.54(18) 140 − 63.28 a= = = = 12.54 b= = 2 324 78 − 54 24 18 78 − 78 − 6 6 140 − 3. so the entire numerator will be the same as with r. we can put them into our equation for a line and start making predictions. This was a portion of the denominator for r as well. a= ∑ Y − b∑ X Notice that we must compute the slope “b” before we can compute Making Predictions Once we have computed the regression coefficients. Y-Intercept The y-intercept is the point at which the regression line crosses the Y-axis. however. That’s because we are at the Y-axis when X=0. − n the y-intercept. b= ∑ XY − ∑X 2 ∑ X ∑Y n (∑ X )2 Notice that we have the covariation of X and Y in the n numerator.7 6 6 6 505 − . if you have already computed the correlation you can use the same values to compute slope. # of Years of College Income X Y 0 15 1 15 3 20 4 25 4 30 6 35 ΣX = 18 ΣY = 140 ΣX 2 X2 0 1 9 16 16 36 = 78 Y2 225 225 400 625 900 1225 ΣY 2 = 3600 XY 0 15 60 100 120 210 ΣXY = 505 Now. 2520 (18)(140) 505 − 6 6 = 505 − 420 = 85 3. In fact. It is also the value we predict for Y when X = 0.

If we average this residual variation for all of our scores. but a prediction on the line that best describes the relationship between X and Y.7 16. Note again that this equation will not yield an actual y-value.7 Y ˆ =30. we can get a measure of the error our equation yields.54X + 12.86 26. So. then we can see how different the Y and predicted Y-values differ.4 Y Here we just insert the X-value.54X + 12.7 Y . # of Years of College X 0 1 3 4 4 6 Income Y 15 15 20 25 30 35 Ŷ 12. and compute Y.32 26. we expect someone with five years of education to make about 30 thousand dollars a year. we can write the equation of the line.24 23.54(5) + 12. Error in Predictions The amount our predicted Y-value differs from the actual Y-value in our data is error or residual variation. If we first plug in each X-value from our data into the regression line to get out a predicted Y-value.7 + 12.94 ˆ =3. ˆ =3.7 Y ˆ =17.86 33. what income would we predict for someone with five years of education? ˆ =3.7 Y Now we simply plug in an x-value that would like to make a prediction about.Once these values are computed. The standard error ( S y − y ˆ ) is the average deviation between actual Y and predicted Y. For example.

86 3. So. the standard error of the estimate we are calculating is the standard deviation of the regression line we computed. Once we find this difference we want to add the differences to get an average.32 -1.Ŷ 12. divide by df.14 33.12 2 ˆ ) = 32.46 9.86 3.85 6−2 4 Our interpretation is that on average actual Y and predicted Y differ by 2.94 Y.3 16.94 1.14 1. all we have left to do is square our residual variations.47 Σ(Y − Y Plugging into the formula we have: S y− y ˆ = 32. # of Years of College X 0 1 3 4 4 6 Income Y 15 15 20 25 30 35 Ŷ Y.7 2.06 ˆ) = 0 Σ(Y − Y (Y. .86 -1.47 32. # of Years of College X 0 1 3 4 4 6 Income Y 15 15 20 25 30 35 Ŷ 12.86 33. we will square the difference score we computed in the last step so that we can add the squared differences.32 26.Ŷ 2.54 11. The formula we are working toward looks like this: S y− y ˆ = ∑ (Y − Yˆ ) n−2 2 So. So.Remember the residual variation is the difference between Y and Ŷ.24 -3.32 26.3 -1.02 3.850. The sum of the deviations about the mean must always equal zero.32 -3.86 26.85 units.12 = 2.86 26.Ŷ)2 5.29 1.24 -1.24 23. In fact.06 ˆ) = 0 Σ(Y − Y Notice that we have the same problem here as with standard deviation. and take the square root of the whole thing.7 16.24 23. sum them. any prediction we make about income using the equation will be off by about $2.47 = = 8.86 1.

I might have the following data: High Performance 20 Compact 14 Mid Size 7 Full Size 9 Notice that our data is now frequency values or how many values in our sample fit into different categories. When analyzing categorical data we say the test is non-parametric. Thus. we will want to analyze categorical or qualitative data as well. If there really was no difference in the frequencies for each level of the variable. all the tests we have learned before this point were parametric tests. What we want to know is if they differ from the frequencies we would observe by chance. Since there a total of 50 claims in our sample. Chi-Square Goodness-of-Fit Test We will learn two different Chi-square tests.Lesson 19 Chi-Square Outline Categorical Data Goodness of Fit Test -observed frequency -expected frequency -X2 statistic Example -hypothesis testing Categorical Data As mentioned at the start of the lesson with correlation. Often times. The test will tell us whether there is a difference in how many values fall at different levels of the single variable (car type). we will analyze frequencies or counts of people falling into different categories or groups. Instead. The values we would expect if there really was no difference in the number of claims made for different car types are what we call the expected frequencies ( f e ). however. then we would expect equal numbers of claims for each car type. The first of these is the goodness-of-fit test. and . We actually took measurements from units in our sample to create our distribution. With the goodness-of-fit test we will test whether the data “fit good” with what we would expect if only chance factors were operating. if I measured the number of insurance claims for different car types. Is there a difference in number of claims for different car types? The values we observe in our sample are the observed frequencies ( f 0 ). For example. For categorical data we will not have a measure of individual units in the sample. all of the data we have been working with so far involve measurement data.

5 12.5 . for this example. and divide by the expected frequency. High Performance Compact Mid Size Full Size 20 14 7 9 Observed Expected 12.81 χ =∑ 2 ( f o − f e )2 Step 3: Run the Statistical Test We have already computed the expected values. Notice that we subtract each expected value from each observed fe value. then we would expect 12. Step 1: Write the Hypotheses for the test. then there must be a difference. If what we observe varies a good bit from the values we would expect if there was not difference.5 12.5 Observed Expected What the Chi-square statistic does is to compare the values we observe to those we would expect if there was no difference. square the difference. so we just need to plug the numbers into the formula. If there really was no difference in the number of insurance claims. We will continue the problem with Alpha set to . For our test degrees of freedom is equal to C – 1.5 claims for each car type.5 Mid Size 7 12.there are 4 different levels of the variable. then we would expect the number of claims to be close to the expected frequencies. We then sum up all of the values we computed. Let’s take a look at the example we have been working on within the context of hypothesis testing. where C is the number of categories Df = 4 – 1 = 3 2 Χ critical = 7. see page A-34.5 12. H1 f o ≠ f e H0: f o = f e Here we are stating that the observed frequencies are the same as the expected for the null. Step 2: Find the Critical Value Again we will use Appendix A to find the critical value. Thus: High Performance 20 12.5 Compact 14 12.5 Full Size 9 12.05.

5) 2 + + + 12.5) χ2 = + + + 12.5 12. . 25 2 .08 Step 4: Make a Decision About the Null Reject the null.5 12.5) 2 (9 − 12. The value we computed in Step 3 is larger than Step 2.χ =∑ 2 ( f o − f e )2 fe (20 − 12.5 12. 25 12 .5) 2 (14 − 12.5 12.5 12.98 = 8.18 + 2.5 56 .42 + 0.5) (1. Step 5: Conclusion Since we rejected the null we say there is a difference in the number of claims made for different car types.5) (−3.5 12.5 χ2 = χ 2 = 4.5) (5.5 2 2 2 2 (7. 25 χ2 = + + + 12.5) 2 (7 − 12.5 + 0.5 12. 25 30 .5 12. so we reject the null.5 12.

So. However. our degrees of freedom will change. In our example: . degrees of freedom for this example are: df = (2 . There are two rows going across and three rows going down. H0: There is no relationship between color preference and personality type (variables are independent). The test of independence starts with frequencies or counts we observe in our sample.1) (C.1) = (1)(2) = 2 We will find the critical value using the same table we used in the goodness-of-fit test. or the observed frequencies.Lesson 20 Chi-Square (Test of Independence) Outline Measuring Independence -observed frequencies -expected frequencies -chi-square Measuring Independence The Chi-square test of independence is similar to the test we just learned in the last lesson. with the null as usual stating that there is no effect or no relationship. Since we have two variables. df = (R . instead of measuring frequencies along only one dimension. Our test is designed to test whether or not these two variables are independent (not related). H1: There is a relationship between color preference and personality type (variables are not independent). For this example. If we reject the null and say the variables are not independent of one another. then we have established that the two variables are related.1) (3. we will measure frequencies for two variables at the same time. a sample of 50 people is used to record personality and color preference is measured: Extroverted Introverted Blue 5 10 Red Yellow 20 5 5 5 __ Observed Frequencies Is there a relationship between personality type and color preference? Our hypotheses will state exactly that.1) where R is the number of rows and C is the number of columns in our table.

Thus. it will be similar to the goodness-of-fit test as well. you must first add up the frequencies for each row and column: fe = Extroverted Introverted Blue 5 10 15 Red 20 5 25 Yellow 5 30 5 __ 20 10 Each value of our six observed values will have separate column frequency ( f c ) and row frequency ( f r ). we will need a formula to compute the expected frequencies instead of just dividing our sample size equally between groups. in our example we must compute six different expected frequencies. For example the observed value in the first row and column (5) has a row frequency of 30 and a column frequency of 15. However. They are the total frequency for the row and column of the individual expected value we are looking for with the computation. So. Extroverted Introverted Blue 5 10 15 Red 20 5 25 Yellow 5 30 5 __ 20 10 So. We will need to compute the expected value separately for each observed value in our sample. the expected frequency is f e = 30 *15 450 = =9 50 50 . In the numerator we have the n frequencies for the “c” column and “r” row. You multiply these values together. fc fr The value n is the total or 50 here.2 Χ critical = 5. So.99 When we start to compute the statistic. Note that our expected frequencies with this test are the values we expect if the null is true. the expected frequencies are the values we expect if there is no relationship or the variables are independent.

Continue finding each expected frequency in this same way for each of the observed values. Whether you compute each individual expected frequency or use the short cut. you will get a complete table of expected frequencies. Once we compute these two expected frequencies all the others are determined (hence the two degrees of freedom). once we have computed these two expected frequencies all of the other values can be found by subtracting out the row or column total. Notice that when we compute a different expected frequency our row and column frequencies will change. must be the number that will make that first column add up to 15. and then divide by the expected frequency for each value in our table. Extroverted Introverted Blue 9 6 15 Red 15 10 25 Yellow 30 6 4 __ 20 10 Expected Frequencies The process for finding the final Chi-square value is the same as before. We will find the difference between our expected and observed frequencies. square the difference. then. All of our expected frequencies must have the same row and column sums as our observed frequencies. There is a shortcut for finding the rest of these values if you feel comfortable with the statistic. for the next value (extroverted and red): Extroverted Introverted Blue 5 10 15 Red 20 5 25 Yellow 5 30 5 __ 20 10 fe = 30 * 25 750 = = 15 50 50 Again. then the remaining number must be 15 – 9 = 6. you continue the process until you have found all the expected frequencies. Since the first expected value is 9. So. The remaining unknown expected value in the first column. So. χ2 = ∑ χ2 = ( f o − f e )2 fe (5 − 9) 2 (20 − 15) 2 (5 − 6) 2 (10 − 6) 2 (5 − 10) 2 (5 − 4) 2 + + + + + 9 15 6 6 10 4 .

Thus.04 Since this value is greater than our critical value.67+2.167+2. we will reject the null here and say there is a relationship. knowing someone’s personality type gives you information about their likely color preference. .5+.χ2 = (−4) 2 (5) 2 (−1) 2 (4) 2 (−5) 2 (1) 2 + + + + + 9 15 6 6 10 4 χ2 = 16 25 1 16 25 1 + + + + + 9 15 6 6 10 4 χ 2 = 1.67+0.78+1.25 χ 2 = 9.

PSY 233 Homework Packet Spring 2010 .

where s p = 1 + s X1 − X 2 n1 + n2 − 2 n1 n 2 1 2 . where sx = n sx single sample t-test formula t = df = n − 1 CI = X ± tcrit ( S X ) independent measures t-test formulas (equal sample sizes only) (X − X 2 ) s2 s2 t= 1 where s X − X = 1 + 2 s X1 − X 2 n1 n 2 1 2 df = n1 + n 2 − 2 -for pooled variances (equal or unequal sample sizes or n’s) s2 s2 (X1 − X 2 ) (n − 1)s12 + (n2 − 1)s22 p p 2 t= where s X − X = . where σ x = n σx X−μ s .2 PSY 233 Formulas sample mean x = ∑X n population mean μ = ∑X N sums of squares sample variance SS = ∑ X 2 (∑ X ) − n 2 s 2 ∑ (X − X ) = n −1 2 population variance 2 σ -OR2 2 ∑ (X − μ ) = N 2 2 s2 = ∑X (∑ X ) − n −1 n σ2 = s= s 2 ∑X (∑ X ) − N N 2 sample standard deviation X−μ σ population standard deviation σ = σ 2 z-score formula z= Z = Xσ + µ z-test formula z= X−μ σ .

... + − 2 k TOT 2 nk N TOT SSWITHIN ⎛ (∑ X 1 )2 2 ⎜ = ∑ X1 − ⎜ n1 ⎝ 2 ⎞ ⎛ ⎟ + ⎜ X 2 − (∑ X 2 ) ⎟ ⎜∑ 2 n2 ⎠ ⎝ 2 ⎛ ⎞ ⎟ + ..3 X1 − X 2 2 ⎡ s12 ( N 1 − 1) + s 2 ( N 2 − 1) ⎤ ⎡ 1 1 ⎤ + ⎢ ⎥⎢ ⎥ N1 + N 2 − 2 ⎣ ⎦ ⎣ N1 N 2 ⎦ t= ANOVA (analysis of variance) formulas (∑ XTOT )2 2 SSTOT = ∑ XTOT − N TOT SS BETWEEN (∑ X ) (∑ X ) = + 2 1 2 2 n1 n2 (∑ X ) (∑ X ) + . + ⎜ X 2 − (∑ X k ) ⎜∑ k ⎟ nk ⎝ ⎠ ⎞ ⎟ ⎟ ⎠ df tot = N − 1 MS Beteween = df Between = K − 1 SS Between SS MSWITHIN = Within df Between df Within df within = N − K F= MS Between MSWithin η2 = SS Between SS total MSWithin n ω2 = SS Between − (k − 1)MSWithin SS total + MSWithin HSD = qα .

4 correlation formulas r= ⎡ ⎢∑ X 2 − ⎢ ⎣ ∑ XY − (∑ X ) n ∑ X ∑Y 2 2 ⎤⎡ ( Y) ⎤ ∑ 2 ⎥ ⎥ ⎢∑ Y − n ⎢ ⎥ ⎥ ⎦ ⎦⎣ df = n − 2 n -ORr= ΣXΣY SPXY where SPxy = ΣXY − n SS X SSY regression formulas ˆ = bX + a Y b= ∑ XY − ∑X 2 ∑ X ∑Y n (∑ X )2 n -ORn −1 n−2 -OR- b= − SPXY SS X a= ∑ Y − b∑ X n a = Y − bX 2 sY −Yˆ = sY (1 − r ) 2 = ∑ (Y − Yˆ ) n−2 goodness of fit chi-square formulas ( f − f e )2 χ2 = ∑ o df = C .1) (C.1 fe Test of independence chi-square formulas ( f o − f e )2 2 fc fr = χ ∑ f fe = df = (R .1) n e .

5 Exam 1 Exam 1 will cover chapters 1-3 in the text. . and Lesson 1-4 online. In Chapter 2 we will not be covering frequency distribution polygons on pages 38-39 of your text.

measurement data and categorical data 2. 4A. Which scale of measurement are the following examples (nominal. senior. independent 2B. freshman. social security number (hint: the number is just a label). faculty member 2C. or ratio)? Select the best answer. graduate. democrat. amount of time it takes a pain reliever to work 2E. interval. a population and a parameter b. descriptive statistics and inferential statistics d. length or width of a room 3. numbers used to identify political affiliation: republican. sophomore. junior.6 Worksheet Chapters 1 and 2 1. What is the dependent variable? . 2A. A recent report concludes that participants on an exercise regimen of running two miles each day had a lower percentage of body fat than participants on no exercise program. 2D. The relation between a sample and a statistic is the same as the relation between a. Are the following examples discrete or continuous variables? Amount of verbal material learned in 30 minutes Number of children in a family 4. a dependent variable and an independent variable c. ordinal. What is the independent variable? 4B.

What is the dependent variable? 6C. or ratio)? 7. ordinal. A study is conducted to determine whether listening to different types of music impairs memory. Two groups of people were selected.7 5. What scale of measurement is the data (nominal. Is the data collected measurement data or categorical data? 6E. ΣXY 7D. ΣX2 7E. Is the independent variable discrete or continuous? 6. Each group is then given a list of 50 words to memorize. interval. 6A. ΣY + 2 7C. A study was conducted to determine whether physically fit persons sleep more hours than those who are not physically fit. (ΣY)2 Y -2 6 7 1 . What is the independent variable? 6B. Use the following data set for 7A through 7G: X 3 4 5 2 7A. They are then given a blank piece of paper and told to write down as many words as they can remember. The other group consisted of people that do not work out at all. One group consisted of people who work out at least 3 times a week. What is the independent variable? 5B. During this 10 minute period. For one week. Participants are given 10 minutes to memorize as many words as they can. What is the dependent variable? 5C. Is the dependent variable discrete or continuous? 6D. subjects slept in a sleep lab and an experimenter recorded the number of hours each person slept. ΣX 7B. one group listens to hard rock. a second group listens to classical music. 5A. and a third group listens to no music at all.

. How many people received less than 9 phone calls? 12. 13. Interval 0-1 2-3 4-5 6-7 8-9 10-11 Real Limits Mid.8 7F. Draw a positively skewed distribution. 9. Twenty FSU students were asked. Σ (X-Y) 8. (ΣX)( ΣY) 7G.Frepoint quency Cumulative Frequency Relative Percentage Cumulative Percentage ______________ ________ ________ ________ ________ _________ __ 10. 10 7 7 6 4 4 6 4 5 5 2 3 3 2 0 2 1 0 11 3 Complete the grouped frequency distribution.5? Interpret. "How many phone calls did you receive last night?" The numbers below are their answers. What percentage of FSU students received between 2 and 3 phone calls? 11. What percentile is associated with a score of 3. What score falls at the 70th percentile? Interpret.

The data are reproduced below: Real Limits f 2.5 5 6.9 Worksheet Chapters 2 and 3 1.5-26.5-22.5-18.5 1 10.5-6.5 5 14.5-10. 9 8 7 Frequency 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Number of Trips 1A.5 0 22.5-38.5 5 30.5 3 26.5-34. 1B.5-30.5 5 Create a histogram for the above data .5 1 34. Compute the mean of the distribution. A retailer created a grouped frequency distribution for the number of weeks individuals spent paying for lay-away items. Compute the median of the distribution 1C.5-14. Compute the mode of the distribution 2. The data is reproduced below. A sample of 44 drivers in South Carolina reported the number of trips they took outside the county of where they lived.5 5 18.

The mode for the above distribution is: a. and 8 4C. all measure of central tendency must be actual values in the distribution . 7 b. 7 5. is normal d. 0 b. the mode d. Median = 20. 1 c. the 50th percentile c. 0 and 7 c. The above distribution: a. 5 d. The only measure of central tendency we are certain to actually observe as a value in our data set is: a. The median is equivalent to: a. 6. is bimodal 4B.10 3. the median c. A distribution of scores has a mean = 30. The distribution: a. is normal d. has a negative skew c. has a positive skew b. 7. is bimodal 4. Which of the following numbers would be considered an outlier in the above distribution? a. 11 and 12 d. and a Mode = 10. Use the following distribution to answer the next three questions Score f 5 1 6 0 7 0 8 0 9 4 10 6 11 7 12 7 4A. has a negative skew c. the 75th percentile d. the 25th percentile b. has a positive skew b. the mean b. none of the above 6.

Which pizza place is most popular among the students surveyed? 5. the mode 4B. What is the best measure of central tendency for this data? a. Pizza Place Frequency Late Night Pizza 5 Papa John’s 6 Pizza Hut 3 Little Caesers 5 4A. Which would be the best measure of central tendency to describe this data set? a. Simpson is innocent or guilty. the median c. not enough information to answer this question .11 Worksheet: Chapters 3 1. the new mean and the old mean will be the same b. how will the mean change? a. The results are as follows. Which is the most commonly used measure of central tendency? a. A survey asked Ohio University students which pizza place they preferred. 4C. What is the mean? 2. the median c. the new mean will be 5 points less than the old mean d. the median c. What is the mode? 1B. "How many phone calls did you receive last night?" The numbers below are their responses. How does it affect the mean when you add a constant to every score? That is. the mode 3. the new mean will be 5 points higher than the old mean c. the mode 4. the mean b. Find the mode of this distribution. What is the median? 1C. the mean b. 0 1 2 0 2 4 2 3 4 5 3 4 5 6 3 10 7 6 7 11 1A. A survey asks whether participants think O. the mean b. if an instructor adds 5 points to everyone's test score. J. A sample of twenty FSU students were randomly selected and asked.

12 6. The lowest possible score on the exam is 0 and the highest possible score is 75. the median will be less than the mode 12 A geography exam was given to samples of high school seniors and college students. the median. (circle one) a.. Their mean income is $200 per week. Based on the two means. What is the median weekly income of the five brothers after Bruce lost his job? Two samples are as follows Sample A: Sample B: 7 13 9 5 10 9 8 1 9 17 12 9 7. Bruce. the median will greater than the mode c. There are five brothers. does it appear that one group is more accurate the other? than . What is the mean for Sample B? 12C. What is the mean for sample B? 9.. the lowest paid. What is the mean for sample A? 10.What is the median for sample B? 11 If a distribution has a positive skew. What is the mode for sample A? 8. gets fired from his $100 a week job and now has an income of $0 per week. What is the mean for Sample A? 12B. The data showing the test scores is below: high school seniors Sample A: 28 college students Sample B: 35 30 33 35 40 40 45 50 50 55 38 40 40 40 40 40 41 42 45 12A. the mean. and their median income is $170 per week. which of the following is true. and the mode will all be the same b.

qualitative d. the number of months survival c. 70% of the sample was specifically taken from urban areas. nominal b. colon cancer 4. The number of months of survival was measured to determine therapy success. The dependent variable was: a. upper real limits d. ratio 2. The independent variable was: a. real limits c.13 Exam 1: Sample Test Multiple Choice (2 points each) A researcher wants to measure the number of pounds of tin the population recycles on average every year. discreet c. gene therapy d. He randomly samples data from 100 recycling plants around the country. What type of scale would be used to measure the tin? a. what should be used to denote the points on the scale of measure? a. the type of therapy b. conventional therapy. Since the researcher knows 70% of the recycling plants are in urban areas. gene therapy d. ordinal c. a random sample of patients is given either the new gene therapy. the number of months survival c. When constructing histograms from a grouped frequency distribution. apparent limits b. The scale used to measure the tin is: a. 3. interval d. mid-point . the type of therapy b. continuous b. or a placebo. parabolic In order to determine whether a new gene therapy will benefit colon cancer patients. colon cancer 5. 1.

5 or less d. such as salaries in the U. In a positively skewed distribution.5. Claire d. 40% of the people got a score of 75. Betty c. ordinal c. Which of the following is not a discrete variable? a. nominal b. The distribution of salaries in the U. nominal b.5 c. If the 40th percentile on an examination is 75. 40% of the people got a score of 75. negatively skewed because rich people represent outliers who earn significantly more than everyone else d. qualitative c. positively skewed because rich people represent outliers who earn significantly more than everyone else 7. less than 40% of the people got a score higher than 75. Measuring the number of times an individual eats during the day is an example of a __________________ variable. Not everything naturally follows a normal distribution.S. amount of time they stayed in a bar d. 60% of the people got a score lower than 75. ratio 11. What scale of measurement is used if you know that one variable is larger than another. a. number of bars a shuffle group visited b. Betty scored the median. Alice scored the mean. number of people who passed out 9. then a. negatively skewed because poor people represent outliers who earn significantly less than everyone else b. discreet 8. and Claire scored the mode. number of tables available c. but you do not know how much larger? a.5 b.14 6. continuous d. Who had the highest score? a. positively skewed because poor people represent outliers who earn significantly less than everyone else c.S. interval d. is: a. They all scored approximately the same 10. Alice b.5 .

number of children in a family c. This is an example of measurement on a scale. nominal b. Which measure(s) of central tendency is/are certain to be changed? a. Compute (∑XY) __________ 18. ordinal c. ratio Use the following data set for the next three problems: (Show your work!) (1 point each) X Y C=3 -3 -2 0 1 -4 1 1 0 17. then a. When a distribution has two separate and distinct medians. Students voted for their preferred professors by ranking them. the mode 13. the median c. the mean and the median d. ratio scaling d. religious affiliation b. The value of one score in a distribution is changed form X = 20 to X = 30. college major 16. a. interval d. inferential statistics 14. nominal scaling c. The concept of generalizing from a few observations to an entire group is central to the area of: a. the mean b. it is bimodal d. a distribution can never have more than one median 15. Compute ∑(X-Y)__________ . it is negatively skewed c. descriptive statistics b. being a registered voter d.15 12. Compute ∑CX __________ 19. it is positively skewed b. An example of a quantitative variable is: a.

complete the grouped frequency distribution. Compute the median 22. Compute the mean 21.16 Use the following population data set for the next few problems 5 10 10 12 15 15 18 18 18 20 20 25 20. Class Intervals Apparent Real Limits Limits 0-4 5-9 10-14 15-19 20-24 25-29 Midpoint Frequency Cum f 23A. The data from these 20 workers are found below: 14 15 16 22 6 12 23 19 27 3 4 2 8 16 15 0 17 14 29 5 (6 points) Relative Percent Cum Relative Percent With the data above. What percentage of workers experienced back pain 14.5 or fewer times? (1 point) . Compute the mode ___________ Show Work! ___________ ___________ (2 points) (1 point) (1 point) 23. A sample of construction workers was asked to report the number of times they experienced back pain on the job in the past month. Twenty workers reported their incidents of back pain every day for a month.

Create a frequency histogram of the above data (from the grouped frequency distribution). What percentile is associated with 14.17 23B. How many workers experienced back pain between 5 and 9 times? 23D. How many times did the 80th percentile experience back pain? 23C. (2 points) .5 incidents of back pain? (1 point) (1 point) (1 point) 23E.

18 Exam 2 Exam 2 will cover Chapters 4-6 in the text. . Also. In chapter 4 we will compute the interquartile range differently than your text pp 79-80. and Lesson 5-8. the online lessons and homework packet contain information about conditional probabilities not covered by your text.

Two samples are as follows Sample A: Sample B: 7 13 9 5 10 9 8 1 9 17 12 9 3A. 0 1 2 0 2 4 2 3 4 5 3 4 5 6 3 10 7 6 7 11 1A. the new standard deviation and the old standard deviation will be the same b. the new standard deviation will half the size (twice as small) as the old standard deviation d. if an instructor divides everyone's score by two. What is the variance for sample B? 4. not enough information to answer this question . The lowest possible score on the exam is 0 and the highest possible score is 75. the new standard deviation will be twice as large as the old standard deviation c. What is the variance? 1B. The data showing the test scores is below: high school seniors Sample A: 28 college students Sample B: 35 30 33 35 40 40 45 50 50 55 38 40 40 40 40 40 41 42 45 . What is the standard deviation? 1C. How does it affect the standard deviation when you divide a constant into every score? That is.19 Worksheet Chapter 4 A sample of twenty FSU students were randomly selected and asked. What is the standard deviation for sample A? 3C. "How many phone calls did you receive last night?" The numbers below are their responses. which sample has more variability? 3B. What is the interquartile range? _______ 2. Just by looking at these data. how will the standard deviation change? a. A geography exam was given to samples of high school seniors and college students.

It should be obvious that the instructor made a mistake in his calculations. does it appear that one group is more accurate than the other? 4B.5.20 4A. An instructor gives his class a 10-point quiz..e. The next day he tells his students that the average score on the quiz was X = 7. Based on the two means. What is the standard deviation for Sample A? 4C. Explain why. What is the standard deviation for Sample B? 4D. has less variability)? 5. Which group is more consistent (i.5 with a standard deviation of s = 13. .

What percentage of men live between 55 and 60 years?__________ .36? 3.43 and a z-score of 1. What percentage of men live 75 years or longer? __________ 6. what is the percentile rank is associated with a score of 65? 4.5. Use this information to answer problems 5-8.25 and a z-score of . What percentage of men live 65 years or less? _________ 8. 5. In a normal distribution of test scores with a mean equal to 57 and a standard deviation equal to 6.3.21 Worksheet Chapter 5 and 6 1.33? 2. What is the percentage area between a z-score of -1. What is the percentage area between a z-score of . The scores on a personality test are normally distributed with μ = 250 and σ = 30. What percentage of people taking the test can be expected to score between 229 and 325? The average man in an industrialized country lives μ = 70 and σ = 6. What percentage of men live between 65 and 75 years? __________ 7.

e.7 is . Men in third-world countries have a life expectancy of μ = 60 and σ = 4. 13. who lived longer relative to their age distribution? In a distribution with a mean of 50 and a standard deviation of 5: 15.22 9. Men in industrialized countries have a life expectancy of μ = 70 and σ = 6. The raw score corresponding to a z-score of -1. 14.3 11. In a distribution of scores with a mean of 1500 and a standard deviation of 250. The z-score corresponding to a raw score of 68.00 is . What raw scores mark the middle 60% of this distribution? . What raw score cuts off the top 10% of this distribution? 18.51 is . 95% of the men will live between the ages of ______ and _________ years find the raw values that mark the middle 95% of the distribution of ages) (i. what raw score corresponds with the 67th percentile? Questions 11 . 12.3. What raw score corresponds with the 14th percentile? 16. The raw score corresponding to a z-score of 0. 10.3. If a man in a thirdworld country lives to be 65 and a man in an industrialized country lives 72.13 refer to a distribution with μ = 60 and σ = 4. What z-score cuts off the top 10% of this (or any) distribution? 17.

Billy wants any beer. He knows they all have very different tastes. heads and tails are mutually exclusive because . What is the probability that the first beverage Jake randomly grabs is a beer? 2B. a. and 12 bottles of Coke. 24 bottles of Molson beer. When flipping a coin. Jake has 12 bottles of Coors beer. Jake is having a party for all of his friends in his apartment complex. it has no influence on whether the coin comes up heads or tails on the next toss. 24 bottles of Heinekin beer. Allison wants a Coke. What is the probability of drawing a red ace out of a standard deck of 52 cards? 6. What is the probability of drawing three cards out of a standard deck of 52 cards. b. sampling is without replacement 2. what is the probability that the second beverage Jake randomly grabs is a Coke? 3. if the coin comes up heads on one toss. What is the probability of drawing a red card out of a standard deck of 52 cards? 5. What is the probability of drawing an ace out of a standard deck of 52 cards? 4. if the coin comes up heads. it cannot also come up tails. 2A. Given that the first bottle grabbed was a Coors. without replacement. 8 bottles of wine coolers. sampling is with replacement d. so he stocks his refrigerator with a large selection. c.23 Worksheet: Chapter 6 1. and have all 3 cards turn up red? .

all are equally likely 9. HHHHH b. 9F. P(F): The probability someone is available full-time 9C..24 7. A letter of the English alphabet is chosen at random. 7A. P(E): The probability someone has coffee shop experience 9B. If I flip a coin 5 times which set of heads (H) and tails (T) outcomes is more likely: a. and some can only work part-time. Coffee Shop No Coffee Shop Experience (E) Experience (not E) Available Full-Time (F) 20 12 Available Part-Time (not F) 42 31 Find each of the following probabilities. Some of the applicants can work full-time. 9E. HTHTH d.. . There are 105 applicants for a job with a new coffee shop. TTTTT c. Some of the applicants have worked at coffee shops before and some have not served coffee before. The exact breakdown of applicants is as follows. P(not E): The probability someone has no coffee shop experience 9D. 9A.. is a vowel (consider y a consonant) 7B. P(E & F): The probability someone has coffee shop experience and is available full-time. P(F | E): The probability someone is available full-time given that they have coffee shop experience. Find the probability that the letter selected. is any letter which follows p in the alphabet 8.. P (not F | not E): The probability someone is available part-time given they have no coffee shop experience.

most scores were below the mean c. all of the raw scores are the same d. it cannot also come up tails.95 d. not enough information to answer this question 5. the new standard deviation and the old standard deviation will be the same b. When flipping a coin. The interquartile range is not the best measure of dispersion because it eliminates 50% of the distribution. none of these because the standard deviation can never be negative 3. if the coin comes up heads on one toss. the probability of getting heart disease . a. . then a. The 50% of the distribution that is eliminated is: a. When the variance is equal to zero.. the probability of it raining c. a. . it has no influence on whether the coin comes up heads or tails on the next toss. the lower 50% d. if everyone's score is divided by 2. heads and tails are independent because . the variance can never be equal to zero 4. the distribution is badly skewed d.80 c. How is the standard deviation affected when you divided a constant into every score? That is. the probability of having being struck by lightning if it is raining d. how will the standard deviation change? a.25 Exam 2: Sample Test Multiple Choice (2 points each) 1. the standard deviation is equal to 1 b.05 2. the probability of being struck by lightning b. Which of the following is a conditional probability a. . the raw scores are negative c. 6. the new standard deviation will be half the size (twice as small) as the old standard deviation d.. most scores were above the mean b. the middle 50% b. . its probability value is a. if the coin comes up heads. the new standard deviation will be twice as large as the old standard deviation c. When the standard deviation has a negative value. If an event can occur once out of 20 times. the lower 25% and the upper 25% 7. b. the upper 50% c.20 b.

What proportion can be expected to score between 19 and 21? Show your work! (3 points) 13. are first multiplied together. To calculate the probability of the joint occurrence of two independent events. are multiplied together d.80 c. are subtracted from each other 9. and then subtracted from 1. The mean of the Stanford Binet IQ is 100 with a standard deviation of 16. the probabilities for the separate events occurring a. 11. . 5 green.35 b. the probability that it will rain and the wind will blow tomorrow 10. what is the probability of drawing a red M & M (which you eat) and then another red one? a.15 The average score on a test of hand steadiness is 20 (μ = 20). . .16 d.0 c. the probability that the wind will blow tomorrow b.26 8. the probability that it will rain or the wind will blow tomorrow d. and 10 yellow M & Ms left in the package. If there are only 10 red. What proportion of individuals can be expected to score higher than 28? Show your work! (2 points) 12. the probability that the wind will blow tomorrow given that it rains c. Which of the following is a conditional probability? a. Use the following population data set to answer the next problem: (Show Work!) 54 29 35 10 28 36 32 45 48 60 Compute the interquartile range _________ (2 points) 14. The standard deviation is 5 (σ=5). are added together b. . .

(3 points) B. Compute the variance ___________ Show Work! 20 20 25 (3 points) 16. What score falls at the 80th percentile. What is the lowest Stanford-Binet IQ you could have and still be eligible to join Mensa? Show your work. Show your work (2 points) D. Mensa is an organization that only allows people to join if their IQs are in the top 2% of the population. What is the probability of obtaining an IQ score lower than 80? ____________ (2 points) Use the following population data set for the next few problems 5 10 10 12 15 15 18 18 18 15. What percentage of the population has a Stanford-Binet IQ score between 84 and 95? Show your work. Compute standard deviation ____________ (1 point) .27 A. (3 points) C.

That is. (3 points each) Problem Frequency Drug 10 Family crisis 20 Other 20 A. what is the probability that they would involve a drug case and a family crisis case. If two of the files were randomly selected one at a time. (Sampling is one at a time with replacement. And the remaining twenty people needed help for miscellaneous reasons. . The numbers are summarized below. she had 50 files. Ten people sought out help for drug related problems. If one of the files were selected at random. B. The psychologist kept 1 file for each person she helped. If one of the files were selected at random. what is the probability that they would both involve drug cases.) Leave your answer in decimal form. If two of the files were randomly selected one at a time.) Leave your answer in decimal form. D. Twenty people needed help for family crisis problems. (Sampling is one at a time with replacement. what is the probability that it would involves a drug case or a family crisis case? Leave your answer in decimal form. C. what is the probability that it would involve a drug case? Leave your answer in decimal form. The psychologist met with 50 employees. A company hired a psychologist to assist their employees in their personal problems.28 17.

29 Exam 3 Exam three will cover Chapter 7 and 8 in the text. . and Lesson 9-12 online.

The researcher finds that a person who exercises regularly has a body fat percentage of 13%. What is the null hypothesis in words and symbols? 2C. Was this a one-tailed or a two-tailed test? 1B. Is this a one-tailed or a two-tailed test? 2B. What was the alternative hypothesis in words and symbols? 2. A study is conducted to determine whether a new drug will improve memory. What was the null hypothesis in words and symbols? 1B. A person taking the new drug is able to recall 35 words from a list of 50 after studying the list for 10 minutes. A researcher predicts that someone who exercises regularly should have a different percentage of body fat than people who do not exercise at all. Does this percentage differ significantly from the general population of people who do not exercise and have a body fat percentage of 20%? 1A. Do they recall more words than the general population that can recall only 25 words? 2A. What is the alternative hypothesis in words and symbols? .30 Worksheet: Chapter 7 and 8 1.

05. a sampling distribution b. the more variability there is in the set of values b. none of the above . 1 . and are less likely to result in rejection of Ho 6. one-tailed test 5. 1 . If we repeatedly sample from a population and form a distribution of sample means it is: a.β c. β b.α 8. α=. and are less likely to result in rejection of Ho d.01. alpha c. power d. the less variability there is in the set of values c. One tail-tests: a. What is the critical value for each of the following? 4A. α =. Someone claims to be a member of the team who is 74 inches tall. none of the above 9. The probability of a Type II error is: a. two-tailed test 4C.76 inches. α d. The mean height of the basketball team is 79 inches high with a standard deviation of 1. What is the probability that someone 74 inches or shorter really is on that basketball team? (Hint: this is mostly a z-score probability problem like we did on Exam 2) 4. one-tailed test 4B. the probability of a Type II error b. The height of the students are normally distributed. the standard error d. the standard deviation 7. predict the direction of the effect. The larger the standard deviation: a. a sampling distribution of the mean c. do not predict the direction of the effect and are more likely to result in rejection of Ho c.31 3. do not predict the direction of the effect. The probability of correctly rejecting the null is: a.01. α =. standard deviation does not indicate variability d. predict the direction of the effect and are more likely to result in rejection of Ho b. The basketball coach likes to recruit tall students.

Also include the probability of each cell. Type I error b. Fill in the blanks with correct decision.. Telling someone to go home and take an aspirin when in fact he needs immediate treatment is an example of .32 10. What is the probability of committing a Type I error given that the null hypothesis is actually true? 12. correct decision 18. Type II error c. Type III error d... correct decision 17. Type II error c. Type I error. Type I error b. Letting a guilty woman go free is an example of. a.. Type II error c. Which cell is power? Decision Reject null Fail to reject null True state of the world Null is true Null is false 15. a. Convicting an innocent woman of a crime is an example of . What is the probability of committing a Type I error given that the null hypothesis is actually false? 11. a.. Type IV error 16. correct decision .. a. and Type II error. Type I error b.. Type II error c.. Telling someone that he has a disease when he does not is an example of . Type I error b.

its mean C. STEP 3: Summarize the data into the appropriate test-statistic.3 days in the hospital.33 Worksheet: Chapter 7 and 8 (Part 2) 1. From the central limit theorem. we know which of the following characteristics of the sampling distribution. all of the above . Test at the . That is. STEP 2: Set up the criteria for making a decision.05 level of significance with a one-tailed test. On the basis of these data. its standard deviation D. can the hospital conclude that the new program has a significant reduction of recovery time. The hospital is trying a new recovery program that is designed to lessen the time patients spend in the hospital.. find the critical value.. The distribution of recovery times is normal with a σ = 1. STEP 1: State your hypotheses (include both H0 and H1). STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? 2. its shape B. The first 10 appendix patients in this new program were released from the hospital in an average of 5. Patients recovering from an appendix operation normally spend an average of 6.5 days. What is the Central Limit Theorem? Why is it so important? 3. A.2 days.

7. In earlier chapters z = X -μ X. Which of these factors does the experimenter have control over before he/she collects data? 6. and subsequently your decision about the null.4 hours per week on the internet. In this chapter the z formula used is z = σ .34 4. From the text.5 hours. The researcher took a random sample of 30 Time Warner cable customers and found that they spend a mean amount of 23. A diligent researchers found that the typical person spends a mean amount of μ = 25 hours per week using the internet. Do Time Warner cable customers spend less time on the internet than others? Set α = .01. σ n What are the differences between the two formulas? Why are the formulas not the same? 5A. with a standard deviation of 2.μ . STEP 1: State the null and alternative hypotheses in words. Label H1 and H0 State the null and alternative hypotheses in symbols Label H1 and H0 STEP 2: Set up the criteria for making a decision (find the critical value). what are some of the factors that affect the likelihood of rejecting H0? 5B. (1 point) . Name the factors that affect the z-score.

what is your conclusion? 7B. Based on your answer above. STEP 4: Evaluate the null hypothesis. Reject or Fail to Reject STEP 5:(conclusion) Based on your evaluation of the null hypothesis. what type of error might you have made in your decision in Step 4? .35 STEP 3: Compute the appropriate test statistic. Show your work.

will approximate the normal curve d. . It is known that there is some risk of sampling error. Professional athletes are now commonly tested for steroid use following competition. the population means are assumed to be equal 6. None of the above is correct . She asked her advisor what alpha level she should choose to minimize the risk of Type I error. Alpha does not have a direct effect on Type I errors. larger.36 Exam 3: Sample Test 1. α does not have a direct effect on Type I errors 4. the standard deviation of the sampling distribution of the sample means b. Type II error d. 3. A psychology student was getting ready to propose her thesis. What would constitute a Type II error on the part of the testing agency. an athlete who is not using steroids tests negative (drug-free) d.01 b. will approximate the normal curve b. less. . Type I error c. fewer. 5. an athlete who is using steroids tests positive (not drug-free) c.05 c. less. According to the Central Limit Theorem.025 c. the __________ likely the sampling distribution of means ___________. will approximate the standard deviation c. When the null hypothesis is rejected. .05 d. if their null hypothesis is that the athlete is drug-free? a. What is the alpha level she should choose to minimize the risk of a Type I error? a. fewer. both b and c 2. but she was very worried about making a Type I error. an athlete who is using steroids tests negative (drug-free) b.01 b. What is the standard error? a. α = . the sample means are assumed to be equal d. a. the _________ the size of the samples selected from the population. but this risk is believed to be minimal. an athlete who is not using steroids tests positive (not drug-free).025 d. a significant difference has been established c. α = . more. then a. Type II error is committed b. α = . A researcher is very worried about making a Type I error. Which of the following gives the least chance of making a Type I error? a.

2. have no effect on d. a biased sample d. the Central Limit Theorem b. Z. you test negative for a disease but you really do have it 8. can’t answer: Not enough information 12. you test negative for a disease and you really do not have it c. What is the critical value for a one tailed-test. the hypothesis that states ‘the error variability is expected to be less than 1’ c.33 d. a. decrease c. If all other factors are held constant. a test of alpha b. decreases in the sample variance will ________ the value of the t-statistic.05? a. the hypothesis that states the number of subjects to be used in the experiment 11.05. 1. you test positive for a disease but you really do not have it b. A directional test means the same as: a. According to your text. variability due to chance 10. a two-tailed test c. a one-tailed test 9. you test positive for a disease and you really do have it d. the hypothesis that states what the experiment was designed to investigate d. or no relationship is expected’ b. the failure to accept the research hypothesis c. 1.64 c. Which of the following would constitute a Type II error? a.58 . increase b. The research (alternative) hypothesis is: a. sampling error means the same as: a. a test of power d. alpha = .37 7. the hypothesis that states ‘no difference.96 b. 2.

(1 point) REJECT or FAIL TO REJECT (circle one) What is the best conclusion. In words: In symbols: STEP 2: Find the critical value.000. according to your decision in STEP 4? (1 point) . (1 point) (4 points) STEP 4: Evaluate the null hypothesis (based on your answers to the above steps). Be sure to clearly label your null and alternative hypotheses.000 with a standard deviation of 6. Do people with only high school degrees earn less than the rest of the company? 13A. Conduct a one-tailed hypothesis test with α = . In a large corporation the mean entry level salary is $27. The entry level salaries for a random sample of 15 employees with only high school degrees is $24.100.38 13.05. (4 points). STEP 3: Compute the appropriate test-statistic. STEP 1 State your hypotheses in both words and symbols.

Be sure to clearly label your null and alternative hypotheses. You want to know. Conduct a TWO-TAILED test STEP 1 State your hypotheses in both words and symbols. Years of population counts have shown African leopards have an average number of spots equal to µ = 25 with a standard deviation of 7 spots. whether Snow leopards have a different number of spots compared to those from Africa. (1 point) Compute the appropriate test-statistic. A biologist claims that Snow leopards have a different number of spots than African leopards. He gets a representative sample of 15 Snow leopards. with a 95% level of certainty. Show your work (4 points) STEP 4 Evaluate the null hypothesis (based on your answers to this point) REJECT or FAIL TO REJECT (circle one) (1 point) . You notice that these leopards have an average of 30 spots. (4 points).39 14. In words: In symbols: STEP 2 STEP 3 Find the critical value.

and 13 in the text. One-tail tests will not be included in the problems here or on the exams. and Lesson 13-16 online. 10. . The formulas for ANOVA (Chapter 13) will vary slightly from the text.40 Exam 4 Exam 4 will cover Chapters 9. Also note that all of our test will now be done as two-tail tests. We will not do the Scheffe test on pp 330-332.

05. Using α = . conduct a t-test. 2-tail. two-tailed. A psychobiologist hypothesizes that the diastolic blood pressure of Type A persons differs from the average person.05. find the critical value.41 Worksheet: Chapter 9 1.Set α = .05. STEP 3: Summarize the data into the appropriate test-statistic. That is. Conduct a t-test to see if a sample of 65 participants with a mean of 83 and a standard deviation of 5. Interpret. the mean diastolic blood pressure is μ = 80.4 is significantly different than a population mean of 80. STEP 1: State your hypotheses (include both H0 and H1). with the standard deviation of S = 18. The sample mean diastolic pressure is X = 93. STEP 2: Set up the criteria for making a decision. . compute the t statistic. 2. STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? Compute 95% confidence limits on μ.76. Set α= . That is. The psychobiologist takes the blood pressure of 22 Type A men whose ages range from 21 and 29. two-tailed. In the population.

42 3. A population has μ = 100 and σ = 50. Find the t-score for each of the following sample means: a. a sample of n = 25 with X = 220, s = 50 b. a sample of n = 4 with X = 230, s = 50 c. a sample of n = 100 with X = 190, s = 50 4. A particular state knows that its officers can run a mile in μ = 7 minutes, and they want to improve this overall running performance of the force. You are the chief statistician for the state-attorney’s general office, and you have been asked to check to see if new recruits hired under a new standard can run faster than the uniformed officers. You plan to compare the mean-mile run time of ten recruits to the average of 7 minutes to determine if it takes them a different amount of time to run a mile. The run times (in minutes) are: 5.2 5.0 6.8 9.3 11.1 7.0 8.4 8.0 9.9 8.4

(hint: you must compute the mean and standard deviation from the sample) 4A. Should you do a one-tailed or a two-tailed hypothesis test? 4B. Conduct the appropriate hypothesis test. STEP 1: State your hypotheses (include both H0 and H1). Set α = .05.

STEP 2: Set up the criteria for making a decision. That is, find the critical value. STEP 3: Summarize the data into the appropriate test-statistic.

STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? 11C. Compute 95% confidence limits. Interpret.

43 5. A manufacturer of flashlight batteries claims that its batteries will last an average of μ = 34 hours of continuous use. After receiving several complaints about the batteries, a consumer protection group predicts that the batteries run in a different amount of time than 34 hours. During consumer testing, a sample of n=30 batteries lasted an average of only X = 32.5 hours with a standard deviation of 3. Conduct a two-tailed hypothesis test with α = .05. STEP 1: State your hypotheses (include both H0 and H1).

STEP 2: Set up the criteria for making a decision. That is, find the critical value.

STEP 3: Summarize the data into the appropriate test-statistic.

STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? 6. In a single-sample t-test, what are the respective critical values for: A. α =.05, n=10, two-tailed test B. α =.01, n=31, two-tailed test C. α =.05, n=40, two-tailed test D. α =.01, n=107, two-tailed test

44

Worksheet: Chapter 10
1. The standard error of the difference (for the independent measures t-test) is an estimate of a. centrality b. normality c. variability d. none of the above 2. If other factors are held constant, increasing the level of confidence from 95% to 99% will cause the width of the confidence interval to: a. increase b. decrease c. not change d. there is no consistent relation between interval width and level of confidence 3. In an experiment, the experimental group has 13 participants with s2 = 3.24 and the second group has 15 participants with s2 = 2.56. Compute the pooled variance

4. Suppose a teaching methods study was designed to test a hypothesis of equal means on the final examination scores for an experimental teaching method and the traditional lecture method. Subjects were randomly assigned to one of the two methods, classes were taught, and final examination scores were recorded. A summary of the data is as follows Experimental: n = 16 X = 87.5 s2 = 38.13 Traditional: n = 16 X = 82.0 s2 = 42.53 Which type of hypothesis testing should be conducted in order to assess whether there is a difference in the final exam scores of the two teaching techniques? a. single sample t-test b. dependent samples t-test c. independent samples t-test Conduct the appropriate hypothesis test. STEP 1: State your hypotheses (include both H0 and H1). Set α = .05, two-tailed.

45 STEP 2: Set up the criteria for making a decision. STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? . find the critical value. STEP 3: Summarize the data into the appropriate test-statistic. That is.

That is.61.46 5. STEP 3: Summarize the data into the appropriate test-statistic.4 with a variance of 10. The sample of phobic patients gave themselves a mean rating of 12.01. STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? . two-tailed. The participants rated themselves on a 1 to 15 scale with higher numbers indicating worse performance. find the critical value. compute the t statistic. Which type of hypothesis test should be conducted in order to assess whether there is a difference in the self report ratings of the two groups? a.5 with a variance of 9. STEP 2: Set up the criteria for making a decision. whereas the nonclinical sample had a mean self-rating of 9. Set α = . independent samples t-test Conduct the appropriate hypothesis test.24. Rapee and Lim (1992) asked 28 persons with social phobias and 33 nonclinical subjects to rate themselves on a public speaking performance that they gave. single sample t-test b. STEP 1: State your hypotheses (include both H0 and H1). That is. dependent samples t-test c.

04). STEP 1: State your hypotheses (include both H0 and H1). lost a mean of 3. Which type of hypothesis test should be conducted in order to assess whether people using the diet pills lost more weight? a. The members of the placebo group.05. STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? . single sample t-test b. The diet pill group lost a mean of 4. The researcher gets two groups of people. That is. Both groups are then instructed to try to lose weight. on the other hand.47 6. That is. STEP 3: Summarize the data into the appropriate test-statistic. A researcher is studying whether diet pills really work. independent samples t-test Conduct the appropriate hypothesis test. The second group of 15 people is given a placebo. The first group of 20 people is given the diet pill to help suppress their appetite.63) during the one month experiment. STEP 2: Set up the criteria for making a decision. find the critical value. The researcher hypothesizes that the people who were given the diet pill will lose more weight. compute the t statistic. dependent samples t-test c. two-tailed.61 pounds (with a variance of 12.78 pounds (with a variance of 10. Set α = .

at least one mean is statistically different from one other mean b. all the means are different from each other c. all the means come from the same population d. When does one conduct an ANOVA? 3. In an ANOVA summary table. greater than 1 d. What is the abbreviation for analysis of variance? 2. we calculate two estimates of the population variance. not enough information given 5.48 Worksheet: Chapters 13 1. When the null hypothesis is true. Between variability can also be thought of as A) between groups variability B) within groups variability C) total variability D) both A and B 8. Which estimate of the population variance is independent of the truth or falsity of H0? 6. then F = MSbetween / MSwithin will be equal to: a. 1 c. If you obtain a significant F statistic you know that: a. the null hypothesis is probably correct 4. what are the sources of variability? 7. Within variability can also be thought of as A) between groups variability B) within groups variability C) total variability D) both A and B . 0 b. To test the truth or falsity of H0.

what are some of the reasons that the groups in an experiment may have different values? (In other words. what are some of the reasons for between groups variability. In this experiment. Suppose I was conducting a study to see which network can make people laugh more on Thursday nights. All participants watch television from 8:00 to 10:00 with a tape recorder. how come everyone in the NBC group does not have the same laughter score?) 13. 10. In this experiment. That is. I have three groups: One group watches NBC. what are some of the reasons for within group variability. The total variability can also be thought of as A) between variability + within variability B) error variability C) within variability D) between variability Use the following example for questions 10 . what are some of the reasons that people in the NBC group have higher laughter scores than people in the CBS group?) 12. What is a multiple comparison procedure (post-test) and why does one need to conduct one when conducting ANOVA? 14. and the third group watches CBS. the second group watches ABC. That is.49 9. what are some of the reasons that the subjects within each group may have different scores? (In other words.12. The experimenter listens to the tape to record laughter. What is the appropriate statistical test? A) Pearson's r B) single sample t-test C) ANOVA D) related measures t-test 11. What is Tukey’s HSD? When does one compute Tukey’s HSD? What does HSD stand for? .

A pool of subjects was randomly divided into five treatment groups. What do eta-squared and omega-square measure? Which one is more accurate? 16. What are two measures of magnitude of effect? Which measure is less biased? 17. Using the . carry out a complete ANOVA on these data. .05 level of significance. STEP 2: Set up the criteria for making a decision STEP 3: Summarize the data into the appropriate test-statistic. 0mg 6 5 3 2 250mg 3 4 5 4 500mg 3 3 4 2 1000mg 4 1 0 3 2000mg 1 0 2 1 STEP 1: State your hypotheses.50 15. The data in the table represent the number of cold and flu viruses reported by the participants as a function of their vitamin C dosage. The groups were administered daily doses of vitamin C over a 12-month period.

51 STEP 4: Evaluate H0. Use Tables D.01. If appropriate.05. α = . dfgroup = 3.4 to determine the critical value for F (Fcrit) for each of the following situations: 20A. Calculate and interpret η2 (eta squared) on the data in question 17. use Tukey's HSD test to perform pairwise comparisons on the means of the data in the above question.3 and D. α = . 20. dferror = 60 20B. dferror = 30 20C. α = . dfgroup = 5. dfgroup = 7. Complete the ANOVA summary table. dferror = 120 20D. Source SS df MS 40 F Group (Between)80 Error (Within) Total 100 14 . (Reject or Fail to reject) Conclusion: 18. dferror = 24 21.01. α = . 19.05. You do not need the raw data to complete this table. dfgroup = 4.

If other factors are held constant. the samples were drawn from populations that were actually dependent rather than independent. d. what term is used to signify (or is equivalent to) variance? a. sum of squares c. difference scores are found for each subject b. the variance of the distribution of differences between means is computed 6. proportion of total variance due to between group differences d. mean square d. there is no consistent relation between interval width and level of confidence 2. the mean of one sample is statistically the same as the mean of the other sample so the decision is that they come from populations that have the same mean value. proportion of total variance due to individual differences 4. increase b. MS group is best described as the a. When conducting an independent measures t-test . the mean of one sample is so far from the mean of the other sample that the decision is that the samples come from populations that have different mean values. the comparison is made against a t-distribution d. if the null hypothesis is rejected: a. b. In an Analysis of Variance test (ANOVA).01 significance level should be used to increase power d. all of the above . only the . decrease c. variability due to individual differences c. In ANOVA. F-ratio b. variance due to between group differences b. increasing the level of confidence from 95% to 99% will cause the width of the confidence interval to: a. When conducting an independent measures t-test: a. c.52 Exam 4: Sample Test 1. not change d. EXCEPT a. the medians of the two populations are assumed to be equal b. Each of the following is part of conducting a independent measures t-test. the null hypothesis is rejected if the calculated t-statistic you compute is more extreme than the critical-t c. both a and c 5. the population variances are estimated c. degrees of freedom 3.

all of the means are significantly different from one another c. when there are more than two means to compare d. the normality assumption b. the repeated measures assumption 10. increase 12. a one-tailed test c. when the population variances are unknown c. if we changed from a one-tailed to a two-tailed test. ten c. between variability < within variability d. infinity d. between variability = within variability c. When do you normally use analysis of variance rather than the independent measures t-test? a. 14 c. one 11. If you obtain a significant F-statistic then you know that: a. when the population means are unknown b. between variability > total variability 8. we would expect power to a. decrease c. 16 . between variability > within variability b. the F ratio is near a. when the data is badly skewed 9. all of the means belong to the same population d. If there is no treatment effect. An independent measures experiment uses two samples with n = 8 in each group to compare two experimental treatments. 15 d.53 7. you decide to reject the null hypothesis. zero b. Keeping everything else constant. 7 b. The assumption that the population variances are the same is called a. homogeneity of variance d. When conducting an ANOVA. then null hypothesis is probably correct 13. Which of the following must be true? a. at least two means are significantly different from one another b. The t-statistic from this experiment will have degrees of freedom equal to a. remain unchanged b.

omega-squared c. when the sample size is less than 30 b. When doing an independent samples t-test. One group of participants (n = 10) watched the movie.86 X = 11.05.6 Conduct a two-tailed test with α = . (1 point) (3 points) Step 4: Based on your answers above. when MUST you pool the variance? a. Which of the following is the least biased measure of magnitude of effect? a.05 15. Both groups were then given a racial attitude test. when you are using an alpha level less than .94 X = 9. A pool of subjects was randomly divided into 4 treatment groups. eta-squared b. state your decision about the null (1 point) REJECT FAIL TO REJECT (circle one) What does your decision lead you to conclude about the research question? In other words. beta d. (1 point) 16. wherein high scores represented a higher level of prejudice. and a control group (n= 10) spent the hour playing cards. delta 17. The groups were administered daily doses of Vitamin C over a 12-month period. A researcher is interested in whether a certain hour-long film that portrays the insidious effects of racial prejudice will affect attitudes toward a minority group. The data in the table . Step 3: Conduct the appropriate statistical test. Step 1: State the null and research hypotheses in symbols: Control 2 s2 = 9. when the samples are of unequal sizes c. state the results of the experiment.54 14. when you are performing a one-tailed test d. Summary data were as follows: Movie s12 = 8.75 (2 points) Step 2: Set up the criteria for making a decision.

Using the . (8 points) F Source Group Error Total SS df MS 2 44 Step 4: Based on your answers above. (1 point) Step 2: Set up the criteria for making a decision (1 point) Step 3: Conduct the appropriate statistical test. 0mg Σx1 = 16 Σx12 = 74 n1 = 4 500mg Σx2 = 12 2 =38 Σx2 n2 = 4 1000mg Σx3 = 8 2 Σx3 = 26 n3 = 4 2000mg Σx4 = 4 2 =6 Σx4 n4 = 4 Step 1: State the null hypotheses in words or symbols. state your decision about the null (1 point) REJECT FAIL TO REJECT (circle one) Based on your decision about the null.55 represents the number of cold and flu viruses reported by the participants as a function of their vitamin C dosage. which group reported the least number of colds and viruses? (1 point) Conduct a test of Magnitude of Effect using the least biased estimator (2 points) . is it appropriate to conduct a post-hoc test? (1 point) YES NO (circle one) Just by looking at the data you used to conduct the test. complete the ANOVA.05 level of significance.

26 X 2 = 18. The omnibus ANOVA was significant with a MSwithin = 36.89. Conduct a Tukey's post-hoc test to determine which of the groups differed from one another. Interpret the effect size you computed above. A researcher conducts an ANOVA test to determine which of 3 treatments (using 33 total subjects) will extend terminal cancer patients lives the longest. (2 points) 19. The mean number of months patients survived for each of the groups is printed below.56 18B. (5 points) X 1 = 28. Set α = .39 X 3 = 17.15 .05.

There will also be a comprehensive section on the final exam worth 15 points. Also. the formulas the homework packet and online lecture notes contain for this chapter differ from those the text uses.57 Final Exam The non-comprehensive part final exam will be worth 50 points (the same as the other exams) and will cover Chapters 15 and 16 in the text. In chapter 15 we will not cover the Spearman correlation pp 404-409. These points will be taken from material on the previous three exams. . and Lesson 17-20 online.

compute the correlation.. That is.. That is. STEP 2: Set up the criteria for making a decision. Compute the correlation coefficient and conduct a hypothesis test using the following steps. Number of hours slept before exam X 10 12 3 8 5 Number of correct answers on exam Y 5 11 0 13 9 1A. .01. find the critical value. how many correct answers should they get if they sleep 9 hours (Hint: Compute the regression equation). STEP 3: Summarize the data into the appropriate test-statistic. STEP 4: Evaluate the Null Hypothesis (Reject or Fail to reject?) What is your conclusion? 1B. Set α = .58 Worksheet: Chapter 15 1. According to the data. She asked a sample of 5 people from her residence hall the number of hours they slept before and the number of correct answers they got on their first exam. The data are as follows. STEP 1: State your hypotheses (include both H0 and H1). two-tailed. A previous student of this class was curious about the relationship between number of hours a person slept before an exam and the number of correct answers on the exam.

64% e. the population correlation is not zero d. not enough information to answer this question 8. What proportion (or percent) of the variability is predicted by the relationship with weather? a.05 +0. the sample correlation is not zero 7.80. What is the difference between the predictor variable and the criterion variable? 5.59 2. 80% b.84 6. Estimate her annual salary.98 Y 2a) Samantha has had 0 years of higher education. the population correlation is zero c. 2b) Tabatha has had 11 years of higher education. ˆ = 2X + 12. Use the regression equation below to predict the yearly salary (in thousands) from the number of years of higher education. the null hypothesis states a.97 +0. Which type of correlation coefficient should be computed when both the X variable and the Y variable are dichotomous? a. 20% d.26 -0. Pearson c. 2c) What is the slope of this regression equation? 2d) What is the intercept of this regression equation? 2e) What is the regression coefficient and y-intercept of this regression equation? 3. For the test for significance of a correlation. the sample correlation is zero b. Put them in order showing the highest to lowest degree of relationship: -0. What is predicted (or predictable) variability (r )? 2 . Suppose the correlation between hot chocolate sales and weather temperature is -0. 40% c. Point biserial d. Spearman 4. What is the "best" fitting line? 9. Listed are 4 correlations. Estimate her annual salary. Phi b.

Use the following data for the next 2 problems. A sample of n = 27 pairs of scores (X and Y values) produces a correlation of r = +0.05 level of significance. twotailed. What is the standard error of estimate for the above data. 11.50. 10B. Are these sample data sufficient to conclude that there is a non zero correlation between X and Y in the population? Test at the . . X 0 1 2 Y 9 7 11 10A. Interpret. Find the regression equation for predicting Y from X from the above data.60 10.

no expected frequencies should be less than 5 . What is the goodness-of-fit test? 3. What type of data does one need to have in order to conduct a chi-square test? 2. Nonparametric tests are referred to as ____ free tests. test of independence 8. distribution c. What is k? 5. What is R? What is C? 7. Degrees of freedom for the test of independence is defined as df = (R .1. contingency test b. goodness-of-fit test d.1). What are observed frequencies? What are expected frequencies? 4. parametric test c. definition b. Which one of the following statements about chi-square is not true? a.1) (C .61 Worksheet: Chapter 16 1. a. measurement d. the observations must be dependent c. parameter 6. chi-square is used primarily with nominal data b. Degrees of freedom for the goodness-of-fit test are defined as df = k . A chi-square test on two categorical variables is called a a.

A potential sponsor would like to know whether local viewers prefer some evening news programs over others.05. STEP 2: Set up the criteria for making a decision STEP 3: Summarize the data into the appropriate test-statistic. (Reject or Fail to reject) Conclusion: . Also state a conclusion. Test the hypothesis that the incidence of depression. Be sure to state your hypotheses.01. as measured in this way. The sponsor conducts a viewer preference survey based on a simple random sample of 1000 households. find your critical value. The table below shows the frequencies of new admissions to a metropolitan psychiatric clinic as a function of season. calculate your test-statistic. Depression Other diagnosis Spring 20 15 Summer 10 15 Fall 12 25 Winter 25 20 10. and evaluate the null hypothesis. Use α = .62 9. KTVO KMDT KLPF KZTV 220 200 300 280 STEP 1: State your hypotheses. STEP 4: Evaluate H0. The results are given in the table. is independent of season. using α = . Perform a goodness-of-fit test on these data.

Also state a conclusion. Using α = . test this hypothesis. find your critical value.63 11. The data in the table were gathered in an investigation of possible gender differences in book-carrying behavior among college students. Be sure to state your hypotheses. The researcher wants to know if men. calculate your test-statistic.05. How does the Chi-square test of independence differ from the chi-square goodness of fit test? . Book-Carrying Styles Down at the Side In Front Other 24 70 6 100 46 4 Women Men 12. compared with women. tend to carry books down at their side rather than in front of them. and evaluate the null hypothesis.

then a.59 d. The direction of a linear relationship between two variables is given by a..00 b.00 6. X.00 or -1. the numerical value of the denominator 5. negative c.00 . 4. correlation between X and Y d. one of the survey questions provided 7 response alternatives. X. slope of a straight line b. when Y increases by two units for each equal single-unit increase in X.. a large value but not +1. the intercept equals +0. the intercept equals +2.63 c. The amount of change in a Y variable that accompanies a given amount of change in X is: a. 30.50 d. the slope equals +2. 7.11 of r. X d. length of the prediction line 3. a. In a survey of 20 individuals. 1. the plus or minus sign c. X b. zero b.11 b. the critical value for the test-statistic would be a. the numerical value b.64 Final Exam: Sample Test 1. non-existent d. either +1. Y c. In regression analysis.. Y. Y-intercept of a straight line c. then the relationship is said to be. The Y-intercept is the value of a. positive b.50 c. 10. a small value but not zero d. neutral 2. a.00 c. 12. Y. If two variables are related so that as values of one variable increase the values of the other also increase.. If the responses were evaluated using a χ2 test for goodness of fit at the α = . Y when the value of is equal to zero. the slope equals +0.00 or -1. A perfect linear relationship of variables X and Y would result in a value of r equal to.05 level of significance. both the sign and the numerical value d.

00 b. if you knew a student's need for achievement score. ρ 12. -1. the line has a slope of +5. α b. μ d.54 between measures of need for achievement and college grade point average. blood pressure tends to increase also d. residual ...00 9. a. there is no relationship between need for achievement and college grade point average ˆ = -1. standard error of estimate c. r = -0. the line has a negative slope and intersects the X axis at +5.. there is a tendency for college grade point to decrease c. The population correlation coefficient is represented by. From this equation we know that 13. you would expect that . β c.. In linear regression the difference between a value of Y and Y a.50 d. higher income is associated with higher blood pressure c. Which correlation coefficient represents the weakest association between the X and Y variables? a.65 8.0 d. standard deviation d. blood pressure tends to decrease 11.0 b.4 c. zero d.60 c. as income increases. as need for achievement scores increase. a. +2. as need for achievement scores decrease. there is a tendency for college grade point to decrease d. -.4 and intersects the Y axis at +5. This study indicates that . you could predict the student's grade point average perfectly b. r = +0. Which of the following values of r allows perfect prediction of the Y score from knowledge of the X score? a. A study has found a negative correlation between a person's income and his or her blood pressure. income and blood pressure are not related b. error of measurement b. The equation of a regression line is Y a. a. X and Y are not linearly related ˆ is known as the .90 10. Given this knowledge. r = +0. as income increases.0. 14. the line has a slope of -1.0 and intersects the Y axis at -1. A psychologist has found a correlation of +0.50 c.4X + 5. r = -0.20 b.

00 c. sample statistics.00 or +1. sample means. When computing a chi-square test of independence one compares a. sample variances. positive slope? 20. If you fail to reject the null hypothesis in a chi-square test for goodness of fit.00 d. If you were to compute a correlation between the X and Y variables for each of the three sets of data. population variances c. variances should be about equal b.00 16. for which set of data would the regression equation have the most negative slope? _____ . observed frequencies. r = -1. 17. for which set of data would the regression equation have the largest. population parameters to . population means b. The standard error of estimate in linear regression will be zero when a. frequencies for the cells should be equal (1 points each) Below are three scattergrams.) A B C _____ _____ 18. which set of data would yield a correlation closest to zero? 19. the slope of the regression line is 10. then the expected and observed a.66 15. expected frequencies d. If you were to construct a regression equation using the X variable to predict the Y variable for each of the three sets of data. the slope of the regression line is 0. (Note: A scattergram may be the correct answer to more than one question. variances should be unequal c. frequencies for the cells should be unequal d. r = zero b. If you were to construct a regression equation using the X variable to predict the Y variable for each of the three sets of data.

(1 point) Based on your results. (Runyon & Haber. and then they indicated their perceived marital happiness using an 11-point scale ranging from very unhappy (1) to perfectly happy (11). Set α = . twotailed. 1991) In a recent study. Twenty-eight married couples volunteered to monitor their daily frequency of sexual intercourse and arguments for 35 consecutive days. Thornton (1977) explored the relationship of marital happiness to the frequency of sexual intercourse and to the frequency of arguments.67 21.74 STEP 4: Evaluate the null hypothesis.05. I have already done this for you: r = -0. Thornton (1977) reported that the Pearson correlation between ratings of marital happiness and number of arguments was -0. That is. STEP 1: State your hypotheses in either words or symbols (2 points) STEP 2: Set the criteria for making a decision. find the critical value points) (2 STEP 3: Summarize the data into the appropriate test-statistic. Do the appropriate statistical test to determine whether there is a significant linear relationship between happiness and arguments.74. is there a relationship in the population between happiness and arguments? YES NO (circle one) (1 point) What proportion of the variability in happiness can be explained by the number of arguments? (1 point) .

Person Weight (X) Y2 XY Time (Y) X2 1 89 11.5 6.5 miles (in minutes) for a sample of 5 individuals. Set α = .68 22. (3 points) Compute the correlation between weight and running time. (Birkes & Dodge.4 33.1 5. Barnett. Istvan.507. alcohol consumption.6 4 92 12.014. and coffee consumption (Zvela. Georgia and Fort Campbell.01 757.921 129.6 5 83 10.464 151.625 102. Soldiers at Fort Gordon. That is.21 732. which included items about cigarette use.29 1. Kentucky completed a questionnaire.5 405 55. One of the questions the researchers wanted to answer was the following: Is there a relationship between smoking and gender in the military? The data are below.72 4.1 4. Gender Smoking Status Current smokers Ex-smokers Nonsmokers Total Male 252 62 170 484 Female 46 29 51 126 Total 298 91 221 610 Perform a chi-square test of independence on these data. & Matarazzo.4 7.6 2 75 10.3 8.356 123. (Set up the appropriate formula to receive credit for your answer.131. (I have already done this for you).05 STEP 1: State your hypotheses. Smedi.5 3 66 11. 1990). H0: Gender and smoking status are independent H1: Gender and smoking status are not independent STEP 2: Set up the criteria for making a decision. find the critical value.96 1.8 23a. (2 points) STEP 3: Summarize the data into the appropriate test-statistic (3 points) STEP 4: Evaluate the null hypothesis (1 point) What is your conclusion? (1 point) 23.255 616.25 871.) . 1993) Below is the weight (in kilograms) and the time to run 1.889 110.

weight is the criterion variable in . weight. 23d. (1 point) is the predictor variable and the regression equation. (5 points) Write the regression equation for predicting running time from weight. (Set up the appropriate formulas to receive credit for your answer.) 23c. time b.69 23b. (1 point) Predict the running time for a child who weighs 77 kilograms. 23e. (1 point) What is the value for the slope of the regression line in 27b. (circle one) a. time.

70 24. A discount store has prepared a customer survey to determine which factors influence people to shop in the store. A sample of 90 people is obtained and each person is asked to identify from a list of alternatives the most important factor influencing their choice to shop in the store. The data are as follows: Convenient Location 30 Low Prices 40 Good Selection 20

On the basis of these data can you conclude that there is any specific factor (or factors) that is most often cited as being important? Test at the .05 level of significance with the goodness of fit chi-square test. Determine the critical region (1 point)

Summarize the data into the appropriate test-statistic

(3 points)

Evaluate H0. (Reject or Retain)

(1 point)

Part 2 25. Compute the median and the mode of the following data set. 9 median mode 7 4 5 7 2

(1 point) (1 point)

71

26. A national test has a mean of 192 and a standard deviation of 10. The author of the exam wants the test to have a mean of 500. What specifically does the author have to do so that her test has a mean of 500 (and the standard deviation remains 10)? (1 point) 27. Which measure of central tendency is used with nominal data? (circle one)(1 point) a. mean b. median c. mode 28. In October of 1981 the mean and the standard deviation on the Graduate Record Exam (GRE) for all people taking the exam were 489 and 126, respectively. Scores on the GRE are normally distributed. 28a. What percentage of students would you expect to have a score between 400 and 500? (1.5 points)

28b. What is the median of this distribution? (1.5 points)

29. A psychologist would like to know how much difference there is between the problem-solving ability of 8-year-old children versus 10-year-old children. A random sample of 10 children is selected from each age group. The children are given a problemsolving test, and the results are summarized as follows: 8-year-olds n = 10 x = 36 s = 3.50 10-year-olds n = 10 x = 39 s = 5.27

30. Perform the appropriate analysis on this data. Set α = .05, two-tailed. STEP 1: State your hypotheses in symbols (1 point)

72 STEP 2: Set the criteria for making a decision. That is, what is your critical value? (1 point) STEP 3: Summarize the data into the appropriate test-statistic. (2 points)

STEP 4: Evaluate the null hypothesis

(1 point)

In a controlled study, more than 70 Dartmouth College students were instructed to use orange-flavored lozenges at the first sign of an incipient cold, sucking on one as often as every two hours. Half the students got zinc lozenges; half the students were given candies that looked and tasted the same, so that none knew who was really taking the zinc. The participants who were given the zinc had a cold for 4.3 days, as against 9.2 days for those who got the look-alike candies 31A. What was the dependent variable in this study? (circle one) (1 point) a. type of cold treatment b. 70 students c. number of days cold continues d. Dartmouth College 31B. What was the independent variable in this study? (circle one)(1 point) a. type of cold treatment b. 70 students c. number of days cold continues d. Dartmouth College 31C. What is the correct analysis for this experiment? (circle one) (1 point) a. independent measures t-test c. chi-square test of goodness of fit b. related measures t-test d. single sample t-test

5 . 12.5-11. music 5B. 11. 168 9.5 10. 2 8. 144 7F.5.5-9.5 5 14 25 6-7 5.5.5-5. continuous 6D. ratio 7A. ratio discrete body fat words recalled 5C. Real Mid.5 8. nominal 2B.5. nominal 2D. 7G.FreCumulative Relative Interval Limits point quency Frequency Percentage 0-1 -0. 6E.5 6. 30% 18 5. ratio 2E. 4A.5 6 9 30 4-5 3.5 4. 14 7C.73 Answers: Chapters 1 and 2 1. 45% of the scores fall at or below 3. 6A. Cumulative Percentage 15 45 70 90 90 100 .5-3. measurement 547E. 13. physical fitness 6B. 70% of the scores fall at or below 5.5 0 18 00 10-11 9.5-7.5 2. discrete amount of sleep 6C.5 4 18 20 8-9 7.5 3 3 15 2-3 1. 45th percentile. A 2A. 5A.5 2 20 10 10. ordinal 2C. 3A.5-1. 55 7D. continuous 3B. exercise regimen 4B. 147B.

5 12.5 1B.5 16.00 8.5 . 8.5 No.5 20.5 32.5 24.74 Answers: Chapters 2 and 3 1A.00 2.00 1C.5 28. 8. of Weeks 3) A 4A) B 4B) C 4C) C 5) B 6) C 36. 6 5 Frequency 4 3 2 1 0 4. 8.

40. 4.1 12.25 2 c 3 a 4A. 9 9. 9. c 4B.6 12. on average both groups are fairly accurate . b 6 170 7 9 8. 9 11 b 12. Papa John's 4C.4 1B.3. 40. Papa John's 5.75 Answers: Chapters 3 1A. No. 2.17 10. 4 1C.

5-14.5-29.5-19.5 15-19 19.5 10-14 14.5 5-9 9.76 Exam 1: Sample Test Answers 1) D 2) A 3) A 4) B 5) D 6) D 7) D 8) C 9) A 12) A 13) D 14) D 15) B 16) B 17) 2 18) -18 19) -6 20) 15.5 22) 18 23A) 50% 23B) 19.5 23C) 3 23D) 50th Class Intervals Apparent Real Midpoint Frequency Cum f Limits Limits -.5-24.5 21) 16.5 25-29 23E) 7 6 Frequency 5 4 3 2 1 0 # of Back Pains 10) B 11) C Relative Percent 20 15 15 30 10 10 2 7 12 17 22 27 4 3 3 6 2 2 4 7 10 16 18 20 Cum Relative Percent 20 35 50 80 90 100 .5 20-24 24.5 0-4 4.5-9.5-4.

2. 4 2. 8. C 3A. on average both groups are fairly accurate 4B. 9.5 points away from the mean on a 10-point scale . No. cannot be 13.83 1B.72 3C. 1. Sample B 3B. 32 4A. 2.77 Answers: Chapter 4 1A.56 4D. Sample B 5. on the average.22 4C.97 1C. Scores.

57. 21.78 Answers: Chapters 5 and 6 1.2 .04% 7. 57. 60 12. 2.4 18. 21. 89th 4.6 16. 1610 11.02 14. 1. 53.31) 15.18% 2.35 10.50% 3. 53.28 17.84% 9.65 to 82. 24. 54. 44. 45. 75.8.18 5. Third-world man lived longer for his distribution (z=1. 4.16) than the other man (z=. 56.48% 8.51 13.48% 6.

9D. 4.5 .3226 . 5. 6.7209 .0385 .0769 .4095 . 7B.38 D .5905 .3048 . 3.79 Answers: Chapter 6 1.15 .1176 .1905 . 9A. 2A. 9F. 7A.19 . 9C. 9E. 9B. 8.75 . 2B. a .

25.8 14B) 21.44 14D) Z= -1. prob = .2 b) 0.75 16) 5.80 Exam 2: Sample Test Answers 1) D 2) D 3) C 4) C 5) B 6) D 7) C 8) C 9) B 10) D 11) .08 d) 0.96% 14C) 113.6 c) .26 17a) 0.1586 13) 19 14A) 132.1056 15) 27.04 .054 12) .

1B. one-tail Participants taking the new drug will recall less than or the same amount of words as the untreated population. 10. 6. 5. .64.33 A B A A C p(Type II error | null is false) = 0 p(Type II error | null is true) = α see text a b a b 4. two-tail The percentage of body fat for exercisers will not differ from those who do no exercise μexercise = 20 The percentage of body fat for exercisers will be different from those who do no exercise μexercise ≠ 20 2A. μnew drug > 25 3.58. 16. Participants taking the new drug will recall more words than the untreated population. 11. 12. . 18.0023 1. 2. 9. 7. 8.81 Answers: Chapters 7 and 8 1A. 1C. μnew drug ≤ 25 2C. 2B. 15. 17. + 2.

H1: Time Warner cable customers spend less time on the internet than the general population H1 µtw < 25 H0: Time Warner cable customers spend the same or more time on the internet than the general population H0: µtw > 25 Critical value = -2.3 critical z = z.11 Reject H0 Patients in the new program are released from the hospital in less time.or two-tailed test 6. the second for finding the probability of a sample of values. 7B. alpha.64 zobtained = -2. sigma 7. distance between means. sigma. two-tail test.51 Reject H0 Time Warner cable customers spend less time on the internet than the general population. 5B. In the same way σ estimates the average difference between μ and X. one-tail vs. D 4.33 Z= -3.3 H1: μ < 6.82 Answers: Chapter 7 & 8 (Part 2) 1. alpha level. σ / n estimates the average difference btw/ µ and X 5A. one. distance between means. Sample size. The first is used for finding the probability of an individual value. H0: μ> 6. see sampling distributions on web page 3. Sample size. 2. N.05 = -1. Type I (because you rejected the null it is the only type of error possible) .

83 Exam 3: Sample Test Answers 1) A 2) A 12) B 3) A 4) A 5) B 6) D 7) D 8) D 9) D 10) C 11) A 13) Step 1: H1: People with a HS degree earn less than other company employees μ HS < 27. 000 / 15 = −2900 −2900 = = − 1.76 7 / 3.64 Step 3: t = 24.81 Step 4: Reject the null conclusion: there is a different number of spots for Snow leopards .873 1.000 6.873 1549.87 6000 / 3.1000 − 27.000 H0: People with a HS degree earn the same or more than other company employees μ HS > 27000 Step 2: -1.96 Step 3: z = 30 − 25 7 / 15 = 5 5 = = 2.19 Step 4: Reject People with a HS degree earn less than other company employees 14) Step 1: H1: Snow leopards have a different number of spots than African leopards H0: Snow leopards have the same number of spots as African leopards H1: μ Snow ≠ 25 H0: μ Snow = 25 Step 2: +1.

31.750 6D. 12 b.84 Answers: Chapter 9 1. a. CI. +2. 4B.045 tobtained = -2. 6C.25 Reject H0 Type A persons have significantly higher blood pressure than the average person. 5. H0: μ = 34 H1: μ ≠ 34 critical t = t.00 tobtained = 4. H0: μ = 80 H1: μ ≠ 80 critical t = +2.021 6B.95 = 6.262 tobtained = 1.262 +2.95 = 84.62 .05 = +2.05 = +2. critical t = t.08 tobtained = 3.73 Reject H0 The batteries last significantly less time than claimed by the manufacturer.32).48 Reject H0 H0: μ = 80 H1: μ ≠ 80 critical t = t.51 ≤ μ ≤ 9. 95% sure that the population of Type A men have a mean blood pressure in this range. ±2. (CI. 4C.2 c. 18 two-tailed H0: μ = 7 minutes H1: μ ≠ 7 minutes 2.47 Retain H0 The sample does not run the mile in less time than the pop. 95% sure that the population the sample of troopers comes from has a mean running time in this range. 4A. 6A.68 ≤ μ ≤ 101.05 = ±2. 3. ±2. 5.

05 = +2. 2. 2. .678 tobt = 3.87 c H0: μem = μtm H1: μem ≠ μtm critical t = t.85 Answers: Chapter 10 1.46 Reject H0 The students in the experimental teaching class performed significantly better on the final exam than students in the traditional class. a 3.042 tobt = 1. c H0: μdiet = μplacebo H1: μdiet ≠ μplacebo critical t = t.042 tobt = 2.83 Reject H0 Social phobic patients rated themselves significantly worse on public speaking performance than did nonclinicals. c H0: µsp = µnc H1: µsp ≠ µnc critical t = t. 6. c 4.05 = ±2. 5. Diet pills are not significantly more effective than placebos in losing weight.05 =±2.02 Fail to reject H0 Diet pills do not work.

η2=.02 20C.g. 3. 8. Subjects taking no Vitamin C and subjects taking 250 mg.01 21. of Vitamin C. 15.) 12. 18. HSD stands for honestly significant difference. Tukey’s HSD is a post-test (multiple comparison procedure).95 20B.06 Fobt = 3..g. Omega-square is less biased. Source SS df MS F Group 80 2 40 23. Some people laugh more than others. Some networks are funnier than others.88 19. Individual differences Error 13. 16. The tape recorder picked up other noise which made it difficult to hear the laughter. of Vitamin C reported significantly more cold and flu viruses than persons taking 2000 mg. 6..g. 7. H0: μ1=μ2=μ3=μ4=μ5 H1: At least one mean is different from the others F.95 Error 20 12 1. 14.15) = 3. 2. ANOVA only tells us that at least 2 means differ. . 17. 2..86 Answers: Chapters 13 ANOVA When you wish to compare more than two sample means.) Error (e. Omega-square is more accurate.29 20D. 10. A B within group variability (variance) Between/ Within/Total . both measure magnitude of the effect. A B A C Individual differences (e. eta-square and omega-square. but not which ones…must do Tukey’s post-hoc test to compare multiple groups and determine which means differ. we can conclude.05(4. After conducting the Tukey HSD. One computes a Tukey’s HSD when the null hypothesis has been rejected to determine which of the groups are significantly different from each other.) Treatment (e. 2.93 Reject H0 At least one group reported more cold and flue viruses than at least one other group.67 Total 100 14 1. 4. 4. 3. 9. 5. 11.51 20A. Tukey’s HSD = 2.

no differences in attitudes between the film and no film group 16) B 17) Step 1: µ1=µ2=µ3=µ4 Source Group Error Total SS 20 24 44 Step 2: 3.1009 = − 2......6 − 11.30 18B) 30% of the variability in number of cold virus reported is due to amount of vitamin C consumed.37 Step 4: Fail to Reject.so. 19..39 Groups 1 and 2 differ (difference=9.11)..46 + 88..24).87). HSD = 6.15 80...56 1.200mg group 18A) ω 2 = .736 [.no. Groups 1 and 3 significantly differ (difference = 11.49 df 3 12 15 Step 3: MS 6.67 2 F 3.99) 2 + 9(3.15 = −1. ..75 9(2.87 Exam 4: Sample Test Answers 1)A 2) C 3) A 4) B 5) A 6) B 7)A 8) C 9) C 10) D 11) B 12) A 13) B 14) B 15A) Step1: H0: µfilm=µnofilm Step 3: t = 9.2] 18 = − 2.33 Step 4: Fail to reject. Groups 2 and 3 do not differ (difference = 1.....14) 2 ⎡ 1 1⎤ + ⎥ ⎢ 10 + 10 − 2 ⎣10 10 ⎦ H1: µfilm ≠ µnofilm Step2: + 2.

10B. 2B.98 c See page 414 of your text -0. -0.56 Fail to reject H0 There is insufficient evidence to conclude that there is a significant linear relationship. 9. 3. 5.45 The standard deviation of points about the regression line (standard error) is 2. 2C.68 or 8 answers.79 X + 1. 2D. 8.26. Yes 1B.05 a d See page 414 of your text variability in Y that is explained by differences in X ˆ = X +8 Y 2.88 Answers: Chapter 15 1A. 2E.98 2 and 12. +0. 10A.45. 2A. The regression equation is Y The answer is 8. 12.959 r = . 11.97. 7. . ˆ =.980 34.980 2 12. 6.84. +0.57. 4. H0: ρ = 0 H1: ρ ≠ 0 rcrit = ± .

H0: observed frequencies are equal to the expected frequencies H1: observed frequencies are not equal to the expected frequencies χ2crit = 7.35 χ2 = 7. 10. Chi-square test of independence: consider 2 variables at once to determine if they are independent (related). Chi-square goodness of fit test: consider 1 variable at a time. 428 of your text See pp. 5.89 Answers: Chapter 16 1.2 Reject H0 Local viewers prefer some evening news programs over others. 442 of your text D B H0: The incidence of depression is independent of season. 8. 4. 2. 7. . 9.99 χ2 = 43. Compares actual data to what we expect by chance. H1: The incidence of depression is not independent of season. 3.69 Reject H0 Men compared with women tend to carry books down at their side rather than in front of them. 6.82 χ2 = 27. 230-431 of your text k stands for k-k-categories (number of groups) A See p.22 Retain H0 The incidence of depression is not independent of season. 12. 11. Categorical or frequency data See p. χ2crit = 11. H0: Book-carrying styles are independent of gender H1: Book-carrying styles are not independent of gender χ2crit = 5.

53 + ...898 rounds to 10.67..2611 .2 = 26 rcrit = ±0.0453 23d) 10.16 + . yes there is a relationship What proportion of the variability. Add 308 to each score. weight.2970 28b).41 23c) . There is a relationship between gender and smoking status 23a) r = +0.90 Final Exam: Sample Test Answers 1) A 2) A 3) C 4) B 5) A 6) C 7) D 8) D 9) A 13) C 14) D 15) B 16) C 17) D 18) B 19) B 20) C 10) D 11) D 12) B 21) STEP 1: H0: ρ = 0 H1: ρ ≠ 0 Null: The correlation does not exist in the population Alternative: The correlation does exist in the population STEP 2: df = n ..7% z = +0.5 STEP 4: Fail to reject null 30a) C 30b) A 30c) A .93 + 1.71 area = .2 = 28 . 489 29)STEP 1: H0: μ1 = μ2 H1: μ1 ≠ μ2 STEP 2: df = 18 tcrit = ±2.63 = 12...so.5476 22) STEP 2: df = (3-1)(2-1) = 2 χ2crit = 5.71 STEP 4: Reject null conclusion? Gender and smoking status are not independent. 27) c. χ 2obtained = 6.101 STEP 3: tobtain = -1.90 23e) a.0359 area = .09 z = -0.2611 = ..02 + 3. Reject 25) median 6 mode 7 26) Adding a constant to each score will change the mean without having an effect on the standard deviation.99 STEP 3: χ2obtained = 1.99..0359 + . mode 28a) 29.5658 = +.374 STEP 4: Reject the null.57 23b) Y = .. time 24) critical = 5.44 + 5.0453X + 7.? .