INFERENTIAL STATISTICS

Notes: Hypothesis Testing 1

Notes: Parametric and Non-Parametric Tests 3

Hypothesis Testing 4

Pearson’s Product Moment Correlation Coefficient 9

Spearman Rho 13

T-test 15

Z-test 20

Chi-Square Test 21

F –test 24

One-Way ANOVA 25

F-test Table 27

GDD, Panabo National High School i

 INFERENTIAL STATISTICS
HYPOTHESIS TESTING

Some Notes on Hypothesis Testing

What is a Hypothesis?
A hypothesis is a tentative explanation for certain events, phenomena or behaviors. It is an assertion
subject to verification or proof. In statistical language, a hypothesis is a statement of prediction of the
relationship between or among variables. In statistical decision making, it is more precise to use the term
statistical hypothesis because research is concerned with assertions that can be tested with statistical
evidence, that is, with information based on samples. Plainly stated, a hypothesis is the most specific statement
of a problem. It is a requirement that these variables are measurable and that the statement specifies how
these variables are related. Furthermore, the hypothesis is testable which means that the relationship between
the variables can be put into test by means of the application of an appropriate statistical test on the data
gathered about the variables.
The definition of “hypothesis” can be narrowed even further, in that, social research is usually based on
samples and sample findings are generalized to populations. Thus a more appropriate definition is:

Hypothesis is a statement about a population parameter that can be tested with a sample
statistic. It is a statement about the possible outcome of the study.
Hypothesis predicts whether or not a possible difference or relationship between variables under
study exists.

Hypothesis and Assumption Distinguished

Hypothesis is a statement of a possibility, subject to verification while an assumption is a statement of

fact which is presumed to be true and does not need to be tested or verified.

Steps in Hypothesis Testing

1. State the Null and Alternative Hypotheses. The first step in traditional hypothesis testing is to
specifically note the assumed value of the parameter before sampling. This assumption to be tested
is known as the null hypothesis, and the symbol for null hypothesis is Ho. The null hypothesis is
usually a statement of no relationship or difference between variables under study.

It is important to note that the null hypothesis is the one that contains the equality relation – that
is, the Ho states that some parameter (mean, percentage, variance, etc.) is equal to specified value.
A test is often carried out for the purpose of trying to show that the Ho is not true. The researchers
hope in this test is that the sample results will show a higher parameter value and thus would not
support the Ho.

If the sample results do not support the null hypothesis, the researcher must conclude something
else. The conclusion that is accepted contingent on the rejection of the null hypothesis is known as
the alternative hypothesis. The alternative hypothesis is a statement of relationship or difference and
the symbol for the alternative hypothesis is H1.

The alternative hypothesis may be either non-directional or directional. When H 1 asserts only
that the population parameter is different from the one hypothesized, it is referred to as a non-
directional or two-tailed hypothesis. Occasionally, H1 is directional or one-tailed. In this instance, in
addition to asserting that the population parameter is different from the one hypothesized, we assert
the direction of the difference.

The selection of an alternative hypothesis depends on the nature of the problem. As with the null
hypothesis, the alternative should be stated prior to actual sampling.
 GDD, Panabo National High School 1
2. State the Level of Significance. The level of significance is the point at which the sample statistic
differs too much from the population expectation to be due to random variation. Establishing a level of
significance is a matter of setting the point at which the sample differs too much from what is
expected given sampling variation. It consists of deciding what the demarcation is between a small
and a large difference. Small differences are those with a high probability of occurrence, whereas
large differences have a low probability of occurrence. The decision as to whether a particular sample
outcome is significant then it is based on its probability of occurrence.

In the social sciences, the common significance levels or cut-off points as the basis for inferring
the operation of non-chance factors are .05 (5%) and .01 (1%). When the event would occur 5% of
the time, the result is probably due to non-chance factors. This cut off point is known as .05 or 5%
significance level. When the event would occur 1% of the time, the result is probably due to non-
chance factors. This cut off point is known as .01 or 1% significance level.

The level of significance set by the experimenter for inferring the operation of non-chance factors
is known as the alpha (a) level. Thus, when employing the .05 level of significance, a = .05; when
employing the .01 level of significance, a = .01.

3. Determine the Statistical Test to be used. Once the level of significance is chosen, it is then
necessary to select the correct statistical test to be used, such as Pearson r, t-test, chi-square,
ANOVA.

4. Define the Rejection or Critical Region. After the level of significance is stated and the proper statistical
test distribution is selected, the next step is to find the boundaries for the critical region of the
sampling distribution. The critical region is also referred to as the region of rejection, because
outcomes of this region call for the rejection of Ho. If the sample observation falls in this area, it is
seen as significantly different from the population expectation.

In a non-directional or two-tailed test, the critical region falls on both ends of the sampling
distribution. In a directional or one-tailed test, the critical region falls on one end of the sampling
distribution, either at the upper or lower end.

Rejecting or accepting the null hypothesis carries the risk of some error. The error involved is of
two types. When we reject Ho, we run the risk of type I error, and when we accept Ho, we run the risk
of type II. Type I error is that of rejecting Ho when it should be accepted, while type II error is that of
accepting Ho when it should be rejected. The difficulty is that there is no way of knowing whether one
has made the correct decision.

5. State the Decision Rule. After we have stated the hypothesis, selected the level of significance,
determined the test distribution to use, and defined the rejection region(s), the next step is to prepare
a decision rule. A decision rule is a formal statement of the appropriate conclusion to be reached
about the null hypothesis based on sample results.

6. Make the Necessary Computation. After all the ground rules have been laid out for the test, the next
step is the actual data analysis. A sample of items is collected and the sample statistic(s) is (are)
computed.

7. Make a Statistical Decision. If the computed value of the sample statistic falls in the critical region or
region of rejection, the null hypothesis is rejected and the alternative hypothesis is accepted.
Otherwise, the null hypothesis is accepted and the alternative hypothesis is rejected.

Tables of significance for rejecting or accepting the null hypothesis are usually provided in
statistics books. If the computed value is equal or greater than the table value, the null hypothesis is
rejected and the alternative hypothesis is accepted. If the computed value is less than the table value,
the null hypothesis is accepted.
 GDD, Panabo National High School 2
Some Notes on

PARAMETRIC AND NON-PARAMETRIC TESTS

I. PARAMETRIC TEST – is a test whose model specifies certain condition about the parameters of the
population from which the research sample was drawn. Since those conditions are not ordinarily tested,
they are assumed to hold. The meaningfulness of the results of a parametric test depends on the validity of
these assumptions.

For example, a technique of inference may be based on the assumption that the scores were drawn
from a normally distributed population. Such technique produces a conclusion which contains qualifiers, e.g.
“If the assumptions regarding the shape of the population(s) are valued, then we may conclude.

Parametric tests also require that the scores under analysis result from measurement in the strength of
at least an interval scale.

Parametric tests are more powerful than non-parametric tests.

EXAMPLES OF PARAMETRIC TESTS:

a. Pearson, r
b. T-test
c. Analysis of Variance (ANOVA)

CONDITIONS TO BE SATISFIED FOR PEARSON r

a. Linear regression of scores in scatter plot.

b. Homoscedasticity – tendency of SD’s or variances of the arrays to be equal.

CONDITIONS TO BE SATISFIED FOR T-TEST:

a. The observations must be independent.

b. The observations must be drawn from normally distributed population.
c. These populations must have the same variances.
d. The variable involved must have been measured in at least an interval scale, so that it is possible to
use the operations of arithmetic (adding, subtracting, multiplying, dividing, finding means, etc.)

CONDITIONS TO BE SATISFIED FOR ANALYSIS OF VARIANCE OR ANOVA (F-TEST)

a. Same conditions for t-test.

b. In addition, the means of the population must be linear combinations of effects due to columns and/or
rows. That is, the effects must be additive.

It must be reiterated that the parametric test are more powerful only when the assumptions underlying
their use are valid. When the assumptions are not met, a non-parametric treatment may be as powerful as the
parametric treatment.

II. NON-PARAMETRIC TEST – is a test whose model does not specify conditions about the parameter of the
population from which the sample was drawn. They are also known as “distribution free” techniques. Having
used one of them, we may say that “Regardless of the shape of the population, we may conclude that . . . .”

Most non-parametric tests apply to data in an ordinal scale, and some, apply also to data in a nominal scale.
Examples of such data are ranks and frequencies.
 GDD, Panabo National High School 3
EXAMPLES OF NON-PARAMETRIC TESTS:

a. Chi-Square
b. Spearman rho
c. Kendall’s test
d. Mann-Whitney U-test
e. The Sign Test
f. Wilcoxon Matched Pairs Signed-Rank Test

ADVANTAGES OF NON-PARAMETRIC TESTS:

1. Probability statements obtained from most non-parametric tests are exact probabilities, regardless of the
shape of the population distribution from which the random sample was drawn. In other words, the accuracy
of the probability statement does not depend on the shape of the population.

2. If sample sizes are very small, there is no alternative in using a non-parametric test unless the nature of the
population distribution is exactly known. However, the power of any non-parametric test may be increased by
simply increasing the size of N.

3. Non-parametric statistical tests are available to treat data which are inherently in ranks as well as data whose
seemingly numerical scores have the strength of ranks.

4. Non-parametric methods are available to treat data which are simply classificatory, as measured in a nominal
scale. No parametric technique applies to such data.

5. Non-parametric statistical tests are typically much easier to learn and to apply than are parametric tests.

DISADVANTAGE OF NON-PARAMETRIC TESTS:

1. If all the assumptions of the parametric statistical model are in fact met in the data, then non-parametric tests
are wasteful of data.

2. There are as yet no non-parametric methods for testing interactions in the analysis of variance model, unless
assumptions are made about additivity.

HYPOTHESIS TESTING

A hypothesis is a statement or tentative theory which aims to explain facts about the real world. Most
hypotheses have their origin in a question to some practical problem. In search for an answer, “educated
guess” and pertinent evidences are brought out which later on are turned into propositions or hypotheses.

HYPOTHESIS – is a prediction of the possible outcome of the study.

- is a statement about the population characteristic based on findings from a
sample.
- it needs proof or verification.

ASSUMPTION – a statement fact presumed to be true without need of proof or verification.

Examples: Conscious behavior is modifiable.

Habits are modifiable.
Learning is faster among the young than the adults.
 GDD, Panabo National High School 4
Two Types of Hypotheses:

1. Null Hypothesis, Ho – a prediction that no difference between groups or relationship between variables
exists.

Ho: There is no significant difference between the pre-test and post-test scores
or The pre-test and post-test scores do not differ.

2. Alternative Hypothesis, H1 – A prediction that a difference between groups or relationship between variables
exists.

Types of Alternative Hypothesis:

1. Non-directional Hypothesis – simply states that a difference or relationship exists, without qualifying the
direction of the difference or relationship.

Example: Pre-test and post-test score differ. Or There is difference in pre-test and post-test scores.

2. Directional Hypothesis – states that a difference or relationships exists with qualification on the direction of
the difference or relationship.

Example: Post-test scores are higher than the pre-test scores.

SAMPLE STUDY:

TITLE: The Relationship of Some Factors with Teaching Efficiency among High School Teachers
of the Three Schools of Panabo City

PROBLEM: This study seeks to determine the possible relationship of some factors such as sex, age, civil
status, and educational attainment on teaching efficiency.

SPECIFIC PROBLEM:

Specifically this study aims to answer the following questions.

In terms of sex: 1. Do male and female teachers differ in their teaching efficiency?

In terms of age: 2. Is there a correlation between age and teaching efficiency?

In terms of civil status: 3. Do married and single teachers differ in their teaching efficiency?

In terms of educational attainment: 4. Do teachers with no M.A. units, with M.A. units and with M.A
degree differ in their teaching efficiency?

GDD, Panabo National High School 5

HYPOTHESIS:

In terms of sex: Ho: Male and female teachers do not differ in their teaching efficiency.
H1: Male and female teachers differ in their teaching efficiency.

In terms of age: Ho: There is no correlation between age and teaching efficiency.
H1: There is a correlation between age and teaching efficiency.

In terms of civil status: Ho: Married and single teachers do not differ in their teaching efficiency.
H1: Married and single teachers differ in their teaching efficiency.

In terms of educational attainment: Ho: Teachers with no M.A. units, with M.A. units and with M.A. degree
do not differ in their teaching efficiency.
H1: Teachers with no M.A. units, with M.A. units and with M.A. degree
differ in their teaching efficiency.

TITLE: The Relationship of Male and Female Students in their NSAT Performances

Problem: This study seeks to determine the relationship between male and female in their NSAT
performances
Specific Problem: Is there a significant difference between male and female students in their NSAT
performances? Or
Do male and female students differ in their NSAT performances?

Null Hypothesis, Ho: There is no significant difference between male and female students in their NSAT
performance. or
Male and female students do not differ in their NSAT performances.

Alternative Hypothesis, H1: There is significant difference between male and female students in their NSAT
(non-directional) performances. Or
Male and female students differ in their NSAT performances.

(directional) H1: Female students have higher NSAT performances than male students.

TYPES OF HYPOTHESIS ACCORDING TO WHAT IS INTENDED TO BE PROVED.

A. Hypothesis of Association – that which tries to prove the significant relationship between two or more factors
or variables

Example: A study aims to determine whether teachers’ age has a relationship with teaching efficiency

Ho: There is no significant relationship between teachers’ age and teaching efficiency.
H1: There is a significant relationship between teachers’ age and teaching efficiency.

Note: Alternative Hypothesis in association or correlation research is always non-directional.

 GDD, Panabo National High School 6
B. Hypothesis of Difference – that which tries to prove the significant difference between two or more groups in
their scores or means.

Example: A study aims to find out whether male and female teachers differ in their teaching performances

Ho: There is no significant difference between male and female teachers in their teaching
performances.
H1: There is a significant difference between male and female teachers in their teaching
performances. (non-directional or two-tailed test)
Female Teachers have higher teaching performance than male teachers.
(directional or one-tailed test)

Types of Error in Hypothesis Testing:

1. Type I or Alpha Error – occurs when a true null hypothesis is rejected when it is supposed to be accepted.
2. Type II or Beta Error – occurs when a false null hypothesis is accepted when it is supposed to be rejected.

Note: There is no way a researcher can detect which error he may commit.
The likelihood of type I error increases as the level of significance increases.
The likelihood of type II error increases as the level of significance decreases.

Level of Significance – the statistical level at which a null hypothesis is to be proved.

Example: .05 or 5% .01 or 1%

Note: When a null hypothesis is accepted or rejected at .05 or 5% level of significance, there is 95%
probability that the conclusion is true and 5% chance that it is false. A 1% level of significance means
that there is 99% probability that the conclusion or decision is true and 1% chances that it is false. For
test involving statistical decision, it is customary to use a level of significance of 5% or 1%.

CORRELATION – the degree of relationship between two or more factors or variables.

Coefficient of Correlation – the degree of relationship expressed in quantitative terms.

Say: qualitative (high, moderate, low)

Quantitative (0.7, 0.3, 1.0)

Scattergram – a graphical representation showing the relation between two variables.

Example: The given data are the following ages and efficiency ratings of teachers in a certain school. Construct
the scattergram to show any indication of relationship.

Teacher Age Efficiency 45 -

Ratings
40 -
A 28 88 35 - Positive
B 25 90
C 40 85 30 -
D 36 94
25 -
E 32 85
0 - , , , , Negative
80 85 90 95
 GDD, Panabo National High School 7
TYPES OF RELATIONSHIP

1. Positive Relationship – occurs when an individual scoring high in one variable, also scores high in the other
variable. Conversely, an individual scoring low in one variable also scores low in the
other variable.

2. Negative Relationship – occurs when an individual scoring high in one variable, scores low in the other
variable. Conversely, an individual scoring low in one variable, scores low in the
other variable.
Math
95 – Perfect
Example: Student Math Science Positive
Score Score 90 – Relationship

A 94 92 85 –
B 91 90
C 87 88 80 –
D 85 84
E 80 82 0 –, , , , , Science
80 85 90 95

Math
95 – Perfect
Example: Student Math Science Negative
Score Score 90 – Relationship

A 93 81 85 –
B 90 83
C 87 85 80 –
D 84 88
E 80 92 0 –, , , , , Science
80 85 90 95

Perfect Positive Perfect Negative High Positive High Negative Low Positive
Relationship Relationship Relationship Relationship Relationship

Low Negative No Relationship Curvilinear Relationship

Relationship
 GDD, Panabo national High School 8
PEARSON, r – a parametric test of relationship devised by Sir Karl Pearson

Parametric Tests – are those statistical tests that require certain assumptions or conditions about the
population from where the samples are drawn; such as:

a. the population distribution is assumed to be normal

Large Samples – are presumed to approach normality in distribution
- samples of 30 cases and above
Small Samples – are presumed to be not approaching normality
- samples of less than 30 cases

b. the type of measurement data is interval or ratio

Example of interval data: grades, I.Q., test scores

Example of ratio data: weights, speed, height

Non-Parametric Test – are those that do not require certain assumptions or conditions about the
population from where the samples are drawn.

Small samples when used in a study are presumed not to be normally distributed less than 30 cases.

Some conditions required in the use of Pearson, r, (It can be used both in large and small samples)

1. Data used is in interval-ratio data. The two factors involved have some linearly as indicated in a scattergram.
2. Scattergram is recommended to be considered when the samples are small or less than 30.

PROBLEM: A study is conducted to determine the relationship between teachers’ age and teaching efficiency.

Specific Problem: Is there a significant relationship between teachers’ age and teaching efficiency?
(always a question)

Hypothesis: Ho – There is no significant relationship between teachers’ age and teaching efficiency.
Or (Teachers’ age and teaching efficiency are not related)
H1 – There is a significant relationship between teachers’ age and teaching efficiency.
Or (Teachers’ age and teaching efficiency are related)

Statistical Test – Pearson, r

Type of measurement data – Interval scale and Ratio scale

Level of Significance - .05 or 5%

Given Data:
X Y X² Y² XY
Teacher Age Efficiency

A 45 86 2025 7396 3870

B 25 84 625 7056 2100
C 30 88 900 7744 2640
D 55 90 3025 8100 4950
E 21 82 441 6724 1722
N=5 ΣX=176 ΣY=430 ΣX²=7016 ΣY²=37020 ΣXY=15282
X=ΣX/N Y=ΣY/N
 X=35.2 Y=86
GDD, Panabo National High School 9
Formula of Pearson, r

ΣXY – N(X)(Y)
A. r =
ΣX² - N(X)² ΣY² - N(Y)²

N(ΣXY) – (ΣX)(ΣY)
B. r=
N(ΣX²) – (ΣX)² N(ΣY²) – (ΣY)²

Solve the value of Pearson r or the Pearson Product-Moment of Correlation by substituting in the two formulas.

Formula A Formula B
(Please show computation step by step) (Please show computation step by step)

r = 0.81

To the significance of the computed Pearson r, refer to the table of critical values of Pearson r and locate the
table value of r at 0.05 level of significance. Degrees of freedom, df = N-2 = 5 – 2 = 3. The value of r = 0.878.

Since the computed r = 0.81 is less than the table value r = 0.878, there is no significant relationship. Therefore
the null hypothesis is accepted. This means that in this study, age does not influence teaching efficiency.
 GDD, Panabo National High School 10
Critical Values of r for the Pearson Product-Moment Correlation Coefficient
(degrees of freedom = number of pairs of scores – 2)

df .05 .01 df .05 .01

1 .997 .9999 21 .413 .526
2 .950 .990 22 .404 .515
3 .878 .959 23 .396 .505
4 .811 .917 24 .388 .496
5 .754 .874 25 .381 .487
6 .707 .834 26 .374 .479
7 .666 .798 27 .367 .471
8 .632 .765 28 .361 .463
9 .602 .735 29 .355 .456
10 .576 .708 30 .349 .449
11 .553 .684 35 .325 .418
12 .532 .661 40 .304 .393
13 .514 .641 45 .288 .372
14 .497 .623 50 .273 .354
15 .482 .606 60 .250 .325
16 .468 .590 70 .232 .302
17 .456 .575 80 .217 .283
18 .444 .561 90 .205 .267
19 .433 .549 100 .195 .254
20 .423 .537

To test the significance of the computed Pearson r, this formula can also be used.

N–2
t=r
1 - r²

Compute the value of t and refer it to the table value of t.

The table value of the computed t = _______

t = _____

Make a decision/conclusion:

________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
 GDD, Panabo National High School 11
CRITICAL VALUES OF T-TEST

1 tailed α = 0.1 0.05 0.025 0.01 0.005

2 tailed α = 0.2 0.1 0.05 0.02 0.01

df = 1 3.078 6.314 12.706 31.821 3.656

2 1.886 2.920 4.303 6.965 9.925
3 1.638 2.353 3.182 4.541 5.841
4 1.533 2.132 2.776 3.747 4.604
5 1.476 2.015 2.571 3.365 4.032
6 1.440 1.943 2.447 3.143 3.707
7 1.415 1.895 2.365 2.998 3.499
8 1.397 1.860 2.306 2.896 3.355
9 1.383 1.833 2.262 2.821 3.250
10 1.372 1.812 2.228 2.764 3.169
11 1.363 1.796 2.201 2.718 3.106
12 1.356 1.782 2.179 2.681 3.055
13 1.350 1.771 2.160 2.650 3.012
14 1.345 1.761 2.145 2.624 2.977
15 1.341 1.753 2.131 2.602 2.947
16 1.337 1.746 2.120 2.583 2.921
17 1.333 1.740 2.110 2.567 2.898
18 1.330 1.734 2.101 2.552 2.878
19 1.328 1.729 2.093 2.539 2.861
20 1.325 1.725 2.086 2.528 2.845
21 1.323 1.721 2.080 2.518 2.831
22 1.321 1.717 2.074 2.508 2.819
23 1.319 1.714 2.069 2.500 2.807
24 1.318 1.711 2.064 2.492 2.797
25 1.316 1.708 2.060 2.485 2.787
26 1.315 1.706 2.056 2.479 2.779
27 1.314 1.703 2.052 2.473 2.771
28 1.313 1.701 2.048 2.467 2.763
29 1.311 1.699 2.045 2.462 2.756
30 1.310 1.697 2.042 2.457 2.750
60 1.296 1.671 2.000 2.390 2.660
120 1.289 1.658 1.980 2.358 2.617
∞ 1.282 1.645 1.960 2.326 2.576
 GDD, Panabo National High School 12
SPEARMAN rho – a non-parametric test of relationship devised by a statistician named Spearman.
- Does not to assume the normality of the population from where the samples are drawn.
- To be used only for small samples, that is, below 30 cases.
- Applicable when data are in ordinal form or in rank.

PROBLEM: A study is conducted to determine the relationship between teachers’ age and teaching efficiency.

Specific Problem: Is there a significant relationship between teachers’ age and teaching efficiency?
(Always a question)

Hypothesis: Ho – There is no significant relationship between teachers’ age and teaching efficiency.
Or (Teachers’ age and teaching efficiency are not related)
H1 – There is a significant relationship between teachers’ age and teaching efficiency.
Or (Teachers’ age and teaching efficiency are related)

Statistical Test – Spearman rho, ρ

Type of measurement data – ordinal or in rank
Level of Significance - .05 or 5%

Given Data:

X Y Rank of X Rank of Y Difference Bet.

Teacher Age Efficiency Rx Ry Rx – Ry (D) D²

A 45
B 25
C 30
D 55
E 21
86 2 3 -1 1
84 4 4 0 0
88 3 2 1 1
90 1 1 0 0
82 5 5 0 0
ΣD² = 2

6 (ΣD²)
Formula for Spearman rho: rho = 1 -
N (N²- 1)

Compute the value of rho and refer it to table value of rho.

Computed rho = _________

Table value of rho = ________

Make decision/conclusion:

________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
 GDD, Panabo National High School 13
Critical Values for Spearman’s Rank Correlation Coefficient
Use this table to determine the significance of your result for this test. For example, if you had 20 pairs of
data and a value of 0.53 then there would be a probability of between 0.01 and 0.005 that it had occurred by
chance. In other words, you might expect to get this result occurring by chance once every 100-200 times. This,
therefore, indicates a very significant correlation between the two sets of data.
Probability that your result occurred by chance

N (number of pairs)

0.1 0.05 0.025 0.01 0.005

4 1.0000 1.0000 1.0000 1.0000 1.0000

5 0.7000 0.9000 0.9000 1.0000 1.0000
6 0.6571 0.7714 0.8286 0.9429 0.9429
7 0.5714 0.6786 0.7857 0.8571 0.8929
8 0.5476 0.6429 0.7381 0.8095 0.8571
9 0.4833 0.6000 0.6833 0.7667 0.8167
10 0.4424 0.5636 0.6485 0.7333 0.7818
11 0.4182 0.5273 0.6091 0.7000 0.7545
12 0.3986 0.5035 0.5874 0.6713 0.7273
13 0.3791 0.4780 0.5604 0.6484 0.6978
14 0.3670 0.4593 0.5385 0.6220 0.6747
15 0.3500 0.4429 0.5179 0.6000 0.6536
16 0.3382 0.4265 0.5029 0.5824 0.6324
17 0.3271 0.4124 0.4821 0.5577 0.6055
18 0.3170 0.4000 0.4683 0.5425 0.5897
19 0.3077 0.3887 0.4555 0.5285 0.5751
20 0.2992 0.3783 0.4438 0.5155 0.5614
21 0.2914 0.3687 0.4329 0.5034 0.5487
22 0.2841 0.3598 0.4227 0.4921 0.5368
23 0.2774 0.3515 0.4132 0.4815 0.5256
24 0.2711 0.3438 0.4044 0.4716 0.5151
25 0.2653 0.3365 0.3961 0.4622 0.5052
26 0.2598 0.3297 0.3882 0.4534 0.4958
27 0.2546 0.3233 0.3809 0.4451 0.4869
28 0.2497 0.3172 0.3739 0.4372 0.4785
29 0.2451 0.3115 0.3673 0.4297 0.4705
30 0.2407 0.3061 0.3610 0.4226 0.4629
 GDD, Panabo National High School 14
T-test – a parametric test of difference used when data are interval or ratio type.
- It is preferably used when samples are small or less than 30.

TWO TYPES OF SAMPLES:

a) Independent Samples (Uncorrelated) – are those where in the selection of the members of one sample does
not influence or affect the selection of the members of the other
sample.

Example: 1. Male and Female Samples

2. Married and Single Teachers Group
3. Students in College A and College B
4. Experimental and Control Groups in unmatched experimental design.

b) Dependent Samples (Correlated) – are those wherein the selection of the members of one group influence or
affect the selection of the members of the other group

Example: 1. Pre-test and Post-test in one sample experimental design

2. Experimental and control groups in a matched experimental design

PROBLEM: A researcher wishes to know whether male and female students differ in their mathematical
abilities. He administered a mathematics aptitude test to 10 male and 12 female students in a
class.

Specific Problem: Do male and female students differ in their mathematical abilities?

Hypotheses, Ho: Male and female students do not differ in their mathematical abilities.
H1: Male and female students differ in their mathematical abilities.

Statistical Test: T-test for independent or uncorrelated samples. Type of data is interval.

Level of Significance: .05 or 5%

Given Data: Math Scores of Male and Female Students

Male (X1) Female (X2) X1² X2²

85 86 7225 7396
91 87 8281 7569
83 90 6889 8100
87 82 7569 6724
80 84 6400 7056
84 79 7056 6241
78 80 6084 6400
76 82 5776 6724
70 71 4990 5041
75 78 5625 5184
ΣX1 = 809 77 ΣX1² = 65805 5929
N1 = 10 80 6400
X1 = 80.9 ΣX2 = 976 ΣX2² = 78764
N2 = 12
X2 = 81.33
 GDD, Panabo National High School 15
Computation of T-test for Independent Sample (Uncorrelated)

Step 1. Compute the sum of the squares of each group.

(ΣX1)² (809)²
For the Male Group: Σχ1² = ΣX1² - = 65,805 - = 356.9
N1 10

ΣX2² (970)²
For the Female Group: Σ χ1² = ΣX2² - = 78,764 - = 355.7
N2 12

Step 2. Compute the T-test.

X1 – X2
T= =
Σχ1² + Σχ2² 1 1
N1 + N2 - 2 N1 + N2

To test the significance of the computed T, refer to critical values of T. Locate in the table the value of T
for two-tailed (non-directional hypothesis) at .05 level of significance and degrees of freedom, df = N1+N2-2.

Computed Value of T = _______

Table Value of T = _______

Make decision/conclusion/interpretation:

________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________

NOTE: In T-test, any negative sign of computed value of T is disregarded. Use the absolute value in compound
to the table value.

SOLVE THIS PROBLEM: (follow the sequence in solving)

PROBLEM: A researcher wishes to know whether a certain food can enhance the memory recall of students.
He conducted an experiment using an unmatched experimental and control groups. A pre-test was
conducted and both groups had similar memory levels. The experimental group was given a
special food. After the experiment, the post-test scores of the experimental and control groups
were compared and subjected to T-test.

HYPOTHESES:
STATISTICAL TEST:
LEVEL OF SIGIFICANCE:
 GDD, Panabo National High School 16
Given Data: Post-test scores of the controlled and experimental groups

Control Group Experimental Group

X1 X2 X1² X2²
26 29
24 23
18 27
20 18
17 31
18 26
21 29

Computation of T-test for Independent/Uncorrelated Samples:

Step 1: Compute the sum of the squares of each group.

Control Group:

Experimental Group:

Step 2. Compute the T-test

Refer to critical values of T in the table at .05 level of significance. Degrees of freedom, df = N1 + N2 – 2.

The table value of T = __________

The computed value of T = _________

Make decision/conclusion/generalization:

________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
 GDD, Panabo National High School 17
CRITICAL VALUES OF T-TEST

1 tailed α = 0.1 0.05 0.025 0.01 0.005

2 tailed α = 0.2 0.1 0.05 0.02 0.01

df = 1 3.078 6.314 12.706 31.821 3.656

2 1.886 2.920 4.303 6.965 9.925
3 1.638 2.353 3.182 4.541 5.841
4 1.533 2.132 2.776 3.747 4.604
5 1.476 2.015 2.571 3.365 4.032
6 1.440 1.943 2.447 3.143 3.707
7 1.415 1.895 2.365 2.998 3.499
8 1.397 1.860 2.306 2.896 3.355
9 1.383 1.833 2.262 2.821 3.250
10 1.372 1.812 2.228 2.764 3.169
11 1.363 1.796 2.201 2.718 3.106
12 1.356 1.782 2.179 2.681 3.055
13 1.350 1.771 2.160 2.650 3.012
14 1.345 1.761 2.145 2.624 2.977
15 1.341 1.753 2.131 2.602 2.947
16 1.337 1.746 2.120 2.583 2.921
17 1.333 1.740 2.110 2.567 2.898
18 1.330 1.734 2.101 2.552 2.878
19 1.328 1.729 2.093 2.539 2.861
20 1.325 1.725 2.086 2.528 2.845
21 1.323 1.721 2.080 2.518 2.831
22 1.321 1.717 2.074 2.508 2.819
23 1.319 1.714 2.069 2.500 2.807
24 1.318 1.711 2.064 2.492 2.797
25 1.316 1.708 2.060 2.485 2.787
26 1.315 1.706 2.056 2.479 2.779
27 1.314 1.703 2.052 2.473 2.771
28 1.313 1.701 2.048 2.467 2.763
29 1.311 1.699 2.045 2.462 2.756
30 1.310 1.697 2.042 2.457 2.750
60 1.296 1.671 2.000 2.390 2.660
120 1.289 1.658 1.980 2.358 2.617
∞ 1.282 1.645 1.960 2.326 2.576
 GDD, Panabo National High School 18
T – TEST FOR DEPENDENT OR CORRELATED SAMPLES.

PROBLEM: A study was conducted to determine the effects on teachers’ efficiency of a seminar on teaching
effectiveness. A group of 10 teachers were involved. A pre-test efficiency rating was obtained from
the subjects. The seminar was conducted for one week, after which the teachers were encouraged
to apply what they have learned. After a semester, a post-test efficiency rating was obtained from
the same subjects.

Specific Problem: Do pre-test and post-test efficiency ratings differ?

Hypotheses: Ho - Pre-test and post-test efficiency ratings do not differ.
H1 - Posttest efficiency ratings are higher than the pre-test efficiency ratings. (directional)
Statistical test: T-test for dependent or correlated samples using a one-sample experimental design.
Level of Significance: .05 or 5%

T-test formula for Dependent/Correlated Samples:

ƩD ÷ N
t =
ƩD2 – (ƩD)2
N
N(N – 1)
Given Data:

Teacher Pre-test Ratings Post-test Ratings Difference, D D²

X1 X2 X1 – X2 (X1-X2)²
A 82 83 -1 1
B 87 85 2 4
C 90 94 -4 16
D 85 89 -4 16
E 92 91 1 1
F 81 81 0 0
G 80 87 -7 49
H 76 82 -6 36
I 83 88 -5 25
J 77 85 -8 64
N = 10 ΣX1 = 833 ΣX2 = 869 ΣD = -32 ΣD² =212

Compute the T-test:

Refer to table value of T at .05 level of significance, one-tailed test directional hypothesis and degrees of
freedom, df = N – 1.

The table value of T = _______

The computed value of T = _______

Make decision/conclusion/generalization:
________________________________________________________________________________________
________________________________________________________________________________________
 GDD, Panabo National High School 19
Z – TEST

Since the T-test is preferred for small samples of less than 30, an alternative test of difference may
be used for large samples of 30 and above. It is a Z-Test.

The Z-Test Formula:

Where: X1 = Mean of group 1

X2 = Mean of group 2
S1² = Squared SD of group 1
S2² = Squared SD of group 2
N1 = Number of cases of group 1
N2 = Number of cases of group 2

To test the significance of computed Z, refer to table value of Z.

Critical Values of Z

Test Type .05 .01

One-Tailed ± 1.645 ± 2.33
Two-Tailed ± 1.960 ± 2.58

SEATWORK:

Make your own data (N=30) of two variables and find the significant difference of these variables
by using T-test and Z-test. Level of significance is .05.
 GDD, Panabo National High School 20
CHI-SQUARE TEST, ____

Chi-Square Test – a non-parametric test of difference used when data are in frequencies or percentages.
- Used for large and small samples.

Two types of Chi-Square Test (___)

1. Test of goodness-of-fit – to be used when only one variable among categories is present.
2. Test of Independence - to be used when more than one variable among categories are present.

Example of Chi-Square Test of goodness-of-fit:

PROBLEM: A researcher wishes to know whether college students in a certain school differ in their perceptions
of an effective teacher.

Specific Problem: Do students differ in their perceptions of an effective teacher?

Hypotheses; Ho: The students do not differ in their perceptions of an effective teacher.
H1: The students differ in their perceptions of an effective teacher.

Level of Significance: .05 or 5%

Statistical Test: Chi-Square Test (because data are in frequencies) of goodness-of-fit

Given Data
Effective Teacher Number of Student
Respondents
Pleasing Personality 12
Intelligence 25
Good Human Relationship 15
Disciplinarian 8
N=60
Contingency Table
Pleasing Intelligence Good Human Disciplinarian Total
Personality Relations
Observed, O 12 25 15 8 60
Expected, E 15 15 15 15 60

The expected frequency for each cell is obtained by dividing the total observed cases by the
number of categories. In this problem, the expected frequency for each cell is 60 ÷ 4 = 15.

Formula of Chi-Square, ___ Σ(O-E)²

= E
Compute the Chi-Square Test.
 GDD, Panabo National High School 21
To test the significance of computed Chi-Square, refer to critical values of Chi-Square. Locate the
table value under .05 level of significance and degrees of freedom, df = k-1 where k is number of categories.

The computed value of Chi-Square, ___ = ________

The table value of Chi-Square ___ = ________

Make decision/conclusion/generalization:

________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________

SEATWORK:

Problem: A researcher wishes to know the college students’ choices of senators for the 2013 senatorial
elections.

Specific Problem:
Hypotheses; Ho:
H1:
Level of Significance:
Statistical Test:

Given Data: Senatorial Candidates Number of Students

Pacquiao 50
Romualdez 40
Delima 49
Lacson 30
Gatchalian 15
Zubiri 6
190

Make a Contingency Table:

Pacquiao Romualdez Delima Lacson Gatchalian Zubiri Total
Observed, 50 40 49 30 15 6 190
O
Expected,
E

Solve/Compute the value of Chi-Square:

The computed Chi-Square = ________

The table value of Chi-Square = ________

Decsion/Conclusion/Generalization:____________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________

GDD, Panabo National High School 22

CHI-SQUARE TEST OF INDEPENDENCE

Problem: A researcher wishes to know whether male and female students differ in their attitudes toward
divorce.

Specific Problem:

Hypotheses; Ho:
H1:

Level of Significance: .01 or 1%

Statistical Test: Chi-Square Test of Independence because data are in frequencies and two variable categories
are involved.

Given Data: Attitudes Sex Total

Male Female

Agree 25 10 35
Disagree 10 45 55
Undecided 15 5 20
TOTAL 50 60 110

Make a Contingency Table

Agree Disagree Undecided

Sex Total
Observed, O Expected, E Observed, O Expected, E Observed, O Expected, E
Male 25 10 15 50
Female 10 45 5 60
Total 35 55 20 110

Complete the Table: For each cell, obtained the expected frequency by multiplying the column and row totals
and to be divided by the grand total.

Solve the value of Chi-Square.

To test the significance of the computed value, refer to table value of Chi-Square at .01 level of significance.
Degrees of freedom, df = (r-1)(c-1), where r and c are the number of rows and columns,
respectively.
The computed value of Chi-Square = ________

The table value of Chi-Square = ________

GDD, Panabo National High School 23

Decision/Conclusion/Generalization
________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________

Critical Values of Chi-Square

df .05 .01 df .05 .01

1 3.84 6.64 16 26.30 32.00
2 5.99 9.21 17 27.59 33.41
3 7.82 11.34 18 28.87 34.80
4 9.49 13.28 19 30.14 36.19
5 11.07 15.09 20 31.41 37.57
6 12.59 16.81 21 32.67 38.93
7 14.07 18.48 22 33.92 40.29
8 15.51 20.09 23 35.17 41.64
9 16.92 21.67 24 36.42 42.98
10 18.31 23.21 25 37.65 44.31
11 19.68 24.72 26 38.88 45.64
12 21.03 26.22 27 40.11 46.96
13 22.36 27.69 28 41.34 48.28
14 23.68 29.14 29 42.56 49.59
15 25.00 30.58 30 43.77 50.89

ANALYSIS OF VARIANCE (ANOVA)

The F-test

The analysis of variance (ANOVA) is a method for dividing the variation observed into different parts,
each part assignable to a known source, cause or factor. The ANOVA was developed by R.A. Fisher and
reported by him in 1923. ANOVA is used to test the significance of the differences between two or more means
obtained from independent samples. (USED IN THREE OR MORE VARIABLES present in the study)

While the t-test is generally used for the comparison between two means, the F-test may also be applied
for the same purpose. The reason for this is that the F-test is directly related to the t-test. However, the F-test is
generally used for determining differences among more than two means.

The F-test is a parametric test, so that the same assumptions ascribed for a parametric test are made:
a) random selection of subjects from a normal population with equal variances,
b) samples or groups are independent and
c) the analyzed data must be interval.

ONE-WAY ANOVA (One-Factor Analysis of Variance) – when there is only one factor being studied or
one category present as the independent variable. This independent variable may de different experimental
conditions, teaching methods, guidance techniques, values education approaches, educational attainment,
socio-economic status or other factors that may have two or more levels.
 TWO-WAY ANOVA (Two-Factor Analysis of Variance) – when there are two factors being studied or
two categories present as the independent variables.

GDD, Panabo national High School 24

ONE-WAY ANOVA
Example:

Problem: A study was conducted to compare the math performance of third year students when grouped
according to curriculum they are enrolled in.

Specific Problem: Is there a significant difference in the math performance of third year students when grouped
according to curriculum?

Hypothesis, Ho: There is no significant difference in the math performance of third year students when grouped
according to curriculum.

Statistical Test: One-Way ANOVA

Level of significance: .05 or 5%

Data: Math Performance of Third Year Students by Curriculum

BEC SPA STE

X1 X2 X3

10 12 16
12 14 18
14 15 20
12 16 22
16 18 20
15 21
11

STEPS in Hypothesis Testing Using ANOVA

1. Compute the sum of scores in each group. ΣX1 = ΣX2 = ΣX3 =

2. Compute the total sum of scores in all groups. ΣXt = ΣX1 + ΣX2 + ΣX3 =

3. Square the scores in each group. Make a column for these scores.

4. Compute the sum of the squared scores in each group. ΣX1² = ΣX2² = ΣX3² =

5. Compute the total sum of the squared scores in all groups. ΣXt2 = ΣX12 + ΣX22 + ΣX32 =
6. Compute the total number of cases in all groups. Nt = N1 + N2 + N3 =

GDD, Panabo National high School 25

2
7. Calculate the correction factor, C = (ΣXt) ÷ Nt =

2
8. Calculate the total sum of squares. SSt = ΣXt – C =

9. Calculate the sum of squares between groups. SSb = (ΣX1)2÷N1 + (ΣX2)2÷N2 + (ΣX3)2÷N3 – C =

10. Calculate the sum of squares within groups. SSw = SSt - SSb =

11. Compute the degrees of freedom between groups. dfb = K-1 = (K is the number of groups)

12. Compute the degrees of freedom within groups. dfw = Nt – K =

13. Compute the mean square between groups. MSb = SSb ÷ dfb =

14.Compute the mean square within groups. MSw = SSw ÷ dfw =

15. Compute the F-ratio. F = MSb ÷ MSw =

16. Summary Table:

Source of Variation Degrees of Freedom Sum of Squares Mean Squares F-computed F-table
Between Groups
Within Groups
Total

17. Make decision/conclusion/generalization

________________________________________________________________________________________
________________________________________________________________________________________
________________________________________________________________________________________
 GDD, Panabo National High School 26
F Distribution critical values

F Distribution critical values for P=0.10

Denominator
Numerator DF
DF 1 2 3 4 5 7 10 15 20 30 60 120 500 1000
1 39.864 49.500 53.593 55.833 57.240 58.906 60.195 61.220 61.740 62.265 62.794 63.061 63.264 63.296
2 8.5264 8.9999 9.1618 9.2434 9.2926 9.3491 9.3915 9.4248 9.4413 9.4580 9.4745 9.4829 9.4893 9.4902
3 5.5384 5.4624 5.3907 5.3426 5.3092 5.2661 5.2304 5.2003 5.1845 5.1681 5.1513 5.1425 5.1358 5.1347
4 4.5448 4.3245 4.1909 4.1073 4.0505 3.9790 3.9198 3.8704 3.8443 3.8175 3.7896 3.7753 3.7643 3.7625
5 4.0605 3.7798 3.6194 3.5202 3.4530 3.3679 3.2974 3.2379 3.2067 3.1740 3.1402 3.1228 3.1094 3.1071
7 3.5895 3.2575 3.0740 2.9605 2.8833 2.7850 2.7025 2.6322 2.5947 2.5555 2.5142 2.4927 2.4761 2.4735
10 3.2850 2.9244 2.7277 2.6054 2.5216 2.4139 2.3226 2.2434 2.2007 2.1554 2.1071 2.0818 2.0618 2.0587
15 3.0731 2.6951 2.4898 2.3615 2.2729 2.1582 2.0593 1.9722 1.9243 1.8727 1.8168 1.7867 1.7629 1.7590
20 2.9746 2.5893 2.3801 2.2490 2.1582 2.0397 1.9368 1.8450 1.7939 1.7383 1.6768 1.6432 1.6163 1.6118
30 2.8808 2.4887 2.2761 2.1423 2.0493 1.9269 1.8195 1.7222 1.6674 1.6064 1.5376 1.4990 1.4669 1.4617
60 2.7911 2.3932 2.1774 2.0409 1.9457 1.8194 1.7070 1.6034 1.5435 1.4756 1.3953 1.3476 1.3060 1.2989
120 2.7478 2.3473 2.1300 1.9924 1.8959 1.7675 1.6523 1.5450 1.4821 1.4094 1.3203 1.2646 1.2123 1.2026
500 2.7157 2.3132 2.0947 1.9561 1.8588 1.7288 1.6115 1.5009 1.4354 1.3583 1.2600 1.1937 1.1215 1.1057
1000 2.7106 2.3080 2.0892 1.9505 1.8530 1.7228 1.6051 1.4941 1.4281 1.3501 1.2500 1.1813 1.1031 1.0844

F Distribution critical values

F Distribution critical values for P=0.05

Denominator
Numerator DF
DF 1 2 3 4 5 7 10 15 20 30 60 120 500 1000
1 161.45 199.50 215.71 224.58 230.16 236.77 241.88 245.95 248.01 250.10 252.20 253.25 254.06 254.19
2 18.513 19.000 19.164 19.247 19.296 19.353 19.396 19.429 19.446 19.462 19.479 19.487 19.494 19.495
3 10.128 9.5522 9.2766 9.1172 9.0135 8.8867 8.7855 8.7028 8.6602 8.6165 8.5720 8.5493 8.5320 8.5292
4 7.7086 6.9443 6.5915 6.3882 6.2560 6.0942 5.9644 5.8579 5.8026 5.7458 5.6877 5.6580 5.6352 5.6317
5 6.6078 5.7862 5.4095 5.1922 5.0504 4.8759 4.7351 4.6187 4.5582 4.4958 4.4314 4.3985 4.3731 4.3691
7 5.5914 4.7375 4.3469 4.1202 3.9715 3.7871 3.6366 3.5108 3.4445 3.3758 3.3043 3.2675 3.2388 3.2344
10 4.9645 4.1028 3.7082 3.4780 3.3259 3.1354 2.9782 2.8450 2.7741 2.6996 2.6210 2.5801 2.5482 2.5430
15 4.5431 3.6823 3.2874 3.0556 2.9013 2.7066 2.5437 2.4035 2.3275 2.2467 2.1601 2.1141 2.0776 2.0718
20 4.3512 3.4928 3.0983 2.8660 2.7109 2.5140 2.3479 2.2032 2.1241 2.0391 1.9463 1.8962 1.8563 1.8498
30 4.1709 3.3159 2.9223 2.6896 2.5336 2.3343 2.1646 2.0149 1.9317 1.8408 1.7396 1.6835 1.6376 1.6300
60 4.0012 3.1505 2.7581 2.5252 2.3683 2.1666 1.9927 1.8365 1.7480 1.6492 1.5343 1.4672 1.4093 1.3994
120 3.9201 3.0718 2.6802 2.4473 2.2898 2.0868 1.9104 1.7505 1.6587 1.5544 1.4289 1.3519 1.2804 1.2674
500 3.8601 3.0137 2.6227 2.3898 2.2320 2.0278 1.8496 1.6864 1.5917 1.4820 1.3455 1.2552 1.1586 1.1378
1000 3.8508 3.0047 2.6137 2.3808 2.2230 2.0187 1.8402 1.6765 1.5811 1.4705 1.3318 1.2385 1.1342 1.1096
 GDD, Panabo National High School 27
NOTES:
 GDD, Panabo National High School 28
NOTES
 GDD, Panabo National High School 29

NOTES
GDD, Panabo National High School 30