Professional Documents
Culture Documents
Hypotheses Testing
Hypotheses Testing
The common problems that confront us are the formulation of test procedures or set of
rules that lead to the acceptance or rejection of some statement or hypothesis about a
particular research. For example, a pharmacist researcher might be required to decide on
the basis of experimental evidence whether a certain newly developed drug is much
superior to the one presently being used, a medical technologist might have to decide on
the basis of sample data whether there is a difference in the results of the two laboratory
procedures, a botanist might wish to establish that a certain plant posses a particular
characteristic different from its family because of the geographical location, or a
biochemist might wish to connect that there is a relationship between the genes and
behavioral pattern of certain animals. The procedures for establishing a set of rules that
lead to the acceptance or rejection of these kinds of statements comprise a major area of
biostatistical inference called hypothesis testing.
Hypothesis Testing
One of the principal objectives of research is comparison: How does one group differ
from another? Specifically we may encounter questions such as these: What is the mean
serum cholesterol level of a group of middle – aged men? How does it differ from that of
women? From that of men of other ages? How does today’s level differ from that of
decade ago? Is the latest drug effective in reducing cholesterol levels? What are the
effects of various diets on serum cholesterol levels?
These are typical questions that can be handled by the primary tools of classical statistical
inference – estimation and hypothesis testing. The unknown characteristic, or parameter,
of a population is usually estimated from a statistic computed from a sample. Ordinarily,
we are interested in estimating the mean and the standard deviation of some characteristic
of the population. The purpose of statistical inference is to reach conclusions from our
data and to support our conclusions with probability statements. With such information,
we will be able to decide whether an observed effect is real or due to chance. In this
lesson, we will use a single sample to explain hypothesis testing.
Definitions
Before getting into the step – by step procedure of a test of significance, you will find it
helpful to look over the following definitions.
Null Hypothesis, Ho. A claim that there is no difference between the population mean µ
and the hypothesized value µo.
Alternative hypothesis, H1. A claim that agrees with the null hypothesis. If the null
hypothesis is rejected, we are left with no choice but to fail to reject the alternative
hypothesis that µ is not equal to µo. Sometimes referre to as the research hypothesis.
Test Statistic. A statistic used to determine the relative position of the mean in the
hypothesized probability distribution of sample means.
Critical region. The region on the far end of the distribution. If only one end of the
distribution, commonly termed “the tail”, is involved, the region is referred to as one –
tailed test; if both ends are involved, the region is known as two – tailed test. When the
computed Z or t falls in the critical region, we reject the null hypothesis. The critical
region is sometimes called the rejection region. The probability that a test statistic falls in
the critical region is denoted by .
Critical Value. The number that divides the normal distribution into the region where
we will reject the null hypothesis and the region where we fail to reject the null
hypothesis.
Significance level. The level that your results will most likely not to be criticized.
Usually from 95% and higher.
This explains how to determine if the test is a left tail, right tail, or two-tail test.
Important Properties:
The type I error and type II error are related. A decrease in the probability of one result is
an increase in the probability of the other.
The size of the critical region, and therefore the probability of committing a type I error,
can always be reduced by adjusting the critical value(s).
An increase in the sample size n will reduce and .
If the null hypothesis is false, is a maximum when the true value of a parameter is close
to the hypothesized value. The greater the distance between the true value and the
hypothesized value, the smaller will be.
The critical value is obtained from the t-table. The degrees of freedom for this test is n-1.
If you're performing a t-test where you found the statistics on the calculator (as opposed
to being given them in the problem), then use the VARS key to pull up the statistics in the
calculation of the test statistic. This will save you data entry and avoid round off errors.
Example 1. To illustrate the basic concepts of a test of significance, let us again consider
the Honolulu Heart Study. Suppose someone claims that the mean age of the population
of 7683 individuals is 53.00 years. How can you verify (or reject) this claim? Start by
drawing say a sample of 100 persons. Suppose the sample mean equals 54.85. Now the
question is, “What is the likelihood of finding a sample mean of 54.85 in a sample of 100
from a distribution whose true mean, µ, is 53, given that = 5.50.
1. H0 : µ = 53
H1 : µ # 53
54.85 53
= 5.5 / 100
= 3.36
4. Critical region
Z = =/- 1.96
Useful links:
http://davidmlane.com/hyperstat/logic_hypothesis.html
http://www.math.virginia.edu/~der/usem170/Chapter11/sld008.htm