The t Test

SHIZA KHAQAN
 The Problem with Z-Scores

The shortcoming of using a z-score for hypothesis

testing is that the z-score formula requires more
information than is usually available.
Specifically, a z-score requires that we know the
value of the population standard deviation.
In most situations, however, the standard deviation
for the population is not known.
In fact, the whole reason for conducting a hypothesis
test is to gain knowledge about an unknown
population.
 Introducing the t statistic

Fortunately, there is a relatively simple solution to

this problem.
When the variability for the population is not known,
we use the sample variability in its place.

The value of standard error:

 Introducing the t statistic

Now we can substitute the estimated standard error

in the denominator of the z-score formula.
The result is a new test statistic called a t statistic:

This formula would be used for calculating the t

statistic by hand. Alternatively, we can use SPSS to
carry out the t test.
 The t Test

The t-Test is a statistical test to determine if there is

a significant difference between two groups.
When a t statistic is used for a hypothesis test,
Cohen’s d can be computed to measure effect size.
 When to use the t test?

When you have a normal distribution

When you want to compare means of two

groups e.g. you may want to compare motivation
among males and females. As these are two groups,
you can use t test. However, if there were more than
two groups you would use a different test.
When you have one categorical and one
continuous variable. In the example above,
gender (male/ female) is categorical and motivation
would be measures on a continuous scale.
 Types of t test

There are three types of t-test:

 The one sample t-test, which is the simplest, determines
whether the observed mean is different from a population
mean.
 The independent t-test is used when comparing means from
two independent groups of individuals.
 The paired t-test is used when comparing the means of two
sets of observations from the same individuals (e.g. repeated
measures design) or from pairs of individuals (e.g. when using
a matched subjects design).
 ASSUMPTIONS OF THE t TEST

Two basic assumptions are necessary for hypothesis tests with the t statistic.
1. The values in the sample must consist of independent observations.
In everyday terms, two observations are independent if there is no consistent,
predictable relationship between the first observation and the second. More
precisely, two events (or observations) are independent if the occurrence of the
first event has no effect on the probability of the second event.

2. The population that is sampled must be normal.

This assumption is a necessary part of the mathematics underlying the development
of the t statistic and the t distribution table. However, violating this
assumption has little practical effect on the results obtained for a t statistic,
especially when the sample size is relatively large. With very small samples, a
normal population distribution is important. With larger samples, this assumption
can be violated without affecting the validity of the hypothesis test. If
you have reason to suspect that the population distribution is not normal, use a
large sample to be safe.
One – Sample t-Test
 One – Sample t - Test

The one sample t test allows us to compare the sample mean

against a known population mean.
Note that there will be just one sample but two means. As
the sample mean will be compared against the sample mean.
Example 1:
 Suppose the population is students of FC, and we know that their mean
CGPA is 3.0. Then we may wish to know if the CGPA of senior
psychology undergraduates is significantly different from the
population.
 So the senior psychology undergraduates would be the sample and if
their mean CGPA is 3.4, then one sample t-test could be used to see if
their CGPA is significantly different from the population.
 One-Sample t-test

Example 2:
 If the average population IQ is 100, we can use the one-sample
t-test to see if a group of advanced Math teachers’ IQ is
significantly higher than the mean population IQ.

Formula for one sample t-test:

Where
t Test for Independent Means
 Independent-measures research design

A research design that uses a separate group of

participants for each treatment condition (or for
each population) is called an independent-
measures research design or a between-
subjects research design.
Example: sex differences, two groups of
participants who undergo different drug conditions:
low dose and high dose
 The independent t test

Formula

The independent t-test

compares the
performance of the
participants in group A
with the performance
of the participants in
group B.
 Examples where independent samples t test would be used

1. Suppose that a researcher wants to compare the

language development among 3 and 4 year olds.
The comparison would be between 2 samples: 3
year olds and 4 year olds and language
development would be measured.
2. A researcher may want to see if there is a
significant difference in the stress levels of
individuals from urban areas and those from
rural areas.
3. If a researcher wanted to look at the level of
academic motivation among men and women.
 Assumptions of the t Test for Independent Means

The first assumption for a t test for independent

means is the same as that for any t test: each of the
population distributions is assumed to follow a
normal curve.
In a t test for independent means, you also have to
be able to assume that the two populations have the
same variance (This assumption is called
homogeneity of variance).
A final assumption of the t test for independent
means is that the scores are independent from
each other.
t Test for Two Related Samples/
Paired t test

T h e P a i r e d t T e s t I s U s ed I n T h e
F o l l o w i n g T w o R e s e a r c h S i t u a ti o n s :

1. M A T C H ED P A I R D E S I GN
2. R E PE A T E D M E A S U R E S D E S I GN
 Matched Pair Design

In a matched-subjects design, each individual in one sample is

matched with an individual in the other sample.
The matching is done so that the two individuals are equivalent (or
nearly equivalent) with respect to a specific variable that the researcher
would like to control.
For example, a researcher could match pairs of subjects on age, gender,
race, and IQ.
As the scores in one set are directly related, one-to-one, with the scores
in the second set, the two research designs (matched pair and
repeated measures) are statistically equivalent and share the
common name related-samples designs (or correlated-samples
designs).
The t-test used for such a design is known by many names: t-test for
dependent means, Dependent t-test, Paired samples t-test, Matched-
pairs t-test
Repeated Measures Design

Time Point 1 Time Point 2

 Repeated Measures Design

A repeated-measures design, or a within-subject

design, is one in which the dependent variable is measured
two or more times for each individual in a single sample. The
same group of subjects is used in all of the treatment
conditions.
The main advantage of a repeated-measures study is that it
uses exactly the same individuals in all treatment conditions.
Thus, there is no risk that the participants in one treatment are
substantially different from the participants in another.
The hypothesis-testing procedure for the situation in which
each person is measured twice (that is, for the situation in
which we have a repeated measures design) is a t test for
dependent means.
 Examples for Repeated measures design

1. A researcher may want participants to take a supplement

and then he may measure memory before and after they
took the supplement.
2. Aggression levels may be measured for a group of children.
Then they may watch an adult behaving aggressively. After
this their aggression levels may be measured again to see if
they changed significantly.
3. A researcher may want to test a new type method of
therapy on people with depression. Their depression
scores would be measured before and after therapy. These
2 sets of scores can then be compared with the
independent sample t test.
 Repeated-Measures (Dependent Means) vs
Independent Measures Designs

Number of subjects: A repeated-measures design typically requires fewer

subjects than an independent-measures design. The repeated-measures design
uses the subjects more efficiently because each individual is measured in both of
the treatment conditions. This can be especially important when there are
relatively few subjects available (for example, when you are studying a rare
species or individuals in a rare profession).
Study changes over time: The repeated-measures design is especially well
suited for studying learning, development, or other changes that take place over
time. Remember that this design involves measuring individuals at one time and
then returning to measure the same individuals at a later time. In this way, a
researcher can observe behaviors that change or develop over time.
Individual differences: The primary advantage of a repeated-measures
design is that it reduces or eliminates problems caused by individual
differences. Individual differences are characteristics such as age, IQ, gender,
and personality that vary from one individual to another. These individual
differences can influence the scores obtained in a research study, and they can
affect the outcome of a hypothesis test.
 Assumptions of the related-samples t test

The related-samples t statistic requires two basic assumptions:

1. The observations within each treatment condition must
be independent. Notice that the assumption of independence
refers to the scores within each treatment. Inside each treatment,
the scores are obtained from different individuals and should be
independent of one another.
2. The population distribution of difference scores (D
values) must be normal. As before, the normality assumption
is not a cause for concern unless the sample size is relatively
small. In the case of severe departures from normality, the
validity of the t test may be compromised with small samples.
However, with relatively large samples (n > 30), this assumption
can be ignored.
 Confidence intervals

A confidence interval consists of an interval of values

around a sample mean, and we can be reasonably
confident that the unknown population mean is located
somewhere in the interval.
The most commonly used confidence level is probably
95%, but values of 80%, 90%, and 99% are also common.
If we use a confidence level of 80%, the resulting interval
is called the 80% confidence interval for μ.
As the confidence interval is constructed around the
sample mean (M), M is located exactly in the center of
the interval.