Chapter

10: The t
Test For
Two
Independe
nt Samples
Chapter 10 Learning Outcomes

Understand structure of research study appropriate

for independent-measures t hypothesis test

Test the difference between two populations or two

treatments using independent-measures t statistic

Evaluate the magnitude of the observed mean

difference (effect size) using Cohen’s d, r2, and/or
a confidence interval

Understand how to evaluate the assumptions

underlying this test and how to adjust calculations
when needed
Independent-Measures
Designs

Allows researchers to evaluate the mean difference between

two populations using data from two separate samples.

The identifying characteristic of the independent-measures

or between-subjects design is the existence of two separate
or independent samples.

Thus, an independent-measures design can be used to test for

mean differences between two distinct populations (such as
men versus women) or between two different treatment
conditions (such as drug versus no-drug conditions).
 The Hypotheses and the Independent-
Measures t Statistic
• Null hypothesis for independent-measures test (no
difference between the population means)

H 0 : 1   2 0
• Alternative hypothesis for the independent- measures test
(there is a mean difference)

H1 : 1   2 0
 The Formulas for an Independent-
Measures Hypothesis Test
• The basic structure of the t statistic:

( M 1  M 2 )  ( 1   2 )
t
s(M  M )
1 2

• t = [(sample statistic) – (hypothesized population

parameter)] divided by the estimated standard error
 Pooled Variance & Estimated
Standard Error
2
s s22
1
s( M  M 2)
 
1
n1 n2
• The above equation shows the standard error concept but is unbiased only if
the two samples are exactly the same size ( n1 = n2)
• Pooled variance (sp2) provides an unbiased basis for calculating the standard
error

2 2
SS1  SS 2 s p s p
2
s 
p
df1  df 2
s( M 1  M 2 )  
n1 n2
 The Final Formula and Degrees of
Freedom
• The independent-measures t statistic = sample mean
difference minus population mean difference divided by
the estimated standard error
• Degrees of freedom (df) for the t statistic is df for first
sample + df for second sample

df df1  df 2 (n1  1)  (n2  1)

Let’s Try This!
Which combination of factors is most likely to produce a
significant value for an independent-measures t statistic?

A. a small mean difference and small sample variances

B. a large mean difference and large sample variances
C. a small mean difference and large sample variances
D. a large mean difference and small sample variances
Let’s Try This!
• T/F
• If both samples have n = 10, the independent-
measures t statistic will have df = 19.
• T/F
• For an independent-measures t statistic, the
estimated standard error measures how much
difference is reasonable to expect between the
sample means if there is no treatment effect.
Let’s Try This!
An independent-measures t statistic is used to
evaluate the mean difference between two
treatments with n=8 in one treatment and n=12 in
the other. What is the df value for the t statistic?
Let’s Try This!
One sample from an independent-measures
study has n=4 with SS=100. The other sample
has n=8 and SS=140.
• Compute the pooled variance
• Compute the estimated standard error for
the mean difference
Let’s Try This!
• One sample from an independent-measures
study has n=9 with a variance of s2=35. The
other sample has n=3 and s2=40.
• Compute the pooled variance
• Compute the estimated standard error for
the mean difference
Hypothesis Testing with the
Independent-Measures t
Statistic
• Step 1.
• State the hypotheses and select the α level. For
the independent-measures test, H0 states that there
is no difference between the two population means.
• Step 2.
• Locate the critical region. The critical values for
the t statistic are obtained using degrees of freedom
that are determined by adding together the df value
for the first sample and the df value for the second
sample.
Hypothesis Testing with the
Independent-Measures t
Statistic
• Step 3.
• Compute the test statistic. The t statistic for the independent-measures
design has the same structure as the single sample t introduced in Chapter 9.
However, in the independent-measures situation, all components of the t
formula are doubled: there are two sample means, two population means, and
two sources of error contributing to the standard error in the denominator.
• Step 4.
• Make a decision. If the t statistic ratio indicates that the obtained difference
between sample means (numerator) is substantially greater than the difference
expected by chance (denominator), we reject H0 and conclude that there is a
real mean difference between the two populations or treatments.
The Critical Region for the Independent-Measures
Hypothesis Test (df = 14; α = .05)
 Measuring Effect Size for the
Independent-Measures t
• Effect size for the independent-measures t is measured in the same way that
we measured effect size for the one-sample t in Chapter 9.
• You can compute an estimate of Cohen’s d or you can compute r2 to obtain a
measure of the percentage of variance accounted for by the treatment effect.

2
estimated mean difference M  M2 t
estimated d   1 r2  2
estimated standard deviation s 2p t  df

• For confidence intervals:

(M1-M2) ± t s(M1-M2)
Confidence Intervals
& Hypothesis Tests
Let’s Try This!

For an independent-measures research study, the value of

Cohen’s d or r2 helps to describe _____.

A. the risk of a Type I error

B. the risk of a Type II error
C. how much difference there is between the two
treatments
D. whether the difference between the two treatments is
likely to have occurred by chance
 • Report whether the
difference between the two
Reporting groups was significant
the Results • Report descriptive statistics
of an (M and SD) for each group
Independent- • Report t statistic and df
Measures t • Report p-value
• Report CI immediately after t,
Test for example, 95% CI [0.782,
7.218]
APA format for reporting
t-tests
An independent-samples t-test was conducted. Alpha was set at .05. The students
who watched Sesame Street as children had higher high school grades (M = 93, SD
= 4.71) than the students who did not watch the program (M = 85, SD = 4.22), t(18) =
4.00, p < .05, d = 1.79.

Or, if using SPSS, report the exact p-value:

t(18) = 4.00, p = .001, d = 1.79.

Or, if reporting confidence intervals:

t(18) = 4.00, p =.001, 95% CI [3.798, 12.202].

Report the use of a one-tailed test explicitly in the research report, for example,
t(14) = –2.67, p ≤ .01, one-tailed
Let’s Try This!

• T/F
• If a researcher reports that t(6) = 1.98, p > .05, then H0
was rejected
Let’s Try This!

An educational psychologist would like to determine whether access

to computers has an effect on grades for high school students. One
group of n=16 students has home room each day in a computer
classroom in which each student has a computer. A comparison
group of n=16 students has home room in a traditional classroom. At
the end of the school year, the average grades for the computer
students are M=86, SS=1005 and grades for the traditional students
are M=82.5, SS=1155.
• Test difference
• Compute a 95% confidence interval
• Write it in APA format
APA Write-up

An independent-samples t-test was conducted. Alpha was set

at .05. At the end of the school year, the students who had a
computer homeroom had the same grades (M = 86, SD =
8.19) as the students who had a traditional homeroom (M =
83, SD = 8.77), t(30) = 1.166, p > .05, 95% CI [-2.626, 9.626].
The Role of Sample Variance
and Sample Size in the
Independent-Measures t Test

• The standard error is directly related to sample

variance
• Because larger variance leads to a larger error,
it also leads to a smaller value of t (closer to
zero) and reduces the likelihood of finding a
significant result
• Because larger samples produce smaller error,
larger samples lead to a larger value for the t
statistic
Two Sample Distributions Representing
Two Different Treatments
Two Sample Distributions Representing
Two Different Treatments
Assumptions Underlying the
Independent-Measures t Formula

• The observations within each sample must be independent

• The two populations from which the samples are selected
must be normal
• The two populations from which the samples are selected
must have equal variances (called homogeneity of
variance)
The Homogeneity of Variance
Assumption

Most hypothesis tests

usually work
reasonably well even if
the set of underlying
assumptions are Requires that the two
violated. populations from which
The one notable the samples are
exception is the obtained have equal
assumption of variances
homogeneity of Necessary in order to
variance for the justify pooling the two
independent-measures sample variances and
t test. using the pooled
variance in the
calculation of the t
statistic
 Hartley’s F-Max Test
• Test for homogeneity of variance

s 2 (largest)
F  max  2
s (smallest)

• Use Table in Appendix B

• Large value indicates a large difference between the sample variances
• Small value (near 1.00) indicates similar sample variances and that the
homogeneity assumption is reasonable
• If homogeneity of variance is violated, use the alternative procedure for
computing the t statistic that does not involve pooling the two sample variances.
• Adjustment will result in a lower df, making it more difficult to reject the null
hypothesis
Let’s Suppose 2 independent
samples each have n=10
with variances of 12.34 and
Try 9.15. For these data, is the
assumption of homogeneity

This!
violated?
 Sample SPSS Output for the Independent-
Measures Hypothesis Test
Group Statistics
VAR00002 N Mean Std. Deviation Std. Error Mean
VAR00001 1.00 8 8.0000 2.92770 1.03510
2.00 8 12.0000 3.07060 1.08562

Independent Sample test

Levene’s Test for Levene’s Test for Test for Test for Equality of
Equality of Equality of Equality of Means: df
Variances: F Variances: Sig. Means: t
VAR00001 Equal variances .000 1.000 −2.667 14
assumed

Equal variances −2.667 13.968

not assumed

Independent Sample test

Test for Test for Test for Test for Equality Test for Equality of
Equality of Equality of Equality of of Means: 95% Means: 95% Confidence
Means: Sig. Means: Means: Std. Confidence Interval of the
(2-tailed) Means Error Interval of the Difference: Upper
Difference Difference Difference: Lower
VAR0000 Equal variances .018 −4.00000 1.50000 −7.21718 −.78282
1 assumed
Equal variances .018 −4.00000 1.50000 −7.21786 −.78214
not assumed
Study this video
and then when
you feel
confident, take
Questions?the
quiz!