Statistics for the Behavioral and Social

Sciences: A Brief Course

Chapter 8
The t Test for Independent Means

October 22nd, 2020

STAT 101
Learning Objectives

8.1 Describe the structure of a research study for an

independent-measures t test
8.2 Explain the logic of the t test for independent means
8.3 Calculate the standard deviation of the distribution of
differences between means
8.4 Conduct hypothesis testing using a t test for independent
means
8.5 List the assumptions for a t test for independent means

STAT 101
Learning Objectives

8.6 Calculate the effect size for a t test for independent

means
8.7 Determine the appropriate sample size required to
ensure high power when using a t test for independent
means
8.8 Compare the features of the three types of t tests
8.9 Summarize how results of t tests for independent means
are reported in research articles

STAT 101
Using t Tests for Independent Means

Objective: Describe the structure of a research study for an

independent-measures t test
• Hypothesis-testing procedure used for studies with two sets
of scores
o Each set of scores is from an different group of people
and population variance is not known.
▪ Common situation in social sciences research
◦ e.g. treatment group vs control group, women vs
men, patients with or without a particular disease

STAT 101
Example 1: Expressive Writing Study

• Two groups
o Expressive writing group
▪ Writes essays about traumatic experiences
o Control group
▪ Writes essays about plans for the day
• Both groups rate their physical health
o Scale 0-100 (0 = very poor health, 100 = perfect health)

STAT 101
The Distribution of Differences between Means

Objective: Explain the logic of the t test for independent means

• The t test for independent means focuses on the difference
between the means of the two groups.
o The comparison distribution is a distribution of differences
between means.

STAT 101
Creating a Distribution of Differences between Means

• Randomly select one mean from the distribution of means from the
first group’s population
• Randomly selecting one mean from the distribution of means for
the second group’s population
• Subtract the mean from the second distribution of means from the
mean from the first distribution of means to create a difference
score between the two selected means.
• Repeat this process a large number of times and you will have a
distribution of differences between means.

STAT 101
The Logic of a Distribution of Differences between Means

• The null hypothesis is that Population M1 = Population M2

• The population variances are estimated from the sample scores.
• The variance of the distribution of differences between means is
based on estimated population variances.

STAT 101
A diagram of the logic of
a distribution of
differences between
means.

STAT 101
Mean of the Distribution of Differences between Means

• With a t test for independent means, two populations are

considered.
o An experimental group is taken from one of these populations
and a control group is taken from the other population.
• If the null hypothesis is true:
o The populations have equal means.
o The distribution of differences between means has a mean of 0.

STAT 101
Estimating the Population Variance

Objective: Calculate the standard deviation of the distribution of

differences between means
• In a t test for independent means, you calculate two estimates of the
population variance.
o Each estimate is weighted by a proportion consisting of its
sample’s degrees of freedom divided by the total degrees of
freedom for both samples.

STAT 101
Pooled Estimate of the Population Variance

• Weighted average
o The estimates are weighted to account for differences in
sample size .
▪ More weight is given to the larger sample.
• The weighted estimates are averaged
• This is known as the pooled estimate of the population variance.

STAT 101
Example of Figuring Pooled Estimate of Population Variance

S2
Pooled =
df1
( )
dfTotal
S1 +
2 df2
( )
dfTotal
S2 =
2 10
40
30
( 60 ) + ( 80 )
40

1 3
= ( 60 ) + ( 80 ) = 15 + 60 = 75
4 4

STAT 101
Figuring the Variance of Each of the Two Distributions
of Means

• If the samples are not the same size, the distributions of means
taken from them do not have the same variance.
o This is because the variance of a distribution of means is
the population variance divided by the sample size.

STAT 101
Example of Figuring the Variance of Each of the Two
Distributions of Means

S 2
SM2 = Pooled = 75 = 6.82
1 N1 11
S 2
S = Pooled = 75 = 2.42
2
M2
N2 31

STAT 101
Figuring the Variance and Standard Deviation of the Distribution
of Differences between Means (1 of 2)

• The Variance of the distribution of differences

2
between means (SDifference )
o variance of Population 1’s distribution of means plus the
variance of Population 2’s distribution of means.
2
SDifference = SM2 1 + SM2 2
• The standard deviation of the distribution of difference
between means (SDifference ) is the square root of the variance.

SDifference = S 2
Difference

STAT 101
Figuring the Variance and Standard Deviation of the Distribution of
Differences between Means (2 of 2)

2
SDifference = SM2 1 + SM2 2 = 6.82 + 2.42 = 9.24

SDifference = S 2
Difference = 9.24 = 3.04

STAT 101
Steps to Find the Standard Deviation of the Distribution of
Differences between Means (1 of 2)

A. Figure the estimated population variances based on each sample.


2
(X − M)
S =
2
(N − 1)

B. Figure the pooled estimate of the population variance.

df (S 2 ) df (S 2 )
SPooled
2 = 1 1 + 2 2
dfTotal dfTotal

df1 = N1 −1 and df2 =N2 −1; dfTotal =df1+df2

STAT 101
Steps to Find the Standard Deviation of the Distribution of Differences
between Means (2 of 2)
C. Figure the variance of each distribution of means.
S 2
SM2 = Pooled S 2
SM2 = Pooled
1 N1 2 N2
D. Figure the variance of the distribution of differences between means.
2
SDifference = SM2 1 + SM2 2

E. Figure the standard deviation of the distribution of differences

between means.
2
SDifference = SDifference
STAT 101
The Shape of the Distribution of Differences between Means

• Since the distribution of differences between means is based on

estimated population variances:
o The distribution of differences between means is a t
distribution.
o The variance of this distribution is figured based on population
variance estimates from two samples.

40(11− 1= 10; 31 − 1= 30; and 10 + 30 = 40).

STAT 101
The t Score for the Difference between the Two Actual
Means
• Figure the difference between your two samples’ means.
• Figure out where this difference is on the distribution of
differences between means.
M1 − M2
t=
SDifference

Example of Figuring the t Score for the Difference between the Two
Actual Means
M1 − M2 198 − 190
t= = = 2.63
SDifference 3.04

STAT 101
Hypothesis Testing with a t Test for Independent Means

Objective: Conduct hypothesis testing using a t test for independent

means
1. Comparison distribution is a distribution of differences between
means (affects Step 2 of the hypothesis testing procedure).
2. Degrees of freedom for finding the cutoff on the t table is based on
two samples (affects Step 3).
3. Samples’ score on the comparison distribution is based on the
difference between your two means (affects Step 4).

STAT 101
Example of the t Test for Independent Means: Step 1

• Restate the question as a research hypothesis and a null hypothesis

about the populations.
o Population 1: students who do expressive writing
o Population 2: students who write about a neutral topic (their
plans for the day)
o Research hypothesis: Population 1 students would rate their
health differently from Population 2 students (two-tailed tests).
o Null hypothesis: Population 1 students would rate their health
the same as Population 2 students.

STAT 101
Example of t Test for Independent Means: Step 2
(1 of 3)

• Determine the characteristics of the comparison distribution.

o The comparison distribution is a distributions of differences
between means.
o Its mean = 0.
o The shape of the comparison distribution will be a t distribution
with a total of 18 degrees of freedom.

STAT 101
t Test for Independent Means for a Fictional Study of the
Effect of Expressive Writing on Physical Health
Expressive Writing Expressive Writing Group Deviation Expressive Writing Group
Group Score from Mean (Score – M) Squared Deviation from Mean
77 −2 4
88 9 81
77 −2 4
90 11 121
68 −11 121
74 −5 25
62 −17 289
93 14 196
82 3 9
79 0 0
 : 790
summation : 790 Blank 850

STAT 101
Control Writing Control Writing Group Deviation Control Writing Group Squared
Group Score from Mean (Score – M) Deviation from Mean
87 19 361
77 9 81
71 3 9
70 2 4
63 −5 25
50 −18 324
58 −10 100
63 −5 25
76 8 64
65 −3 9
680 Blank 1002

STAT 101
 850 1002
M1 = 79.00; S12 = = 94.44; M2 = 68.00; S22 = = 111.333
9 9
N1 = 10; df1 = N1 − 1= 9; N2 = 10; df2 = N2 − 1= 9
dfTotal = df1 + df2 = 9 + 9 = 18
2
SPooled =
df1
dfTotal
(S1 ) + df (S2 ) = 18 (94.44) + 18 (111.33) = 47.22 + 55.67 = 102.89
2 df2 2 9 9
Total
2
2 SPooled 102.89
S M1 = = = 10.29
N1 10
2
2 SPooled 102.89
S M2 = = = 10.29
N2 10
2
SDifference = SM2 1 + SM2 2 = 10.29 +10.29 = 10.29
2
SDifference = SDifference = 20.58 = 4.54

Needed t with d f = 18, 5% level, two-tailed =  2.101

(M1 − M2 ) (79.00 − 68.00)
t= = = 2.42 Decision: Reject the null hypothesis.
SDifference 4.54
STAT 101
Example of t Test for Independent Means: Step 2 (2 of 3)
• Determine the characteristics of the comparison distribution.
o Figure the estimate population variances based on each
sample.
S12 = 94.44 and S22 = 111.33
o Figure the pooled estimate of the population
variance.2
SPooled = 102.89
o Figure the variance of each distribution of means.
SM2 1 = 10.29 2
SPooled
= SM2
SM2 2 = 10.29 N

STAT 101
Example of t Test for Independent Means: Step 2 (3 of 3)
• Determine the characteristics of the comparison distribution.
o Figure the variance of the distribution of differences between
means.
▪ Adding up the variances of the two distributions of means
would come out to
SDifference = 20.58

o Figure the standard deviation of the distribution of difference

between means.

SDifference = S2
Difference = 20.58 = 4.54

STAT 101
Example of t Test for Independent Means: Step 3

• Determine the cutoff sample score on the comparison

distribution at which the null hypothesis should be rejected.
o You will use a two-tailed test.
o For a significance level of .05
▪ t scores are 2.01 and -2.01

STAT 101
Example of t Test for Independent Means: Step 4

• Determine your sample’s score on the comparison distribution.

t=
( M1 − M2 )
SDifference

t=
( 79 − 68 )
4.54
11
t= = 2.42
4.54

STAT 101
The
distributions
for a t test for
independent
means for the
expressive
writing
example

STAT 101
Example of t Test for Independent Means: Step 5

• Decide whether to reject the null hypothesis.

o Compare your samples’ score on the comparison distribution to
the cutoff t score.
o Your samples’ score is 2.42, which is larger than the cutoff score
of 2.10.
o You can reject the null hypothesis.
o Research hypothesis is supported:
▪ Students who do expressive writing report a higher level of
health than students who write about a neutral topic.

STAT 101
Assumptions of the t Test for Independent Means

Objective: List the assumptions for a t test for independent means

• The population distributions are normal.
• The two populations have the same variance.
• Even if the distributions are not exactly normal or the variances are
not exactly equal, the t test is still pretty accurate.

STAT 101
Effect Size for the t Test for Independent Means

Objective: Calculate the effect size for a t test for independent means
• Estimated Effect Size

(M1 − M 2 )
S pooled

M1 − M2 6.00 − 3.00
Estimated Effect Size = = = 1.47
SPooled 2.04

STAT 101
Power for the t Test for Independent Means

Objective: Determine the appropriate sample size required to ensure

high power when using a t test for independent means
• Power
o Determined using a power table, computer software, or a
power calculator (found online)

STAT 101
Approximate Power for Studies Using the t Test for Independent Means
Testing Hypotheses at the .05 Significance Level
One-tailed test
Number of Participants Effect Size Small (.20) Effect Size Medium Effect Size Medium
in Each Group (.50) (.80)
10 .11 .29 .53

20 .15 .46 .80

30 .19 .61 .92

40 .22 .72 .97

50 .26 .80 .99

100 .41 .97 *

STAT 101
Two-tailed test

Number of Participants Effect Size Small (.20) Effect Size Medium Effect Size Medium
in Each Group (.50) (.80)
10 .07 .18 .39

20 .09 .33 .69

30 .12 .47 .86

40 .14 .60 .94

50 .17 .70 .98

100 .29 .94 *

STAT 101
Power When Sample Sizes Are Not Equal

• Power for a t test for independent means can be determined using a

t table, a power software package, or an Internet power calculator.
• Harmonic Mean is used as the equivalent of each group’s sample
size when determining power when the number of scores in the two
groups differ

STAT 101
Approximate Number of Participants Needed in Each Group
(Assuming Equal Sample Sizes) for 80% Power for the t Test for
Independent Means, Testing Hypotheses at the .05 Significance Level

Blank Effect Size Small (.20) Effect Size Medium (.50) Effect Size Large (.80)

One-tailed 310 50 20

Two-tailed 393 64 26

STAT 101
The t Test for Independent Means in Research Articles

Objective: Summarize how results of t tests for independent means

are reported in research articles
• When found in research articles, the results of these tests are
accompanied by reporting of the means and sometimes the
standard deviations and effect sized.

t (18) = 4.72, p < .01.

STAT 101
STAT 101