Module 6: t Tests

The Applied Research Center

Module 6 Overview
}  Types of t Tests
}  One Sample t Test
}  Independent Samples t Test
}  Paired Samples t Test
}  Examples
t-Tests
}  Used for inferences concerning one or two means
}  3 main types:
}  One-sample t-test
}  Independent samples t-test
}  Paired samples t-test
One Sample t-Test
}  Used to test whether the mean of single variable differs
from a specified constant.
}  Example
}  A researcher wants to test whether the average IQ score of a
group of students differs from 100.
}  A stats professor wants to determine whether the average
grade on Assignment 1 differs significantly from 23 (an A
average).
One Sample t-Test
}  Step 1: State the Null and Alternate Hypotheses

}  Ho = The average grade on Assignment 1 is equal to 23.

}  Ha = The average grade on Assignment 1 is not equal to

23.

}  Is this a directional or nondirectional Ha?

One Sample t-Test (cont d)
}  Step 2: Input each student s grade into SPSS
}  Step 3: Run the Analysis.
}  Analyze à Compare Means à One Sample T-test
}  Test variable = assign1
}  Test value = 23
}  Click OK
 One Sample t-Test (cont d)

One-Sample Statistics

Std. Error
N Mean Std. Deviation Mean
assign1 15 21.0333 1.54072 .39781

One-Sample Test

Test Value = 23
95% Confidence
Interval of the
Mean Difference
t df Sig. (2-tailed) Difference Lower Upper
assign1 -4.944 14 .000 -1.96667 -2.8199 -1.1134
One Sample t-Test (cont d)
}  Step 4: Make a decision regarding the null
}  M = 21.03, SD = 1.54
}  t (14*) = -4.944
}  p < .001

}  What is the decision regarding the null?

}  *14 = df = n-1 = 15-1 = 14

One Sample t-Test (cont d)
}  Using the level of significance = .05, do we reject or fail to
reject the null?
}  If p < .05, we reject the null
}  if p > .05, we fail to reject the null

}  According to SPSS, p < .001

}  .001 < .05, therefore, we reject the null!

One Sample t-Test (cont d)
}  Step 5: Write up your results.

}  The null hypothesis stated that the average grade on

Assignment 1 is equal to 23. A one sample t-test revealed
that the average grade on Assignment 1 (M = 21.03, SD =
1.54) differed significantly from 23, t (14) = -4.944, p < .
001. Consequently, the null hypothesis was rejected.
Independent t-Test
}  The independent samples t-test is used to test
comparative research questions
}  That is, it tests for differences in two group means or
compares means for two groups of cases.
Independent t-Test (cont d)
}  Example:

}  Suppose the stats professor wanted to determine

whether the average score on Assignment 1 in one stats
class differed significantly from the average score on
Assignment 1 in her second stats class.
Independent t-Test
}  Step 1: State the Null and Alternate Hypotheses

}  Ho = There is no difference between class 1 and class 2

on Assignment 1.

}  Ha = There is a difference between class 1 and class 2 on

Assignment 1.

}  Is this a directional or nondirectional Ha?

Independent t-Test (cont d)
}  Step 2: Input each student s grade into SPSS, along with
which class they are in

Grade Class
20.00 1.00
20.50 1.00
21.00 1.00
20.50 1.00
20.00 1.00
24.50 2.00
23.50 2.00
20.00 2.00
20.00 2.00
Independent t-Test (cont d)
}  Step 3: Run the Analysis.
}  Analyze à Compare Means à Independent Samples T-test
}  Test variable = assign1
}  Grouping variable = class
}  Define Groups:
}  Type 1 next to Group 1
}  Type 2 next to Group 2
}  Click Continue
}  Click OK
 Independent t-Test (cont d)

Group Statistics

Std. Error
class N Mean Std. Deviation Mean
assign1 1.00 14 21.1786 1.48851 .39782
2.00 13 21.9038 1.93525 .53674

Independent Samples Test

Levene's Test for

Equality of Variances t-test for Equality of Means
95% Confidence
Interval of the
Mean Std. Error Difference
F Sig. t df Sig. (2-tailed) Difference Difference Lower Uppe
assign1 Equal variances
4.519 .044 -1.096 25 .283 -.72527 .66152 -2.08771 .637
assumed
Equal variances
-1.086 22.530 .289 -.72527 .66810 -2.10894 .658
not assumed
Independent t-Test (cont d)
}  Step 4: Make a decision regarding the null
}  Class 1 (M = 21.18, SD = 1.49)
}  Class 2 (M = 21.90, SD = 1.94)

}  Which row do we look at on the output?

Independent t-Test (cont d)
}  Step 5: Levene s Test for equal variances
}  Ho = The variances of the two variables are equal.
}  Ha = The variances of the two variables are not equal.

Independent Samples Test

Levene's Test for

Equality of Variances
p = .044, t-test
which is <.05;
Therefore, we reject
F Sig. the
t null and
df doSig.
not(2-tailed
assign1 Equal variances
assumed
4.519 .044 assume
-1.096 equal25 .283
Equal variances
not assumed
variances!
-1.086 22.530 .289
 Independent t-Test (cont d)
}  Looking at the Equal variances not assumed row (the
bottom row)

Independent Samples Test

Levene's Test for

Equality of Variances t-test for

F Sig. t df Sig. (2-tailed)

assign1 Equal variances
4.519 .044 -1.096 25 .283
assumed
Equal variances
-1.086 22.530 .289
not assumed
Independent t-Test (cont d)
}  Make a decision regarding the null
}  t (22.5) = -1.086
}  p = .289

}  Using the level of significance = .05, do we reject or fail to

reject the null?
Independent t-Test (cont d)
}  Remember
}  If p < .05, we reject the null
}  if p > .05, we fail to reject the null

}  According to SPSS, p = .289

}  .289 > .05, therefore, we fail to reject the null!!

Independent t-Test (cont d)
}  Step 5: Write up your results.

}  The null hypothesis stated that there is no difference

between class 1 and class 2 on Assignment 1. An
independent samples t-test revealed that the average
grades on Assignment 1 did not differ significantly from
Class 1
(M = 21.18, SD = 1.49) to Class 2 (M = 21.90, SD = 1.94),
t (22.5) = -1.086, p = .289. Consequently, the researcher
failed to reject the null hypothesis.
Paired Samples t-Test
}  Used to compare the means of two variables for a single
group.
}  The procedure computes the differences between values
of the two variables for each case and tests whether the
average differs from 0.
Paired Samples t-Test
}  Example
}  A researcher wanted to know the effects of a reading
program. The researcher gave the students a pretest,
implemented the reading program, then gave the students a
post test.
Paired Samples t-Test
}  Step 1: State the Null and Alternate Hypotheses

}  Ho = There is no difference in students performance

between the pretest and the posttest.

}  Ha = Students will perform better on the posttest than

on the pretest.

}  Is this a directional or nondirectional Ha?

NOTE for One-tailed Tests!!
}  Remember when we have a directional hypothesis, we
conduct a one-tailed test.
}  When we have a non-directional hypothesis, we conduct a
two-tailed test.
}  SPSS (unless given the choice) automatically runs a 2-
tailed test, IF you have a directional alternate hypothesis
(and a 2-tailed test was run), you MUST divide the p-value
by 2 to obtain the correct p-value!
Paired Samples t-Test
}  Step 2: Set up data

Pre Post
20.00 25.00
21.00 24.00
19.00 23.00
18.00 22.00
20.00 24.00
21.00 25.00
Paired Samples t-Test

}  Step 3: Analyze the Results

}  Analyze à Compare Means à Paired Samples t-Test
}  Paired variables: pre--post
 Paired Samples t-Test (cont d)

Paired Samples Statistics

Std. Error
Mean N Std. Deviation Mean
Pair pre 19.8333 6 1.16905 .47726
1 post 23.8333 6 1.16905 .47726

Paired Samples Test

Paired Differences
95% Confidence
Interval of the
Std. Error Difference
Mean Std. Deviation Mean Lower Upper t df Sig. (2-tailed)
Pair 1 pre - post -4.00000 .63246 .25820 -4.66372 -3.33628 -15.492 5 .000
Paired Samples t-Test (cont d)
}  Step 4: Make a decision regarding the null
–  Pretest (M = 19.83, SD = 1.17)
–  Posttest (M = 23.83, SD = 1.17)
–  t (5) = -15.49
–  p < .001 (two-tailed)

}  What is the decision regarding the null?

 Paired Samples t-Test (cont d)
}  We have a directional alternate, therefore we have to
divide the p-value by 2.

Paired Samples Test

Paired Differences
95% Confidence
Interval of the
Std. Error Difference
Std. Deviation Mean Lower Upper t df Sig. (2-tailed)
0 .63246 .25820 -4.66372 -3.33628 -15.492 5 .000
Paired Samples t-Test (cont d)
}  .000/2 = 0
}  p < .001
}  What is the decision regarding the null?
Paired Samples t-Test (cont d)
}  Using the level of significance = .05, do we reject or fail to
reject the null?
–  If p < .05, we reject the null
–  if p > .05, we fail to reject the null

}  According to SPSS, p < .001

}  .001 < .05, therefore, we reject the null!

Paired Samples t-Test (cont d)
}  Step 5: Write up your results.

}  The null hypothesis stated that there is no difference

in students performance between the pretest and
the posttest. A paired samples t-test revealed that
students scored significantly higher on the posttest
(M = 23.83, SD = 1.17) than they did on the pretest
(M = 19.83, SD = 1.17), t (5) = -15.49, p < .001.
Consequently, the null hypothesis was rejected.
Directional Hypothesis Example
}  Suppose:
–  Ha = Class 1 will score higher on Assignment 3 than Class 2.
–  à Must be based on literature (or prior data/test scores).
}  Run everything the same, only difference is final p-value!
}  According to SPSS, p = .289/2=.145
}  à still fail to reject the null!
Another Example
}  If Ha = Class 1 will score higher on Assignment 3 than
Class 2.
}  And SPSS reported a p-value of .08. When.08/2 = .04,
which IS significant; in this case, we would reject the null!
Module 6 Summary
}  Types of t Tests
}  One Sample t Test
}  Independent Samples t Test
}  Paired Samples t Test
}  Examples
Review Activity
}  Please complete the review activity at the end of the
module.
}  All modules build on one another. Therefore, in order to
move onto the next module you must successfully
complete the review activity before moving on to next
module.
}  You can complete the review activity and module as many
times as you like.
Upcoming Modules
}  Module 1: Introduction to Statistics
}  Module 2: Introduction to SPSS
}  Module 3: Descriptive Statistics
}  Module 4: Inferential Statistics
}  Module 5: Correlation
}  Module 6: t-Tests
}  Module 7: ANOVAs
}  Module 8: Linear Regression
}  Module 9: Nonparametric Procedures