Transcript Dependent t-Test - Southeast Missouri State University

Dependent t-Test
CJ 526 Statistical Analysis
in Criminal Justice
When to Use a Dependent tTest
1.
Two Dependent (related) Samples
1.
2.
Repeated-Measures Design (before-after)
Matched-Subjects Design
Example of a Dependent t-Test

A forensic psychologist wants to determine
whether physical exercise in a boot camp
program has an effect on muscular
strength. He/she measures the number of
pull-ups 25 program participants complete
at the beginning of the program (M =
3.28, SD = 1.88) and at the end of the
program (M = 3.6, SD = 1.73).
Example of a Dependent t-Test
-- continued
1.
2.
Number of Samples: 2 (before &
after)
Nature of Samples: dependent
(same subjects at two different
points in time)
Example of a Dependent t-Test
-- continued
3. Independent Variable: participation
in boot camp--exercise
4. Dependent Variable and its Level of
Measurement: number of pull-ups
(ratio)
5. Target Population: boot camp
participants
Example of a Dependent t-Test
-- continued
6. Appropriate Inferential Statistical
Technique: t test, related samples
7. Null Hypothesis: no difference between
the groups before and after boot camp
exercise
8. Alternative Hypothesis: there will be a
difference, boot camp participants will be
able to do more pull ups after training
9. Decision Rule:
1.
If the p-value of the obtained test statistic is less
than .05, reject the null hypothesis, one-tail test
Example of a Dependent t-Test
-- continued
10. Obtained Test Statistic: t
Decision: accept or reject null
hypothesis
12. D.f. = n-1
(in this case n – 1 = 25 – 1 = 24
11.
Results Section
 The
results of the Dependent t-Test
involving participating in a physical
exercise program as the independent
variable and number of pull-ups as
the dependent variable were not
statistically significant.
Discussion Section
 It
appears that participating in a
physical exercise program did not
have an effect on developing
muscular strength among
participants in a boot camp program.
SPSS Paired-Samples t-Test
Procedure
 Analyze,
Compare Means, PairedSamples t-Test
 Move pair of variables over to Paired
Variables
SPSS Paired-Samples t-Test
Sample Printout
Paired Samples Statistics
Pair
1
Score on Drink Index
Score on Drug Index
Mean
26.25
13.10
N
Std. Deviation
14.646
13.924
20
20
Std. Error
Mean
3.275
3.114
Paired Samples Correlations
N
Pair
1
Score on Drink Index
& Score on Drug Index
Correlation
20
Sig.
.783
.000
Paired Samples Test
Paired Differences
Mean
Pair
1
Score on Drink Index Score on Drug Index
13.15
Std. Deviation
9.444
Std. Error
Mean
2.112
95% Confidence
Interval of the
Difference
Lower
Upper
8.73
17.57
t
6.227
df
Sig. (2-tailed)
19
.000
SPSS Paired-Samples t-Test
Printout
 Paired
Sample Statistics
– Paired variables
– Mean
–N
– Standard Deviation
– Standard Error of the Mean
SPSS Paired-Samples t-Test
Printout -- continued
 Paired
Samples Correlations
– Paired variables
–N
– Correlation
– Sig
 p-value
of correlation coefficient
SPSS Paired-Samples t-Test
Printout -- continued
 Paired
Samples Test
– Paired variables
– Paired Differences
 Mean
of the Difference
 Standard Deviation of the Difference
 Standard Error of the Mean of the Difference
 95% Confidence Interval of the Difference
– Lower
– Upper
SPSS Paired-Samples t-Test
Printout -- continued
 t:
obtained test statistic
 df: degrees of freedom
 Sig: p-value
– Divide by 2 to get one-tailed p-value