niversity of North Texas Dr. J. Kyle Roberts © 2004 Unit 5: Comparing Two Sample Means Lesson 1: The Paired Samples t Test EDER 6010: Statistics for Educational Research Dr. J. Kyle Roberts University of North Texas To Reject or Not to Reject? Next Slide
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
University of North Texas Dr. J. Kyle Roberts © 2004
Unit 5: Comparing Two Sample Means
Lesson 1: The Paired Samples t Test
EDER 6010: Statistics for Educational Research
Dr. J. Kyle Roberts
University of North Texas
To Reject or Not to Reject?
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Paired Samples t Test
The same people measured on the same construct on two different occasions
Pre-test ↔ Post-test
(a.k.a., Dependent Samples t Test)
Assessing the effectiveness of a reading intervention with 8 students in a single classroom
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Null Hypothesis
postpre XXH :0
0:0 postpre XXH
0:0 DiffXH
Pre Post Diff
5 8 -3
6 9 -3
7 8 -1
7 9 -2
6 8 -2
4 7 -3
5 7 -2
4 6 -2
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Paired Samples t Test in SPSSAnalyzeCompare MeansPaired-Samples T Test
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Paired Samples t Test in SPSS
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
SPSS OutputPaired Samples Statistics
5.5000 8 1.19523 .42258
7.7500 8 1.03510 .36596
READPRE
READPOST
Pair1
Mean N Std. DeviationStd. Error
Mean
Paired Samples Correlations
8 .808 .015READPRE & READPOSTPair 1N Correlation Sig.
Paired Samples Test
-2.2500 .70711 .25000 -2.8412 -1.6588 -9.000 7 .000READPRE - READPOSTPair 1Mean Std. Deviation
Std. ErrorMean Lower Upper
95% ConfidenceInterval of the
Difference
Paired Differences
t df Sig. (2-tailed)
Mean of the difference scores
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Comparing Studies
Study 1
Pretest Posttest
Mean = 80 Mean = 90
SD = 10 SD = 10
N = 20 N = 20
Study 2
Pretest Posttest
Mean = 80 Mean = 90
SD = 10 SD = 10
N = 121 N = 121
t-crit = 2.093on
19 df
Which study is more likely to produce results that are statistically significant?
t-crit = 1.980on
120 df
Study 2
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Comparing Studies
Study 1
Pretest Posttest
Mean = 80 Mean = 90
SD = 10 SD = 10
N = 20 N = 20
Study 2
Pretest Posttest
Mean = 80 Mean = 90
SD = 10 SD = 2
N = 20 N = 20
t-crit = 2.093on
19 df
Which study is more likely to produce results that are statistically significant?
Study 2
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Comparing Studies
Study 1
Pretest Posttest
Mean = 80 Mean = 90
SD = 10 SD = 10
N = 20 N = 20
Study 2
Pretest Posttest
Mean = 80 Mean = 85
SD = 10 SD = 10
N = 20 N = 20
t-crit = 2.093on
19 df
Which study is more likely to produce results that are statistically significant?
Study 1
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Factors Affecting t-calc
1. The magnitude of the difference between the two means
2. The “spreadoutness” of the scores
3. The number of people in the study
4. The level at which α is set
Just like the Single-Sample t Test
Next Slide
University of North Texas Dr. J. Kyle Roberts © 2004
Unit 5: Comparing Two Sample Means
Lesson 1: The Paired Samples t Test
EDER 6010: Statistics for Educational Research
Dr. J. Kyle Roberts
University of North Texas
To Reject or Not to Reject?
Related Documents
