Reference no: EM133651074
Question 1: Detective Mute's first thought is that it makes suspects "clam up" and talk less. After a bit more reflection he thinks it may induce more "nervous chatting". He decides to design an experiment to investigate the question by randomly assigning a sample of volunteer informants to either a low-stress group, high-stress group, or no-stress control group. The dependent variable is the percentage of time talking during a mock interrogation. The data is shown below.
| Participant |
Low
Stress
|
Participant |
High Stress |
Participant |
No Stress |
| 1 |
31 |
16 |
41 |
31 |
36 |
| 2 |
5 |
17 |
93 |
32 |
80 |
| 3 |
2 |
18 |
12 |
33 |
7 |
| 4 |
78 |
19 |
0 |
34 |
35 |
| 5 |
32 |
20 |
3 |
35 |
0 |
| 6 |
17 |
21 |
29 |
36 |
24 |
| 7 |
60 |
22 |
0 |
37 |
11 |
| 8 |
30 |
23 |
2 |
38 |
0 |
| 9 |
62 |
24 |
1 |
39 |
18 |
| 10 |
79 |
25 |
4 |
40 |
1 |
| 11 |
62 |
25 |
42 |
41 |
36 |
| 12 |
65 |
27 |
18 |
42 |
13 |
| 13 |
37 |
28 |
60 |
43 |
49 |
| 14 |
58 |
29 |
24 |
44 |
19 |
| 15 |
60 |
30 |
2 |
45 |
0 |
Create an SPSS datafile using this data. Name the independent variable "Stress" and code the no-stress, low-stress, and high-stress groups using values 1, 2, and 3 respectively. Name the dependent variable "PercentTalk". Then create a one-way ANOVA that also includes a planned orthogonal contrast that consists of a first contrast that compares both stress groups (i.e. a combination of the low and high stress groups) to the control group and a second contrast that compares the mean for the low-stress group to the mean for the high-stress group. In the "options" menu tick both "Descriptives" and "Homogeneity of Variance test".
(13) Paste the output boxes for Descriptives, Test of Homogeneity, and the summary table for the ANOVA into your word document.
(14) Is there a significant overall (omnibus) F?
(15) What does it conclude on the basis of the F test?
(16) Paste the "Contrast Coefficients" and "Contrast Tests" output boxes into your word document.
(17) Is Contrast 1 (that compares the mean of the combined group that experienced some level of stress to the mean of the control group) significant?
(18) Is Contrast 2 (that compares the low and high stress means to each other) significant?
(19) Does the independent-groups t-test comparing the low- versus-high stress groups and paste the output here:
(20) Compare the obtained t-statistic for the independent-groups t-test to Contrast 2 in the one-way ANOVA. Why do they differ?
(21) Sometimes the planned orthogonal contrasts do not include all the comparisons of interest. create a protected t-test that compares the mean for the control group to the mean for the low stress mean. {Hint. See the protected t-test slide for the One-Way ANOVA chapter. Or, just add another contrast with the weights +1, -1, 0}
What is the value of the protected t and is it significant?
(22) What is your overall conclusion?
(23) Use the chart builder to generate a bar chart with 95% confidence intervals showing the results of this experiment.
Suppose Detective Mute considered this experiment to be completely exploratory and did not set up any orthogonal contrasts or protected t-tests in advance. Further assume that he wanted to be conservative with respect to the probability of a Type 1 error (that is, rejecting the null when in fact the null is true) and did not want to inflate the familywise alpha. Note that the variances in the three groups are approximately equal and that the Levene's test was non-significant. Also note that there are an equal number of participants in each group. Rerun the oneway ANOVA and choose a post hoc test that seem appropriate.
(24) Paste the output of the post hoc test here:
(25) Which pairwise comparisons are significant?