5. Let’s say we collect data in a sample of 5 people regarding pretest aggressio

Question

5. Let’s say we collect data in a sample of 5 people regarding pretest aggression. The aggression measure is on a scale of 1 = low aggression to 10 = high aggression. After initial data collection, we do an intervention aimed at improving self-control and reducing aggression. After the intervention, we measure the same 5 people with the same aggression measure to get a posttest measure of aggression. Here are the (made-up) data…

Subject ID number

Pretest Aggression

Posttest Aggression

1

5

4

2

1

1

3

8

7

4

3

3

5

4

2

We need to determine if pretest aggression significantly differs from posttest aggression and decide to do a dependent-samples t-test. Compute the observed dependent-samples t test value. Report the t test to three decimal places. HINT: To avoid confusion regarding sign (+/-) of the observed value, subtract the posttest from the pretest to obtain D (pretest-posttest). For full credit, be sure to show all of your work. (2 point)

Subject ID number

Pretest Aggression

Posttest Aggression

1

5

4

2

1

1

3

8

7

4

3

3

5

4

2

Explanation / Answer

H0: There is no difference between means of two dependent groups

H1: There is a difference between means of two dependent groups

alpha:5%

Test statistic: we use paired t test

Difference Scores Calculations

Mean: -0.8
= 0
S2 = SSdf = 2.80/(5-1) = 0.70
S2M = S2/N = 0.70/5 = 0.14
SM = S2M = 0.14 = 0.37

T-value Calculation

t = (M - )/SM = (-0.8 - 0)/0.37 = -2.14

Conclusion:

P value= 0.099301.

P>0.05 we do not reject the H0. Hence claim is not significant

Pretest Aggression Posttest Aggression Diff (T2 - T1) Dev (Diff - M) Sq. Dev 5 4 -1 -0.2 0.04 1 1 0 0.8 0.64 8 7 -1 -0.2 0.04 3 3 0 0.8 0.64 4 2 -2 -1.2 1.44 M: -0.8 S: 2.80

Related Questions

Navigate

• Browse (All)
• Browse 5
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.