Assignment 1 – Simulating Between-Subject and Within-subject t-tests (Due Jan23/25)

Connect to the simulation page (http://www.ruf.rice.edu/~lane/stat_sim/index.html). Select Repeated Measures. Follow the instructions for how to change the sample size, the mean values, the standard deviation and the value of rho (the correlation between the two groups). Please complete the following steps. Place the percent significant values along with the values used for each simulation in a summary table (you can use the table provided in ACME). Please answer ALL questions and type your answers on another sheet. For Parts A-E use the ‘Simulate 5000’ button. For Part F use the ‘Simulate’ button.

Please hand in both your written answers and the summary tables.

Part A

1) Enter the following numbers into the appropriate spaces in the simulation table.

N	Population Mean A	Population Mean B	Rho (rAB)	Population s.d.
15	18	24	0	8

Now select between-subjects and select the 5000 Simulations button. Then select within-subjects and select the button for 5000 Simulations (make sure that rho = 0). Report the percent significant for both types of t-tests. What does this tell you about the relative power of the two tests when there is no correlation between the two groups?

2) Now do the same thing, but for the within-subjects test use the default setting for rho (default setting for rho = 0.5). What happens to the percent significant? What do these simulations tell you about the relative power of the two tests when there is a relationship between the two groups?

Part B

3) Now change the N to 36 and conduct both tests. Make sure that rho = .5 for the within-subject test. Compare your between-subject result to the between-subject result of question 2. Compare the within-subject result to the within-subject result of question 2. What happens to the percent significant in each case?

4) Change the N to 8 and do the tests. Compare your between-subject and within-subject results to the comparable results of question 2. What happens to the percent significant?

Part C

5) Now change the population s.d. from 8 to 10 and repeat what you did in the question 2 (make sure that rho = .5 for the within-subject test). Compare your between-subject and within-subject results to the comparable results of question 2. What happens to the percent significant?

6) Now change the population s.d to 6 and conduct the simulation. Compare your between-subject and within-subject results to the comparable results of question 2. What happens to the percent significant?

Part D

7) Return the population s.d. value back to 8 as in question 2. Change the Population Mean A to 20 and conduct both a between-subject and a within-subject simulation. Compare these between-subject and within-subject results to the comparable results you obtained in question 2. What happens to the percent significant?

8) Change Population Mean A to 16 and re-do the simulations. What happens to the percent significant for the between-subject and within-subject relative to the results for question 2?

Part E

9) Change the Population Mean A to 24 and re-do the analysis. What is the percent significant for both types of tests?

10) With both population means at 24, change the population s.d. to 10. What is the percent significant for each test?

11) Now change the population s.d. to 6. What is the percent significant for each test?

12) Return the value of the population s.d. to 8. Change the value of rho to .1 for a within-subject test and do the simulation. What is the percent significant for the within-subject test?

13) Change the value of rho to .8 for the within-subject test. What is the percent significant?

14) What have you learned about the percent significant when there is no difference between the means of the two groups?

Part F –use the ‘Simulate’ button not the ‘Simulate 5000’ button for the remaining simulations

15) Enter the following numbers into the simulation table.

N	Population Mean A	Population Mean B	Rho (rAB)	Population s.d.
8	10	15	0	10

Now select within-subjects and then hit the ‘Simulate’ button 8 times. You will see a series of lines appear on the grid below, as well as the t value appear. Observe the lines and record the t-value for each of the 8 simulations.

16) Now change the value of rho to .9 and hit the ‘Simulate’ button 8 times. Observe what happens to the lines and record the t-value for each of the 8 simulations.

17) In general, what happens to the variability of the lines as rho increases?

18) In general, what happens to the t-value as rho increases?

19) What does this tell you about the relationship between rho (r in the equation) and the value of t?

20) Enter the following information into the simulation table.

N	Population Mean A	Population Mean B	Rho (rAB)	Population s.d.
15	10	15	0	10

Select the between-subjects button. Hit the ‘Simulate’ button 8 times, recording the numerator and denominator values, the df and the critical t value each time.

Now enter the following information into the simulation table.

N	Population Mean A	Population Mean B	Rho (rAB)	Population s.d.
15	10	15	0.5	10

Select the within-subjects button. Hit the ‘Simulate’ button 8 times, recording the numerator and denominator, the df and the critical t-value each time.

Calculate the mean numerator and denominator values for each type of test. Also, calculate the mean t-value and provide the appropriate df and critical value for each type of test.

What do you notice about the mean numerator and denominator values across the two types of tests?

What do you notice about the mean t-values, df and critical values for the two types of tests?