Research question: Can exercise reduce stress?

Researchers are interested in how exercising can be beneficial for mental health.

They set up a study in which 150 UCG students are asked to participate in a training program for 6 months. The students are split over 3 groups: Group 1 will follow a program of 50 yoga classes Group 2 will follow a program of 50 kickboxing classes

Group 3 will follow a program of 50 fitness classes

Each student has been asked the same four questions twice to measure their level of stress (before and after following the program). 1. How chaotic was your week on a scale? (0-100) 2. How emotional did you feel this week? (0-100)

3. How positive did you feel this week? (0-100) 4. How did you sleep this week? (0-100)

The reliability of these items has been tested and the Cronbach's alpha of the composite scale is .85. These four questions have been appropriately compiled into the variable 'Pre_stress' and 'Post_stress' that range from 0-100 (0 meaning not stressed at all, 100 meaning very stressed).

1. What are the independent and dependent variables? And what is the level of measurement of the variables? Be specific. Open the data set exercise.sav 2. Investigate whether you can assume the data is normally distributed and explain how you investigated this and what is your conclusion (in max. 300 words). 3. Report the mean and standard deviation for each group as well. 4. Check whether the residuals have a constant variance. Write down which test you used to check this and what your conclusion is based on the test. Explain. 5. Check if there are any outliers. Describe how you did this. If there are outliers, what should you do with them? Explain. 6. What do you think of the operationalization of stress? Explain. 7. Suppose that researchers first want to know if there was a difference between stress levels overall before and after the experiment, regardless of the groups from which the data come. What parametric statistical test could you use to answer this question? Explain your answer. 8. What non-parametric test could you also have used to answer question 7? 9. In this case we will perform the parametric test from your answer in question 7. State the appropriate null and alternative hypothesis. 10. Perform the statistical test that you chose in SPSS. What results did you obtain? Report test statistics, and p-value. Do you think there is sufficient evidence to reject your null hypothesis? Make a new variable 'Stress_difference' by subtracting the post-test stress from the pre-test stress. 11. Now suppose that researchers want to know the differences in means in stress between the exercise programs. Run an ANOVA to find out whether this is the case. 12. State the appropriate null and alternative hypothesis. 13. Perform the statistical test that you chose in SPSS. What did you find? Report test statistics, and p-value. Do you think there is sufficient evidence to reject your null hypothesis? 14. Now suppose that researchers had an hypothesis based on theory before running the experiment, which is that Yoga reduces stress most compared to the other training programs. Run the ANOVA again and include a test for planned comparisons to test this hypothesis. Interpret the planned comparisons and present your results. Indicate what coefficients you used for the planned comparisons.

(exercise.sav)

https://drive.google.com/file/d/1NScaEuA9txNOHN3sBSx9Opzebr-7JK8W/view?usp=sharing