KINE 5300:Research Methods in KinesiologyDue Wednesday, May 4, 2005

FinalExam (100 pts)

Directions: Read each question/problem very carefully. Use your notes and the textbook to complete the questions/problems.

  1. (8 pts) Design an experimental, cross-sectional study containing:
  2. 1 discrete (scale) dependent variable
  3. 2 categorical independent variables
  4. 1 ordinal, between-subjects, active independent variable
  5. 1 nominal, between-subjects, attribute independent variable
  1. (8 pts) Design an experimental, longitudinal study containing:
  2. 1 continuous dependent variable
  3. 3 categorical independent variables
  4. 2 ordinal, within-subjects, active independent variables
  5. 1 nominal, between-subjects, attribute independent variable

  1. (15 pts) Traditionally, if a resistance training program is designed, there are specific rest periods (between sets) that can be prescribed to achieve a specific goal. For example, if the primary goal is to increase muscle strength, a rest period of 2 – 5 minutes between sets is recommended. If muscle hypertrophy is the primary goal, 30 – 90 seconds between sets is recommended. If muscular endurance is the primary goal, 20 – 30 seconds is recommended. These recommendations are based largely on older research studies, personal experiences, and basic logic; however, very little scientific evidence is available to support these recommendations.
    Therefore, a researcher conducted a study to examine the effects of different rest period durations on muscle strength, hypertrophy, and endurance in men and women. Sixty men and 60 women (n=120) volunteered to participate in this study. To examine muscle strength, each participant completed a 1RM test for the back squat exercise. To examine muscle hypertrophy, magnetic resonance images were captured for the thigh muscles, and for muscle endurance, each participant completed as many repetitions of the back squat exercise as possible with 50% of 1RM. Participants were matched for strength and randomly assigned to 4 groups with equal numbers of men and women: (1) long rest periods, n=30; (2) medium rest periods, n=30; (3) short rest periods, n=30; and (4) control, n=30. The training program consisted of 12-weeks of resistance training. Groups 1, 2, and 3 completed the same 8 resistance training exercises (matched for volume) 3 times per week, and the control group did nothing. The testing for muscle strength, hypertrophy, and endurance was conducted at the beginning of the program (week 0), 6 weeks into the program (week 6), and immediately after the program (week 12).
    Answer the following questions regarding this study. :
  1. Categorize this experimental design as cross-sectional or longitudinal.
  2. List all of the independent variables in this study.
  3. Classify each independent variable as active or attribute.
  4. Classify each independent variable as within-subjects or between-subjects.
  5. Explain the number of levels for each independent variable.
  6. List all of the dependent variables in this study.
  7. Classify each independent and dependent variable as qualitative or quantitative.
  8. Classify each qualitative variable as ordinal or nominal.
  9. Classify each quantitative variable as discrete or continuous.

  1. (4 pts) Which of the following is a FALSE statement and explain why:
  2. A study can have good internal validity, but poor external validity.
  3. A study can have good internal and external validity simultaneously.
  4. A study can have poor internal validity, but good external validity.
  5. A study can have poor internal and external validity simultaneously.

Explain:

  1. (6 pts) Suppose you are conducting a study that examines the effects of an 8-week fitness program (i.e., walking at 4 mph, 30 min, 4 times/wk) on body composition in middle-aged women (30 – 50 yrs) that work at UTA. You randomly assign 50 participants to either the treatment group (walking program) or the control group (no fitness program). However, 1 week after your study begins, UTA starts the construction of a new parking garage on campus, and as a result, 75% of the women in your study are forced to park about ¾ mile away from their office each day and walk to work. How will this affect the validity of your study?
  1. Poor external validity due to multiple-treatment interference.
  2. Poor external validity due to the interaction effect of testing.
  3. Poor internal validity due to experimental mortality (attrition).
  4. Poor internal validity due to maturation.
  5. Poor external validity due to the reactive effects of the experimental setting.
  6. Poor internal validity due to history.

Explain your answer:

Explain how you would have counteracted this problem:

  1. (6 pts) Suppose you are conducting an experiment that examines the effects of an 8-week circuit resistance training program (i.e., 12 resistance training exercise stations, 30-s per station, 30-s rest between stations, 3 times/wk) on maximal oxygen uptake (VO2max) in sedentary middle-aged men (30 – 50 yrs). You randomly assign 50 participants to either the treatment group (circuit resistance training) or the control group (no training). However, after you conduct the pre-training VO2max test, you meet with each participant to explain their individual VO2max results and their percentile rank based on the ACSM Guidelines manual. As a result, each of your participants realized that they weren’t very “fit” and they all began dieting and increasing their daily physical activity levels during the course of your 8-week study. After you analyzed your data at the end of your study, you realize that both the treatment and control groups increased their VO2max. How would this affect the validity of your study?
  1. Poor external validity due to multiple-treatment interference.
  2. Poor external validity due to the interaction effect of testing.
  3. Poor internal validity due to selection.
  4. Poor internal validity due to maturation.
  5. Poor external validity due to the reactive effects of the experimental setting.
  6. Poor internal validity due to testing.

Explain your answer:

Explain how you would have counteracted this problem:

  1. (5 pts) Why is a control group necessary for a true experimental or quasi-experimental design?

  1. (10 pts) Imagine that you are designing a study to examine the effects of exercise training on muscle flexibility. During a prospective pilot study, you notice that there is great variation among the flexibility measures of your subjects before the training begins: some are very flexible and others are inflexible. This could be a problem, since subjects who are less flexible are more apt to increase their flexibility than those who are already very flexible, which may result in erroneous findings.
    Describe 2 differentmethods of control for the initial flexibility differences among your subjects: (1) before the study begins and (2) after the study is complete.

  1. (10 pts; 2 pts ea) Complete the following graphs:
  1. (12 pts) Calculate the sample size needed to find a significant difference (α=0.05) with a statistical power of 0.80 for the following six examples (show all of your work!):
  1. Example 1 (two independent samples):
  2. Mean of group 0 (μ0) = 153
  3. Mean of group 1 (μ1) = 168
  4. Standard deviation (σ) = 28
  5. Example 2 (two independent samples):
  6. Mean of group 0 (μ0) = 145
  7. Mean of group 1 (μ1) = 168
  8. Standard deviation (σ) = 28
  9. Example 3 (two independent samples):
  10. Mean of group 0 (μ0) = 153
  11. Mean of group 1 (μ1) = 168
  12. Standard deviation (σ) = 12
  13. Example 4 (one dependent sample):
  14. Mean of the group at time=0 (μ0) = 153
  15. Mean of the group at time=1 (μ1) = 168
  16. Standard deviation (σ) = 28
  17. Example 5 (one dependent sample):
  18. Mean of the group at time=0 (μ0) = 145
  19. Mean of the group at time=1 (μ1) = 168
  20. Standard deviation (σ) = 28
  21. Example 6 (one dependent sample):
  22. Mean of the group at time=0 (μ0) = 153
  23. Mean of the group at time=1 (μ1) = 168
  24. Standard deviation (σ) = 12

  1. (6 pts) Using Table 2 (p. 216) of Beckham and Earnest (2003), calculate the sample sizes necessary to find a significant difference (α=0.05) from Androstenedione, day 0 – Androstenedione, day 28, with a statistical power of 0.80 for the following variables:
  1. Weight (kg)
  1. % Body Fat
  1. Total cholesterol (mmol/l)
  1. Hint: Values in parentheses are standard errors (SE). To calculate the standard deviation (σ), use the following equation:
  1. Reference: Beckham, S.G. and C.P. Earnest. Four weeks of androstenedione supplementation diminishes the treatment response in middle aged men. British Journal of Sports Medicine, 37: 212-218. 2003.

  1. (10 pts) Suppose you just completed a study to quantify the test-retest reliability for the 1RM bench press in 8 sedentary, college-aged women. You conducted 1RM bench press trials on Monday and Thursday with no interventions in between. Here are the results:

Trial #1Trial #2

146161

148162

170189

90100

157175

156171

176195

205219

  1. What is the intraclass correlation coefficient (ICC1,3) for the test-retest reliability of this measurement?
  1. Briefly interpret your ICC value. Is it high or low or good or bad, and explain your interpretation.
  1. What is the standard error of measurement (SEM, √MSE)?
  1. Briefly interpret your SEM value. Is it high or low or good or bad, and explain your interpretation.
  1. What is the minimum increase in strength (minimum difference, MD) that is needed in order to be considered “real” for this measurement?
  1. Briefly interpret yourMD value. Is it high or low or good or bad, and explain your interpretation.
  1. If you were to increase the number of subjects in your sample from 8 to 16, what would happen to the MD value? Why?
  1. Briefly describe and interpret the systematic error that was involved in this measurement. Is it high or low or good or bad, and explain your interpretation.
  1. If it was “good,” provide evidence and explain why it was good.
  1. If it was “bad,” provide evidence, explain why it was bad, and suggest a mechanism to correct the bad systematic error.