Questions for this homework relate to the following journal article.
Campbell RA, Gorman SA, Thoma RJ et al. Risk of Concussion During Sports Versus Physical Education Among New Mexico Middle and High School Students. American Journal of Public Health 2018 108:93-95.
Module 11 HW Concussion article.pdf Download Module 11 HW Concussion article.pdf
Please answer the following questions about this publication and the study it describes.
After reviewing this article (especially the methods), describe 3 possible sources of information bias in this study. Choose one out of the 3 and describe how you might design the study better to minimize the potential bias.
Review the data in Table 1.
a. Compute the risk difference and risk ratio to compare the rates of concussion among boys to girls, both for playing basketball.
b. Show how to compute the risk ratio for concussion while involved in physical education, comparing boys to girls. Did you get the same answer as the authors? How would you interpret the Confidence Interval associated with the relative risk shown in the publication?
3. Turns out that on average, girls participating in sports and physical education in high school are older than boys, because boys are more likely to drop out of high school. During the 4 years of high school, the rate of concussion is highest in 9th grade and then declines over time, with the lowest rate in 12th grade (data not shown here but in other publications).
a. Describe the criteria for confounding using this study as an example, showing how age (such as school year) might be considered a confounder in this study, when comparing the rate of concussion in physical education between boys and girls.
b. If age is a confounder, what analysis or design strategies would you suggest to the investigators to make valid comparisons in the rates of concussion between boys and girls ?
4. Describe 2 possible sources of Selection Bias in this study. Choose one and describe how you might improve the study design or execution to minimize possible selection bias.
