Research question


Types of Variables

Classify the following variables as dichotomous, nominal, ordinal, continuous, or discrete numerical. Could any of them be modified to increase power, and how?


A. History of heart attack (present/absent)

B. Age

C. Education (college degree or more/less than college degree)

D. Education (highest year of schooling)

E. Race

F. Number of alcohol drinks per day

G. Depression (none, mild, moderate, severe)

H. Percent occlusion of coronary arteries

I. Hair color

J. Obese (BMI ≥ 30)/non-obese (BMI<30)

Types of Validity

An investigator is interested in studying the effect of resident work hour limitations on surgical residents. One area she wishes to address is burnout, and she plans to assess it with two questions (answered on a 7-point scale) from a more extensive questionnaire: (a) “How often do you feel burned out from your work?” and (b) “How often do you feel you’ve become more callous toward people since you started your residency?” The investigator sets out to assess the validity of these questions for measuring burnout. For each of the following descriptions, name the type of validity being assessed:

1. Residents with higher burnout scores were more likely to drop out of the program in the following year.

2. These items seem like reasonable questions to ask to address burnout.

3. Burnout scores increase during the most arduous rotations and decrease during vacations.

4. A previous study of more than 10,000 medical students, residents, and practicing physicians showed that these two items almost completely captured the emotional exhaustion and depersonalization domains of burnout as measured by the widely accepted (but much longer) Maslach Burnout Inventory (West et al. J Gen Intern Med 2009;24:1318–1321).