Distinction between Observational and Experimental Studies Observational the researcher collects information on attributes or measurements of interest but DOES NOT influence the results. Experimental the researcher deliberately influences events (via random assignment of treatments ) and investigates the effects of the intervention. In this brief handout we will discuss several designs which we will encounter throughout the semester, including the strengths and weaknesses of each design. The following table summarizes some study types we will encounter: Types of Study Designs Observational Cross-Sectional Case-Control (Retrospective) Cohort (Prospective) Experimental Randomized Clinical Trial (RCT) Definitions: Cross-Sectional: the study population is ascertained at one point in time, and measurements are taken at that point. Case-Control: two groups of individuals are initially identified: (1) a group that has the disease under study (the cases) and (2) a group that does NOT have the disease under study (the controls). An attempt is then made to relate their prior health habits to their current disease status. This is also called a retrospective study. Cohort: a group of disease-free individuals (the cohort) are identified at one point in time and are followed over a period of time until some of them develop the disease. The development of disease over time is then related to other variables measured at baseline (exposure variables). This is also called a prospective study. Randomized Clinical Trial: this is a type of research design for comparing different treatments, in which the assignment of treatments to patients is by some random mechanism. 1
Example 1.1: Consider the article by Warburton, Bersellini, and Sweeney titled An evaluation of a caffeinated taurine drink on mood, memory, and information processing in healthy volunteers without caffeine abstinence. Source: Psychopharmacology. 2001 Nov; 158(3): 322-328. The authors of this study suggest that the consumption of caffeine can lead to improved attention. Suppose that we are interested in testing this theory further. Specifically, we want to know whether regular consumption of the energy drink Red Bull is associated with improved academic performance among college students studying for the GRE. Where should we start? Study Design #1: When students register for the GRE, a questionnaire is administered. Survey items include the amount of red-bull consumed per week for the previous year, age, sex, undergraduate institution, and place of birth. The outcome of interest is the student s score on the GRE. 2. Why would we collect information on covariates such as age, sex, etc.? 3. Can we use such a study to conclude that Red Bull consumption causes an 2
Study Design #2: A group of graduating seniors who took the GRE and were accepted to highly competitive graduate programs are identified. Another group of graduating seniors who took the GRE and were not accepted are also identified. Again, information on the amount of red-bull consumed per week for the previous year, age, sex, undergraduate institution, and place of birth is obtained. 2. How would you go about selecting the controls to be used in the study? 3. Can we use such a study to conclude that Red Bull consumption causes an 3
Study Design #3: A group of students scheduled to take the GRE are selected for the study. Some are randomly assigned to daily consumption of Red Bull, while others are assigned to daily consumption of a placebo. The GRE score is recorded as the outcome variable. 2. Should subjects know which group to which they have been assigned? Why or why not? 3. Can we use such a study to conclude that Red Bull consumption CAUSES an 4
Example 1.2: Concerned about the possible relationship between cell phone use and head and neck cancer, investigators recruit 10,000 subjects for a study. All are examined and determined to be cancer-free at the start of the study. About half of them are regular cell-phone users, and the other half claim to never have used one and promise not to use one until after the study is complete. The investigators follow the subjects for five years. At the conclusion of the study, the number of occurrences of head or neck cancer is compared across the groups. 2. Can we use such a study to conclude that cell phones cause head or neck cancer? Why or why not? 3. Describe the strengths and weaknesses of this type of design. 5