POL 51 Lecture Notes - Lecture 3: Stereotype Threat, Internal Validity, Time Series
Document Summary
Research design i and ii: experimental and observational studies. Dealing with regression to the mean: an extreme value followed by a period in which values return towards the average. Regression to the mean may be the cause for a causal relationship, rather than 2 variables being related to one another. Research design 1: compare pre-intervention score to post intervention score. Research design 2: assign the intervention to cases at random. Potential selection bias: there is something different about 1 group for it to be participating in the actual event vs. the other group taking the placebo. Researcher randomly assigns subjects to one of two groups: the experimental group (take drug) and the control group (no drug) Researcher controls the introduction of the experimental factor (the value of the. Measure the two groups on the dependant variable of interest (blood pressure) If there is a difference, you can attribute it to the experimental variable.