Unformatted text preview: - Controls for major threats to internal validity- Two alternative theoretical explanations (H1 & H0) (place 1 & 0 below H). 3.) Selecting subjects: Probability sampling is best, but . ..- Randomizing- Increases confidence that small samples are representative- Why? Because randomization. ..randomizes everything- Matching- What attributes you match on depends on your RQ- Randomization s better- most statistics assume it....
View Full Document
- Fall '08
- representative, light improved participants