01:920:311 Lecture Notes - Lecture 4: Experiment, Random Assignment, Quasi
18 Mar 2016
School
Department
Course
Professor
Document Summary
Recap: true experiments (with experimental group and control groups) Dependent variable is decided with control group. How do true experiments address the requirements of causality: empirical association the results support the hypothesis, time ordering intervention came before the post-test, non-spuriousness systematic random assignment to groups. Has some but not all features of a true experiment (when one cannot be done: to hold a true experiment could be unethical, or when independent variables cannot be assigned at random. Non-equivalent control group designs: the experimental and control groups are determined before the treatment and are not randomly assigned = non-equivalent , ex. a teacher agrees to teach a new math curriculum whiles others use the hold. We already know there is something different about this teacher (more excited/motivated maybe) Use groups that are similar (ex. using students compared to siblings raised in same home) Or similar work groups at different locations.
