COMM 1 Lecture Notes - Lecture 13: Ice Cream Van, Experiment, Hawthorne Effect
23 May 2020
School
Department
Course
Professor
Document Summary
Naturally over time people mature/get better over time. If you find someone with test scores at the extreme, and then they improve: mortality: particular people of interest in the study might drop out because they are sensitive to the subject of study. But there will be mortality across all conditions: you might be in trouble if people drop out unequally across, reactivity effects: participant"s reaction to being studied, rather than to groups because of the manipulation. Same thing again (true experiment etc. ) but also: Use double blind procedure: neither the researcher nor the participant knows which group is which (they don"t know if they have a real pill or a placebo) No random assignment, but they try to make up for that. Quasi-equivalent control group design: y1, y1. Y2 (group 2: these two groups are the same in some aspect that controls for extraneous variables. Use pre-test scores to match groups before manipulation.
