APA 2180 Lecture Notes - Lecture 13: Blind Experiment, Selection Bias, Interrupted Time Series
Document Summary
Experiment: modify/intervene with one group, and not with a control group. Usually conducted with issues narrow in scope: basic or natural sciences, social sciences. Manipulaion of iv: modiies an aspect of a situaion, limited to research that can manipulate condiions (smoking example, random populaion?) Control is crucial but not possible to control everything. Researchers someimes use decepion and confederates (milgram student as prison guard) in experimental research. Quasi-experimental: variaions of classical or true experiment. Pre-experimental designs (pilot tesing: no random assignment. Decide how to introduce the iv (treatment) Two group post-test only design: random assignment, treatment, post- test, missing pre-test. Interrupted ime series: no random assignment 1 group, muliple pretests and post-tests. Equivalent ime series: no random assignment, pretest treatment post-test treatment post-test. Lain square design: random assignment but treatments given in diferent orders, post-tests, ex: measuring agility, strength, & power.