SOCI 211 Lecture 10: Lecture 10 - Experiments
Document Summary
Interaction between some aspect of the experiment and the experimental variable. If you pretest someone then their sensitivity to the experimental variable will be changed. Though there is the real effect you can see, it is possible that the results are not generalizable for a larger population only when the pretest is present do you get that experimental effect. Since there is a pretest it is possible the pretest influences. The presence of an interaction between some aspect of the experiment (ex: pretest) and experimental variable. If you have a sample that isn"t weighted, you don"t have that character for research design then you probably don"t have a problem. The meaning is that you have two things present, one the experimental stimulus, two the character of the experiment (pretest, bias, associated selection, etc. ) the combination produces the threat to generalizability. Subject to sampling error, we should have groups with similar characteristics.