CRIM 220 Lecture Notes - Lecture 5: Statistical Conclusion Validity, Regression Analysis, Random Assignment

• School: Simon Fraser University
• Department: Criminology
• Course: CRIM 220
• Professor: William Glackman

Document Summary

Chapte(cid:396) (cid:1011) expe(cid:396)i(cid:373)e(cid:374)tal a(cid:374)d (cid:395)uasi- expe(cid:396)i(cid:373)e(cid:374)tal desig(cid:374)s. An experiment is a process of observation in criminal justice research, to be carried out in a situation expressly brought about. Especially well suited for hypothesis testing components of a true experiment. Subjects are initially measured in terms of d, prior to association with iv (pretested) Re-measured in terms of dv (post-tested) e. g. , diagram of depression study. Dv: improvement in employee satisfaction and productivity e. g. , experimental group - pretest - antidepressant drug - post-test control group - pretest - no drug - post-test placebo, double blind experiments selecting subjects. Goal is to be able to attribute changes in treatment group to the intervention. Group should not be different prior to the intervention. Assign participants to groups randomly experiments meet causal inference criteria. Produces experimental and control groups that should be equivalent. Decide on target population - group which results of your experiment will apply.