COM 381 Lecture Notes - Lecture 9: Internal Validity, Experiment, Quasi
Document Summary
Examines cause and effect, but has poor internal validity. Controls very few of the threats to internal and external validity. Does not have randomization to treatment and control groups. Provides some type of treatment to experimental group. In one-shot case study we expose a group to a treatment x and measure the outcome y. It lacks a pretest y and a control group. It has no basis for comparing groups, or pre- and post-tests. In one-group pretest/posttest design we include the measurement of y before and after treatment x. It has no control group, so no group comparisons. You can state whether change in outcome/dv has taken place. In static-group comparison we have experimental and control groups, but no pre-test. It allows for comparisons among groups, but no pre- and post-tests. No subject selection control = no external. Definition the purpose is to determine efficacy of a treatment. Provides structure to the evaluation of cause and effect.