Quasi-Experimental Design And External Validity In Research

Quasi-Experimental Designs and External Validity In research the goal is to design a study that is generalizable to other populations, settings, and times, however this is not an easy task (Campbell & Stanley, 1963). Researchers try to create studies that have strong external validity, while attempting to avoid possible threats. External validity determines the extent to which the data collected in one study can be generalized to other populations, setting, treatments variables, and measurement variables. Threats to external validity compromise our ability to relate our findings to other groups.
There are four main threats to external validity including, reactive or interaction effect of testing, interaction effects of selection biases, reactive effects of experimental
The Posttest Only Control Group designs solves the reactive or interaction effect of testing by removing the use of a pretest, while the Soloman Four Group Design addresses the same issue by using experimental and control groups with and without pretests, the interaction of testing and the treatment are controlled. Additionally, this threat effects all Quasi-Experimental Designs since they lack a control group, the threat can be eliminated by using design features that include a control group. The second threat is the interaction effects of selection biases and the experimental variable, where using only posttest measures may be suggested as selection biases. Separate-Sample Pretest-Posttest designs account for this threat, by creating random assignment procedures and applying treatment before to some groups and after to others. The third threat to external validity is reactive effects of experimental arrangements, which are typically due to a subject having knowledge that they are participating in a study. This threat can affect the Equivalent Time Samples Design since it has limitation