Randomized Experiments

So You've Decided to Do A Randomized Experimental Design

First, Let's define what a randomized experiment is. A randomized is one in which subjects were assigned to the treatment and comparison groups randomly. That is, once you determine what/who your target population is going to be, your subjects are then randomly assigned to the comparison and treatment groups. This random assignment gives you Probabilistic Equivelance. This means that any differences between the treatment and comparison groups is due to chance: After randomized assignment, the groups are equivelant in probability. This notion Probabilistic Equivelance is the reason that randomized experiments ("true experiments") are the best for establishing cause, or establishing that your treatment or program work. If the randomization was done properly, you know that your groups (in probability) were the same before your treatment (program or intervention), and that any difference is due to your treatment, and most likely not due to differences between your two groups due to the assignment process.

PRE and POST Tests

Now that you have decided to do a randomized design, will you be taking a pre-treatment measurement (a pre-test) as well as a measurement to determine your treatment outcome (a post-test)? If you have the resources, and an appropriate pre-test measure, and, you want to get an idea of where your subjects stand before your treatment, I suggest that you do a pre-test measure.

So, will you do both a pretest and a post test measure, or just a post-test? PRE&POST POST ONlY

Rhonda BeLue