Rules For Research: A Brief Checklist of Things to Consider for a Credible Experiment
In chapter 1 it was stated that one of the ways of minimizing incorrect conclusions from observations is to have two groups of differently treated people simultaneously compared with each other (as in the contrast of an experimental and control condition). Comparing two or more conditions then, is one of the features of good experimental design. There are a few other rules that are generally followed, along with procedures and constraints that are only appropriate under some conditions. The five most consistently followed rules are specified and discussed in this chapter. These rules can constitute a kind of checklist for proper procedures. Additional concerns that are only relevant in particular circumstances are discussed in appropriate contexts. The rules begin with the familiar requirement of more than one condition in an experiment.
There must be more than one condition in the experiment. There should be at least one condition which can be recognized as a control condition.
The classic distinction between conditions is between an experimental and a control condition. A control condition can only be