The simplest form of controlled clinical trial is one which duplicates exactly the conditions of the original observations, and which ensures that only the variable of interest is manipulated. Typically, one group receives the AA-B form of treatment and follow-up, and another group receives a treatment that is similar in every respect except that the active ingredient is missing. In human clinical trials, this control group receives all the nonspecific ingredients that might account for the apparent improvement, such as expectancies of being helped, spontaneous improvement, temporary improvement due to alterations in routine, and extra attention. Another important factor controlled in such designs is the bias of the observer who typically is not blind to the treatment, who may be biased in its favor, and who honestly sees improvement.
KeywordsControl Diet Baseline Period Teacher Rating None None Informal Comment
Unable to display preview. Download preview PDF.