Autism Therapy: study design

Scientific experiments are designed, not to prove a therapy works, but rather to put a therapy under a hard test to see if even the most suspicious observer could be convinced that the therapy works. The study design is important, because if the design is poor, the results will be difficult to interpret and will not persuade the suspicious observer. Most studies described on this site are designed to test the effect of a given therapy on a population of children with autism. Ideally, the scientists must state the effect that they are looking for prior to beginning the study. The study then measures whether or not this beneficial effect happens, and whether or not there are any unintended harmful effects of the therapy. Most scientists believe that studies with large numbers of subjects (greater than 100) are more reliable than studies with smaller number of subjects. Other important aspects of study design are whether or not the study includes a placebo or other control.

The best study designs are made by people who truly question whether or not a certain therapy works and really want to put the therapy to the test. The worst study designs come from those who already believe that the therapy works and are trying to create more believers through the use of a false version of the scientific process.