Parallel group study design

There are two forms of the f-test, and each is applicable for sets of measurements that have been obtained in different ways. The method of data collection precisely and uniquely determines which of these two forms of statistical analysis is appropriate. Section 7.6 introduced the independent groups Mest, which is appropriate for the analysis of data collected during a study employing a parallel group study design. Another form of the Mest is called the dependent measures t-test. This test is sometimes called the related measures Mest, the repeated measures Mest, or the Mest for matched pairs. The name dependent measures Mest has been chosen here since the contrast with the word independent in the name independent groups Mest is clear. 

A cross-over design can be used. Where this is not appropriate a parallel group study design should be chosen. 

Parallel-group studies are typically thought of as the most straightforward design case. In fact, a bewildering array of variations exists within this class. 

Other important issues are to do with the basic structure of trials. For example, what are the appropriate comparator treatments How many should be studied simultaneously Will the object of the trial be to find differences or to demonstrate equivalence Should a parallel group study or a cross-over trial be used How many centres should be used What should the block size be If a cross-over trial is employed will it be necessary to use incomplete blocks (These are designs in which each patient receives only a subset of the treatments being studied.) How long should the treatments be applied When should values be measured How should they be combined Should the results be analysed sequentially If the trial is to be blinded, how should this be achieved Is stratification desired What factors should be included in the analysis What is the general approach to be adopted What is the strategy for dealing with multiplicity of outcomes ... 

Finally, trials can follow parallel or crossover study designs. In a parallel trial, patients are assigned to a therapy that they remain on, and they are compared with patients in alternate therapy groups. In a crossover trial, patients switch or change therapy assignments during the course of the trial. 

The parallel-group, double-blind, placebo-controlled study design represents the golden standard of acute treatment trials of depression, mania and anxiety disorders. This design is intended to limit bias, in particular selection and measurement bias. Trials based on this design are expected to provide information about the effect size of a new compound and its side-effect profile. 

---