Lecture
An experimental design or scheme is a plan for presenting all levels of the IV to participants or groups of participants in order to test a causal hypothesis. For a factorial experiment, the design includes a specification of the conditions in which the levels of two or more variables are combined. According to the principle of isolated conditions, the functional control of each IV occurs independently of the other.
A complete design encompasses all the formally assumed combinations of the independent variables.
Besides complete designs, patchwork designs are used, in which, for some reason, not all the groups of DV measurements assumed by a complete design are present. Sometimes they arise from the need to evaluate a newly emerged competing hypothesis, or else for reasons of economy.
Example: Merton, a thought experiment. Are those whose abilities manifested themselves earlier better recognized within the scientific community? The first variable is «age», the second is «the success of subsequent scientific activity». But we run into the problem of the unfeasibility of assembling the sample (highly successful subsequent activity combined with late entry into higher education: Lomonosov, and that is all).
It turns out that a patchwork design can be used here (selecting three groups rather than four).
Campbell assigns patchwork designs to quasi-experimental schemes, demonstrating their indispensability when it is impossible to select equivalent groups.
Example: testing the hypothesis about the influence of the command and flight training of cadets at an air force school on their attitudes toward leadership and on the change in their attitude toward superiors and subordinates. It is impossible to assemble a group of cadets who have not undergone command training (one cannot deprive them of their education!). Two groups with different lengths of service were compared.
Group 1: one week remained until the end of the training period (the control group)
Group 2: was tested twice: at the beginning and at the end of training (the experimental group).
The first comparison was carried out on the DV measured simultaneously in both groups (in the experimental group it was the beginning of training).
The second comparison concerned the «before-after» effect in the experimental group.
Comments