Lecture
Planning an experiment is organizing its conduct in accordance with the presumed experimental model, through mental comparison with a flawless model of its embodiment, in order to obtain reliable data.
Planning presupposes that in the EH the IV and DV are specified and, at the methodological level, operationalized.
A distinction is drawn between formal and substantive planning.
Formal planning is directed at choosing the scheme, the plan for organizing the interactions, and specifying the minimal effect (in differences of the DV).
The experimental design includes specifying the scheme for presenting the conditions of the IV to different groups of subjects or the sequence of IV levels presented to a single subject, as well as the number of trials or subjects (n). It also includes the plan for recording the DV.
Classifications:
1. The classification of experimental schemes into intra-individual, between-group, and cross-individual is one of the main criteria for classifying experiments, and the choice of scheme is the outcome of planning any experiment.
Intra-individual: «regular alternation», «random sequence», «positional balancing».
Between-group: 1) a simple design for an experimental and a control group without preliminary testing; a simple design for two equivalent groups; 2) a design with preliminary and final testing; 3) the Solomon design.
Cross-individual schemes:
1) the reversal balancing scheme (conditions are presented to two groups of subjects in the forward and reverse order); 2) full balancing for a multi-level experiment (the Latin square, the balanced Latin square).
2. Campbell's criterion – the criterion of a rigorous (true) experiment, in relation to which one can distinguish pre-experimental, experimental, and quasi-experimental designs.
This author linked the identification of true experimental studies to the possibility of randomization as a strategy for assigning subjects to groups and assigning different levels of the experimental factor to equivalent groups.
Pre-experimental designs: the «single-case analysis», the «before-after» design (preliminary and final testing) on a single group of subjects.
The criterion for classifying experiments by object (an individual or a group of people) and by the manner of assigning IV levels (to different groups, or a sequence of trials to a single subject) must be distinguished from the form in which the experiment is conducted.
Comments