Lecture
Correlational studies are those in which hypotheses about relationships are tested and the psychological variables are either manifestations of different aspects of one and the same basic process, or accompany one another, while the question of their mutual determination remains open.
Based on the results of correlational studies, the relationships between variables may be as follows:
1. The observed dependence is probably causal, but the direction of the relationship may be any.
2. Variables A and B are not linked by a causal dependence, but are part of a complex of interacting variables such that other causal dependences among some other variables generate the correlation between A and B.
- The relationship between variables may be mediated by one or several intervening variables.
- A and B may be consequences of a single cause (spurious correlation).
It is impossible to choose among the cases considered without experimental control.
The designs of correlational studies should be regarded as forms of control in obtaining empirical data.
Example: Rushton's correlational study. (the personality characteristics of two groups of teachers: «researchers» and «educators»).
1. Designs with a single group of participants.
- each subject of the single group is represented in it by at least two measurements of variables, whether these are two different indicators or the same indicator assessed at a different point in time.
Hypotheses about relationships may in this case presuppose different ways of collecting data, since the data-collection design depends on taking into account substantive interpretations.
Variants of designs with a single group:
1) Variables X and Y reflect different properties of the reality under study, which are represented by the indicators of different methods. When a scatter diagram is used, the ranges of dispersion of each of the indicators are set on the axes, and synchronous correlations are usually computed. It is assumed that the variates are characterized by values measured at one and the same point in time. Synchrony may mean that the temporal order in which the methods are applied is irrelevant, since the indicators are relatively stable.
2) Variables X1, X2…Xm reflect the dispersions of the values of one and the same variate, but measured at different points in time. Here autochthonous correlations are computed (the correlation of a variate with itself).
The adequacy of a correlational study design includes assessing the possibility of covering the entire range of the measured variable, or of refining it in order to formulate the hypothesis about relationships correctly. The researcher often does not know this range and is therefore forced to resort to increasing the number of measurements (to increasing the sample) in order to control non-systematic confoundings.
When using a single-group design, the researcher resorts to statistical control, which means controlling the variability of variables (variates and extraneous variables) by increasing the number of participants in the sample.
2. Designs with two or more groups of participants.
The design must also include forms of controlling the extraneous variable. Three main forms of control:
1) Stabilization of the extraneous variable, or selection of the levels of the values of the main variables so that the extraneous variable appears in the form of accounted-for levels that set up a factorial design of its interrelationships with at least one of the two measured variables.
2) Statistical control (a large n; selection of a sample that varies across the entire range).
- Design with homogeneous groups (the equivalence of the groups and the difference between them along the main variate, which acts as an analogue of the independent variable, are controlled); here the extraneous variable is fixed at a particular level or is functionally controlled by forming subgroups with particular levels of these extraneous variables.
A study of the dependence of IQ on birth order
3) Matching of pairs; used in cases where the initial number of participants is small.
The method of contrasting groups; a study of adjustment
In the method of contrasting groups, the difference set by the main variate is amplified by the said accompanying variables, and individuals with intermediate scores are excluded. Criticism: such a strategy reduces the samples of participants to sizes at which there arises a threat to covering the full range of values of the main variables; one cannot be sure that all differences within pairs are controlled and that a significant extraneous variable has not been overlooked.
Comments