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 processes, or accompany one another, while the question of their mutual determination remains open.
The plans of correlational studies must 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 «teachers»).
1. Plans with one group of subjects.
- each subject of the single group is represented within it by at least two measurements of variables, whether two different indices or one and the same index assessed at a different point in time.
2. Plans with two or more groups of subjects.
The plan must also include forms of control of the extraneous variable. Three basic forms of control:
1) Stabilization of the extraneous variable, or selection of the levels of the values of the main variables such that the extraneous variable appears in the form of accounted-for levels defining a factorial plan of its interrelations with at least one of the two measured variables.
2) Statistical control (a large n; selecting a sample that varies across the entire range).
- The plan with homogeneous groups (the equivalence of the groups and the difference between them on the main variate, which serves as an analog of the independent variable, is controlled); the extraneous variable here is fixed at a particular level or 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; applied in cases where the initial number of subjects is small.
The method of contrasting groups; A study of adjustment
In the method of contrasting groups, the difference defined by the main variate is amplified by the aforementioned accompanying variables, and individuals with average scores are excluded.
Quasi-experiments are the term for those schemes of organizing data collection that make it possible to test hypotheses while orienting toward the use of the norms established in the construction of studies and inferences on the basis of the experimental method.
Here a reduction in the forms of control is assumed.
Two main directions of the reduction of forms of control in the formation of experimental and control groups can be identified:
the failure to fulfill the condition of randomization in the selection of subjects;
treating as an analog of the independent variable the difference between the groups that was introduced as the basis for the groups' non-equivalence.
1. The strategy of selecting groups by a given attribute.
Matching pairs of subjects – makes it possible to regard a quasi-experimental study as similar to a correlational one. The attribute chosen as the variate here can be formulated either as an analog of the dependent variable or as an analog of the independent variable (the reconstructed causally acting variable). The subjects matched into a pair will differ on some attribute. It is precisely this distinguishing attribute that will lie at the basis of the difference between the homogeneous groups in a quasi-experiment.
Example: Bandura and Walters – the connection between adolescent aggression and rejection in the family. The adolescents in each pair differed in their aggression scores.
2. The plan with a nonequivalent control group.
In educational research, the plan with a nonequivalent control group is widespread (one of the quasi-experimental plans with control reduced down to the organization of interventions).
3. The mixed plan of controlling external and internal conditions.
A variable of internal conditions (individual differences among people) is confounded with a variable of external conditions: for example, propensity for risk (internal) affected the dependent variable (the decision to remand into custody), and it was confounded with variables of external conditions (the availability of free places, the situation in which the decision was applied)
Forms of control common to quasi-experiments and correlational studies:
1. The application of statistical control.
Increasing the number of subjects in order to detect confoundings of the basic process with basic extraneous processes.
Unlike in correlational studies, here the inferences are built differently.
One part of quasi-experimental schemes is built on the basis of true (in Campbell's sense) experimental plans, but here there are limitations in the control of threats to validity (e.g., because of the non-equivalence of the groups).
The other part – involves only the researcher's choice of when and on whom to conduct the measurements, and is characterized by the absence of experimental control.
Here statistical control is carried out – control after the main empirical part of the study has been completed (post factum control).
For this it is necessary to additionally measure the basic extraneous processes (the sources of confounding).
- Post factum control
The general scheme constitutes a system of comparing the influence of basic extraneous processes, measured as individual differences, on the basic process being studied.
Control of the inference includes 3 stages:
1) The stage of the initial selection of groups of differing composition, differing in the levels of the main basic process;
2) The experimental comparison of the values of the dependent variable in these groups, the search for the «effect» of the basic process;
3) Control of the confounding of the identified main effect with basic extraneous variables.
- The temporal trend (there is nothing like this in correlational studies)
The time-series scheme entails the possibility of repeated (prolonged) measurement of the dependent variable both before and after an influence regarded as experimental.
The resulting curves reflect the dynamics of temporal changes – these are the temporal trends. In analyzing the temporal trend, time is an analog of the independent variable. If the experimental influence changes the level during a particular time interval, this is a disruption of the temporal trend.
Comments