Lecture
In a psychological experiment a cause-and-effect (causal) hypothesis is tested.
By means of organizing experimental manipulations and other forms of control, the experimenter ensures a well-grounded conclusion that the empirical data obtained correspond to the statement «variable X (the IV) acts on variable Y (the DV) in such a way that…»
A cause can only come before an effect.
That is, variable X must precede variable Y in time.
In psychology there are also other conceptions of causality, in the context of which this condition may not hold. (K. Lewin: dynamic causality; Piaget: synchronic causality).
Covariation (a regular connection) between the DV and the IV
The manipulation of the IV takes the form of functional control of the conditions in the experiment and makes it possible to move to conditional statements of the «if…then…» type.
The absence of competing (third) hypotheses.
That is, the elimination of threats to the conclusion that X affects Y coming from other explanations of the changes in Y (explanations by «third variables,» for example).
The confounding (extraneous) variables (EVs) must be duly controlled – the accompanying or systematic changes in variables not included in the hypothesized relationship between X and Y. Otherwise there is a threat to internal validity.
Bredenkamp's scheme:
artifactual conclusion: X→ EV →Y
sound conclusion (provided the third variable is controlled):
X →
X × EV → Y
EV →
An artifactual conclusion may be a consequence of procedural features of how the experiment is conducted.
Thus, experimental control must include: managing the IV, stabilizing the EVs or randomly varying them across levels so as to rule out a correlation of the IV with the EVs.
The problem of interpreting psychological causality is closely connected with the theoretical premises and methodological positions of authors regarding the construction of a psychological explanation. There is a need to distinguish between levels of hypotheses.
An experimental hypothesis pertains to empirically established regularities; a theoretical hypothesis establishes a principle of explanation proceeding from the tenets of one or another psychological theory.
Comments