Lecture
In addition to manipulating the independent variable and measuring the dependent variable, the researcher must take into account the additional variable, the level of which determines the possibility of subsequently generalizing to the type of activity, the populations, and so on being studied.
The additional variable enters into the formulation of the experimental hypothesis (unlike the extraneous variable) as a specification of the conditions under which the action of the independent variable is expected.
Additional variables make the hypothesis less precise, since they presuppose interrelationships of the basic variable with other influences that must be taken into account. At the same time, additional variables also make the hypothesis more concrete, since they indicate the range of action of the regularity under study.
Example: a study of behavior under conditions of risk using the material of lotteries. The participants made hypothetical payments. Lichtenstein and Slovic: the additional variable was the conditions of observing the behavior of gamblers in real life (real bets in Las Vegas). The decisions made under conditions of real payments corresponded to the same regularities that had been discovered under conditions of hypothetical payments.
The researcher always chooses the criterion with respect to which the established dependence can be transferred to other types of reality. The chosen levels of the additional variable will always limit these possibilities of transfer, while at the same time making them more compelling.
The use of groups of participants differing in their level of motivation, of different experimental material, and the variation of other aspects of the experimental conditions are often aimed at broadening the scope of generalization of the dependence under study.
An additional variable present in the experimental hypothesis – the population of potential participants, the type of experimental influences, the ways of recording the dependent variable – are all potential sources for developing factorial designs.
By introducing a secondary independent variable, one controls confoundings and also refines the form of the functional dependence.
An example of the simplest of the functional (quantitative) dependences: the Yerkes-Dodson law of the optimum of motivation (an experiment with mice: three levels of difficulty of the tunnels, with the level of motivation being the strength of the electric shock, which served as reinforcement in learning).
The introduction of the second variable made it possible to clarify precisely the preservation of the form of the established relationship at other levels of task difficulty.
Additional variation is a way of raising the level of generalization of the conclusions from an experimental study. Instead of equating the conditions of the additional variable, it involves systematically varying them for the purpose of subsequently transferring the causal dependence to a multitude of other situations, populations, and so on.
Additional variation involves either using additional levels of the independent variable in order to broaden generalizations about the dependence being established (extrapolating variation), or using intermediate levels of the independent variable in order to predict and refine the form of the relationship between different levels of the independent and dependent variables (interpolating variation).
Interpolating variation involves covering all variations of the variable within the framework of a particular research procedure.
The main assumption here is the assumption of the monotonicity of changes in the variable, and hence the possibility of extending the generalization to those levels that were not represented in the study but that can be additionally included in it.
Extrapolating variation is the extrapolation of a generalization beyond the values of the variable in the study, on the assumption that these additional levels cannot be encompassed by experimental control (a «breakthrough» in generalization)
Comments