Vicarious Functioning and Representative Design
I am currently finishing my Ph.D. on legal decision making and will be looking for a postdoc next year. I spent the last year indulging in Brunswikian thinking in Berlin, at the Center for Adaptive Behavior and Cognition, Max Planck Institute for Human Development.
I've been interested in two Brunswikian concepts, namely vicarious functioning and representative design.
Vicarious Functioning. How valid is the process of vicarious functioning as cue substitution rather than cue integration when describing human judgment behavior? Brunswik highlighted that an individual copes with uncertainty in the environment (as defined by probable cause-effect relationships) through the process of vicarious functioning.
Although Brunswik proposed that vicarious functioning could involve either cue substitution or cue integration, he placed more emphasis upon the latter.
Traditionally, the process of cue integration has been represented by a regression model, which was first proposed by Brunswik and later adopted by social judgment theorists.
In a set of studies on legal decision making (working with Peter Ayton, City University, London) and medical decision making (collaborating with Clare Harries, University College, London), I compared the relative descriptive and predictive validity of models that substitute cues and models that integrate cues when deciding on a case.
We found that a simple process model (called the Matching Heuristic) which substitutes cues and is nonlinear and noncompensatory proved better at describing and predicting professionals' judgment behavior than three different structural models (including a regression model), that integrate cues and are characterized by linear, compensatory processing of information.
These findings suggest that the "ritualistic" practice of using regression models to describe human judgment behavior needs to be seriously reconsidered.
Representative Design. What is the nature and extent of the use of representative design in judgment and decision-making research? And does representative design really matter?
Brunswik argued that if we want to understand how an individual adjusts to a probabilistic environment, then we must conduct studies that enable the participant to express any mechanisms developed for achievement.
An experiment must therefore present stimuli that preserve the structure of the environment (i.e., present all available cues, maintain cue values, cue distributions, cue intercorrelations and ecological validities of cues).
Not only is an experiment employing representative design capable of capturing an individual's natural process of vicarious functioning, it will also yield results generalizable to the individual's environment beyond the experimental situation.
Representative design is arguably Brunswik's most important contribution to psychology, although it has largely been ignored. Social judgment theorists have explicitly expressed their commitment to representative design.
But theory and practice are two different things. I have been working on a review paper (with Ralph Hertwig and Ulrich Hoffrage, Max Planck Institute for Human Development, Berlin) that maps the development of representative design, examines the implementation of representative design in empirical studies in the field of judgment and decision making, and assesses the effects of representative design upon research results.
We hope to present the results at the next Brunswik meeting.