Cognitive structural models of unemployment and employment

Basing predictions on the concepts of a mental model and social representation, we use a network diagram technique, which is a graphical extension and a statistical development of Lunt's matrix method (Lunt, 1989, 1991), to elicit individual models of the causal factors affecting a person's prospects of unemployment (Expt 1) or employment (Expt 2). Such diagrams comprise both direct paths from a factor to the phenomenon and indirect paths in which the connection to it is via some other factor. As we expected, the representation of unemployment is politicized in a way the representation of employment is not. rn both studies, individuals also rated the effectiveness of different actions, based on the diagrammed factors, to affect the target phenomenon. These ratings were best explained by the perceived strength of factor paths rather than simply by the number of paths emanating from that factor. Experiment 2 also showed that the relative importance of direct path strengths in predicting these ratings depended on the order of completing the diagram and rating tasks consistent with expectations derived from the theory of mental models. Finally, using factor analysis and item response theory we show that individual diagrams for employment can be conceptualized as deriving from a single consensual representation. We discuss a view of social representation consistent with these data.