Flies in the ointment: Methodological problems in the analysis of the percentage of variance due to persons and situations.

Abstract

The use of omega-squared ratios for the analysis and interpretation of data pertaining to the individual-differences-situationism controversy is critically examined. It is shown that these ratios are formally inappropriate indices when the issue is defined in terms of the limits of transsituational generalization. Coefficients of geheralizability are shown to be the correct indices, and the differing properties of the two indices are discussed. Attention is also paid to the logical and empirical unjustifiability of interpreting the relative size of interaction terms without subsequent multivariate demonstration that these interactions are nonartifactual, replicable, and meaningfully patterned. The need is stressed for methodologically correct analyses of ecologically representative data bases in the development of sophisticated theories that take the joint influence of individual differences and situational factors into account.

Individual differences in situation–behavior profiles: A triple typology model.

“…The model results in a parsimonious representation that allows one to grasp the complex interplay between persons, situations, and behaviors. As such, it may be a useful tool to describe individual differences in situation-behavior profiles as advocated by authors such as Golding (1975), Olweus (1976), Shoda et al (1993Shoda et al ( , 1994, and Wright and Mischel (1987).…”

“…A second important theoretical aspect is to reveal, beyond a formal and parsimonious representation, the psychological mechanisms at the basis of the three typologies and their interrelations (e.g., Golding, 1975;Mischel & Peake, 1982;Mischel & Shoda, 1995;Olweus, 1976;Shoda et al, 1994). We have shown how this can be done by relating the triple typology model to active psychological features of situations, behaviors, and persons.…”

“…Why do different persons (situations, behaviors) belong to the same person (situation, behavior) class (Bem, 1983;Funder & Colvin, 1991;Golding, 1975;Mischel & Peake, 1982;Olweus, 1976;Shoda et al, 1994)? What is the substantive basis of the hierarchical relations and of the if-then rules of the model?…”

“…Several authors other than Bern have underscored the importance of studying personality in terms of meaningful patterns of stability and change in well-defined groups of behaviors in relation to well-defined groups of situations (Allport, 1937;Claeys, Timmers, & Phalet, 1993;Golding, 1975;Mischel & Shoda, 1995;Olweus, 1976;Pervin, 1976;Shoda, Mischel, & Wright, 1993Wright & Mischel, 1987). In particular, Wright and Mischel, building on earlier work of Alston (1975), proposed a conditional approach to the study of personality, in which the fundamental unit of observation is not the unconditional probability of behavior but rather the conditional probability of a certain type of behavior given types of situations or psychological conditions.…”

Interpersonal Aspects of Dangerousness

“…

Although there has been considerable controversy about the relative contributions of personality and environment in the familiar equation B = f(p,e), no one argues that the social context of behavior can be completely ignored (Bowers, 1973;Golding, 1975;Mischel, 1973). Golding writes that the major scientific task is the specification of the manner in which person and situation combine to produce behavior, rather than trying to specify the amounts of variance each accounts for.

…”