Person/Situation (Environment) Assessment

Introduction

This entry gives an overview of the importance of person/situation interaction for the psychological assessment of either persons or situations, and provides guidelines for the assessment of person/situation interaction.

Any individual's specific behaviour such as a stress reaction varies across situations and can be graphically represented as an intraindividual situation profile of that behaviour. A behaviour shows a person/situation interaction if different individuals vary differently across situations in that behaviour (graphically: their situation profiles are not parallel). In such a case, the interindividual differences within situations are not consistent across situations; hence the behaviour correlates lower than 1 between the situations.

Figure 1 illustrates three typical cases: (a) no person/situation interaction between situations 1, 2 (note that the curves of the two individuals differ, however, in the level of the behaviour); (b) ordinal person/situation interaction between situations 2, 3 due to a ceiling effect (the interindividual rank-order is constant across situations but the sizes of the interindividual differences vary); (c) disordinal person/situation interaction between situations 3, 4 (the interindividual rank-order varies across situations; graphically, the profiles cross).

Figure 1. Three typical cases of (no) person/situation interaction. The two curves refer to two individuals (see text).

Importance of Person/Situation Interactions

The importance of person/situation interaction for the assessment of personality traits was first recognized during the consistency debate in the 1970s and 1980s (Mischel, 1968; Kenrick & Funder, 1988). Single behaviours were found to show low cross-situational consistencies, which questioned the assumption of broad personality traits that would allow the prediction of behaviour in heterogeneous situations. However, these studies confounded person/situation interaction and measurement error, and hence overestimated the amount of systematic person/situation interaction. Empirical studies that have controlled measurement error by aggregating trait-relevant behaviour over time or across similar situations have found that the size of person/situation interaction varies greatly between different traits, from virtually zero interaction for subjective well-being to maximum interaction (equivalent to zero cross-situational consistency) for sociability in work versus recreational situations (Diener & Larsen, 1984). Thus, attempts to obtain general estimates of the percentage of variance attributable to persons, situations, and person/situation interaction (e.g. Endler & Hunt, 1966) are in vain.

Furthermore, it has been increasingly recognized that the size of person/situation interaction strongly depends also on the sampling of trait-relevant situations. Whereas it is relatively easy to obtain person samples that are representative of populations such as national age groups or applicants for particular jobs, it is still an unresolved issue how a representative situation sample should be obtained for a behaviour or a personality trait (Ten Berge & De Raad, 1999). The amount of person/situation interaction found will very likely depend on the heterogeneity of the situations in the situation sample. In addition, the extremity of the situations will also influence the amount of person/situation interaction because ceiling and floor effects induce ordinal interactions (see Figure 1).

Finally, person/situation interactions depend on whether an ecologically valid or an experimental design is chosen. In the first case, the frequency of persons in situations is controlled by the persons themselves such as in beeper studies that obtain at random times during the day reports of the current situation as well as experience or behaviour in the situation. In contrast, in experimental designs all persons are assigned to the same situations by the experimenter one by one, and are observed in them or are asked how they would react to them. The size of person/situation interaction tends to be smaller in ecologically valid designs to the extent that persons have opportunities to choose or to avoid certain situations.

...