skip to content

DFG Research Unit FOR 2150 | Relativity in Social Cognition

Taking a Measure of Comparative Thinking: Revisiting Assimilation and Contrast

Roland ImhoffRon DotschPaul Barker

The way humans navigate their social environment and respond to situational affordances is inherently relative. In judging other individuals, people use comparison standards to calibrate their judgment whether this particular person is relatively intelligent or not, exhibiting either assimilative tendencies to adjust their judgment in the direction of a comparison standard or contrastive tendencies to move further away. Although there exist prominent models to explain the antecedents of such effects (most prominently, the selective accessibility model), a major drawback of current comparison research is the lack of a sensitive measure. There is no well-validated measure how much people rely on comparative information and to what extent they exhibit assimilative or contrastive tendencies. Without such a widely accepted measure inconsistencies throughout the literature remain ambiguous as to whether they speak to the underlying theory or are mere collateral damages of certain operationalizations. Specifically, inconsistent results may simply follow from people’s diverging perceptions of a given standard, or because a chosen standards was simply just not extreme enough to prompt contrast, or not moderate enough to prompt assimilation. In the present project we propose to further advance the development of a newly invented Comparative Judgment Task (in the first project phase) into a fine-grained measure of comparative processing. Adopting a novel curve-fitting approach will liberate us from the need to provide a consensual definition of what constitutes extreme, respectively moderate standards. Independent of the exact values and allowing for individual and situational differences thereof, any distribution of judgments as a function of comparison standards in a sine-wave form will a) confirm the predictions of the selective accessibility model and b) allow the extraction of exact parameters of comparative processing and assimilative vs. contrastive tendencies. In the present project phase 2 we will further validate this measure across a host of dimensions (work package 1), scrutinize its psychometric properties (work package 2), and provide illustrations of its experimental validation (work package 3). In the unexpected case of difficulties establishing validity in work package 1, we propose an alternative work package 4 that will scrutinize the existing literature to better understand the boundary conditions of typical effects of assimilation and contrast.

1 / 4


Prof. Dr. Roland Imhoff
PI | Curve-Fitting

Johannes Gutenberg-Universität Mainz

Sozial- und Rechtspsychologie
Binger Str. 14-16
55122 Mainz | GERMANY

Telefon +49-6131-39-39291
E-Mail roland.imhoff(at)