Assimilation and contrast in perceptual judgments

An attempt is made to specify the differential conditions underlying two opposite context effects: assimilation and contrast. Assimilation occurs when the judgment of the current stimulus shifts in the direction of the preceding stimulus; contrast occurs when the judgment shifts in the opposite direction. Both effects are brought together and demonstrated in a single paradigm. It is hypothesized that the number of contextual stimuli is the differentiating factor. Two alternative hypotheses, the differential stimulation hypothesis and the temporal position hypothesis, are also tested. The differential stimulation hypothesis proposes that the intensity difference between the contextual and judged stimulus determines whether assimilation or contrast occur, and the temporal position hypothesis proposes that varying the temporal position of the contextual stimulus is the critical factor. While the results do not support either of these alternative hypotheses, it is shown that the number of contextual stimuli does determine whether assimilation or contrast occur.     

Assimilation and contrast illusions: Differences in plasticity

Three experiments show differences in the plasticity of the contrast and assimilation portions of the Delboeuf, Ebbinghaus, and Ponzo illusions. Contrast illusions show decrement in illusion magnitude with free inspection, whereas assimilation illusions do not. A model to explain both the original distortion and the differential susceptibility of the two classes of illusion to decrement is offered.     

Assimilation and contrast to group primes: The moderating role of ingroup identification

This research examined the conditions under which behavioral contrast would be observed in relation to ingroup and outgroup primes. The authors tested the hypothesis that differing levels of commitment to the ingroup would predict diverging behavioral responses to outgroup but not ingroup primes. Across two studies, featuring both age and gender groups, we found that ingroup identification predicted responses to outgroup primes with higher identifiers showing an increased tendency to contrast, that is, behave less like the outgroup, and more like the ingroup. Ingroup identification did not predict responses to ingroup primes. The implications of these findings for social comparison and social identity theories are discussed.     

Assimilation is due to one perceived whole and contrast is due to two perceived wholes

There is evidence that assimilation occurs when stimuli form one perceived whole. The assumption that a single internal event results in both a perceived whole and assimilation follows. There is also evidence that a transition from assimilation to contrast can occur. This assimilation should be due to one internal event, and this contrast is argued to be due to an interaction between the internal events for two perceived wholes. Another proposal is that at any one moment in time all stimuli result in an outcome that has a specific position on a continuum that ranges from one strong perceived whole to two interacting perceived wholes to two independent perceived wholes.     

Assimilation and contrast in geometrical illusions: a theoretical analysis.

It is argued that an unwarranted dichotomization of geometrical illusions has produced theories whose explanatory scope is limited to either assimilation or contrast. Since recent attempts at integration lack precision, a 1977 unified model of the perception of extent by Brigner was revised to predict accurately both assimilative and contrastive phases of the parallel lines illusion. Application of the revised model to other geometrical illusions was discussed and a means of parsimoniously accounting for variations in assimilative illusions occurring with age and with lightness of the stimulus figure was suggested.     

On the consequences of priming: Assimilation and contrast effects

Two experiments investigated the effects of priming (activation of a category by unobtrusive exposure to exemplars of that category) on subsequent judgments in an unrelated task. Subjects were primed with one of four levels of ferocity (size) in the course of a “color perception” experiment, and were later asked to judge the ferocity (size) of real (unambiguous) and unreal (ambiguous) animals. An interaction between ambiguity of judged stimuli (real vs unreal animals) and extremity of primed exemplars (moderate vs extreme levels of ferocity or size) was revealed. Assimilation effects (judgments consistent with the primed category) occurred only when moderate exemplars were primed and ambiguous stimuli judged. Contrast effects occurred when extreme exemplars were primed and ambiguous stimuli judged and, irrespective of extremity of the primed exemplar, when unambiguous stimuli were judged. The results are interpreted in terms of an integration of social judgment and social cognition perspectives.     

Assimilation for affiliation and contrast for control: complementary self-construals

People have knowledge about relationships (i.e., relational schemas) that is based on their experiences. Because most people have experience with complementary behavior (interaction partners behaving similarly in terms of affiliation but oppositely in terms of control), they expect complementary behavior in their relationships. Like other beliefs about relationships, expectations of complementarity affect self-construal. The authors provide evidence for complementary self-construal; people assimilate to relevant relationship partners on the affiliation dimension and contrast on the control dimension. Consistent with the proposed role of relationship knowledge in these effects, complementary self-construal was moderated by the familiarity of the target, whether people focused on their relationship with or the appearance of the target, and whether the context was relevant for the interpersonal dimension.     

Assimilation resistance and affective response in problem solving

In a problem solving experiment, predictions were based upon a theory of motivation and action, the scheme theory, and a psychometric model for the component analysis of rating data, the sector model (Eckblad, 1981; 1980). Pattern guessing tasks were presented to 32 students, 8 to each subject, in individual sessions lasting more than one hour. Number of errors, rated attractiveness, and guessing latencies (the process variables) were recorded during each task. After completion of each task, the affective reaction to it was assessed by means of ratings on six scales, including the scales Pleasant, Interesting, and Complex. On the basis of the sector model, an index of assimilation resistance was computed from the ratings of each task/subject unit, and the index was used to predict mean values and gradients for the process variables. Results show that the sector structure is clearly present in the rating data, that the process variables fit into the structure in the predicted manner, and that mean values and gradients for the process variables differ highly significantly and as predicted among tasks of different levels of assimilation resistance.     

From contrast to reinforcement: role of response contingency in anticipatory contrast

Intake of a 0.15% saccharin solution is suppressed if access to the saccharin is followed by access to 32% sucrose in brief daily pairings. The present series of four experiments was concerned with factors that lead to this anticipatory contrast effect (suppressed saccharin intake) rather than a reinforcement effect. In Experiment 1, anticipatory contrast was obtained with an autoshaping procedure (no lick requirement on the initial tube), and degree of contrast did not vary as a function of intersolution interval in the range of 0-15 s. Experiments 2 and 3 showed that requirements of 10, 100, 200, or 400 licks on the first tube available led to a reinforcement effect in latency, but a requirement of 0 licks (autoshaping procedure) led to a contrast effect in licks and latency. In Experiment 4, a group with a 200-contingent-lick requirement showed a reinforcement effect in latency, but a group yoked to this contingent group showed a contrast effect in both latency and licks. Overall, the results suggest that anticipatory contrast occurs under conditions of a "relaxed" instrumental contingency. The data are discussed in terms of control of behavior by stimulus-stimulus, response-stimulus, and stimulus-response associations, and the results are related to behavioral contrast, to flavor-outcome associations, and to "misbehavior" produced by Pavlovian-instrumental interactions.     

Assimilation and contrast in persuasion: the effects of source credibility in multiple message situations

The present research explores a contextual perspective on persuasion in multiple message situations. It is proposed that when people receive persuasive messages, the effects of those messages are influenced by other messages to which people recently have been exposed. In two experiments, participants received a target persuasive message from a moderately credible source. Immediately before this message, participants received another message, on a different topic, from a source with high or low credibility. In Experiment 1, participants' attitudes toward the target issue were more favorable after they had first been exposed to a different message from a low rather than high credibility source (contrast). In Experiment 2, this effect only emerged when a priming manipulation gave participants a dissimilarity mindset. When participants were primed with a similarity mindset, their attitudes toward the target issue were more favorable following a different message from a high rather than low credibility source (assimilation).     

Sensitivity to attachment, alignment, and contrast polarity variation in local perceptual grouping

A number of leading theories (e.g., Grossberg & Mingolla, 1985; Kellman & Shipley, 1991; Rensink & Enns, 1995) commonly assume that perceptual grouping by contour alignment occurs preattentively across reversing contrast polarity elements. We examined this notion in seven visual search experiments. We found that only grouping by attachment supported preattentive visual search and that grouping by contour alignment required attention in order to operate. Both attachment grouping and grouping by contour alignment were sensitive to contrast reversals. Further results showed that contour alignment was a strong grouping cue only among elements with the same contrast sign but that it did not facilitate grouping across reversing contrast. These results suggest that grouping by contour alignment operates only on inputs of consistent contrast polarity.     

Assimilation and contrast in social comparisons as a consequence of self‐construal activation

We extended existing research about self‐construal activation to the study of social comparisons, specifically to self–other similarity ratings. Independent self‐knowledge substantiates the notion of dissimilarity, whereas interdependent self‐knowledge implies similarity with others. Therefore, perceived self–other similarity was predicted to decrease after independent and increase after interdependent self‐knowledge priming. However, we expected such assimilation effects to occur only if the priming was subtle, but contrast effects if it was overt. In order to test this hypothesis, we developed a scrambled sentences test for priming the respective self‐knowledge. The unscrambled sentences described the self either in terms of independence or interdependence. The subtlety of this priming was manipulated by having participants write down either the full sentences (overt priming) or only the remaining word in each item (subtle priming). Results confirmed the predictions. Underlying cognitive processes of the effects are discussed. Copyright © 2000 John Wiley & Sons, Ltd.     

Assimilation and contrast: the two sides of specific interference between action and perception

Perception and action have long been treated as relatively independent and serial processes. More recent views, however, consider perception and action as relying on a common set of processes and/or representations. The present paper will focus on a variety of specific (content-based) perception–action interactions that have been taken as support for such views. In particular, the following aspects will be considered: direction of influence (perception on action vs. action on perception), temporal type (concurrent vs. non-concurrent), functional relation (related/unrelated), and type of movements (biological vs. non-biological). Different extant models of the perception-action interface are discussed and a classification schema proposed that tries to explain when contrast and when assimilation effects will arise.