Pseudocontingencies: an integrative account of an intriguing cognitive illusion

Seeing is misbelieving: Consumers wrongly believe that unhealthy food tastes better when there is more of it

Recent studies have shown that people can believe that unhealthy foods taste better, even if healthy and unhealthy foods are equally as tasty. Specifically, when tasty and unhealthy foods are frequent in one context but rare in another, people perceive unhealthy foods to taste better, even if health and taste are unrelated. Given that people often consume food in one context, the current study investigated whether false beliefs about the health-taste relationship in foods can also occur in just one single context, in which either healthy or unhealthy foods are predominant, when there is no contrasting context where the respective other food is predominant. In two experiments (N = 342), we presented participants with pictures of meals from a single context and varied the frequency of healthy and unhealthy foods between participants. Although healthy and unhealthy foods tasted equally as good, participants believed that (un)healthy foods tasted better when there were more of them. This research demonstrates that health-taste beliefs might be changed by increasing the relative frequency of healthy foods in the environment overall, not by just offering some healthy and tasty foods.

Contingency inferences from base rates: A parsimonious strategy?

The pseudocontingency framework provides a parsimonious strategy for inferring the contingency between two variables by assessing the base rates. Frequently occurring levels are associated, as are rarely occurring levels. However, this strategy can lead to different contingency inferences in different contexts, depending on how the base rates vary across contexts. Here, we examine how base-rate consistency influences base-rate learning and reliance by contrasting consistent with inconsistent base rates. We hypothesized that base-rate learning is facilitated, and that people rely more on base rates if base rates are consistent. In Experiment 1, the base rates across four contexts implied the same (consistent) or different (inconsistent) contingencies. Base rates were learned equally accurately, and participants inferred contingencies that followed the base rates but deviated from the genuine contingencies within contexts, regardless of consistency. In Experiment 2, we additionally manipulated whether the context was a plausible moderator of the contingency. While we replicated the first experiment's results when the context was a plausible moderator, base-rate inferences were stronger for consistent base rates when the context was an implausible moderator. Possibly, when a moderation-by-context was implausible, participants also relied on the base-rate correlation across contexts, which implied the same contingency when base rates were consistent but was zero when the base rates were inconsistent. Thus, our findings suggest that contingency inferences from base rates involve top-down processes in which people decide how to use base-rate information.

Self-truncated sampling produces more moderate covariation judgment and related decision than descriptive frequency information: The role of regressive frequency estimation

Covariation judgment underlies a diversity of psychological theories and influences various everyday decisions. Information about covariation can be learned from either a summary description of frequencies (i.e., contingency tables) or trial-by-trial experience (i.e., sampling individual instances). Two studies were conducted to investigate the impact of information learning mode (i.e., description vs. self-truncated sampling) on covariation judgment and related decision. In each trial under the description condition, participants were presented with a contingency table with explicit cell frequencies, whereas in each trial under the self-truncated sampling condition, participants were allowed to determine when to stop sampling instances and thus the actual sample size. To eliminate sampling error, an other-yoked (i.e., between-subject) design was used in this research so that cell frequencies shown in a trial under the description condition were matched with those experienced in a trial under the self-truncated sampling condition. Experiment 1 showed that the self-truncated sampling condition led to more moderate covariation judgments than the description condition (i.e., a description-experience gap). Experiment 2 demonstrated further that the same gap extended to covariation-related decisions in terms of relative contingent preference (RCP). Regressive frequency estimation under self-truncated sampling appeared to underlie the consistent gaps found in the two studies, whereas the impact of regressive diagnosticity (i.e., the same sample of instances was viewed as less diagnostic under description than under self-truncated sampling) or simultaneous overestimation and underweighting of rare instances under experience was not supported by the observed data. Future research might examine alternative accounts of the observed gaps, such as the impacts of belief updating and stopping rule under self-truncated sampling.

Toward Parsimony in Bias Research: A Proposed Common Framework of Belief-Consistent Information Processing for a Set of Biases

One of the essential insights from psychological research is that people's information processing is often biased. By now, a number of different biases have been identified and empirically demonstrated. Unfortunately, however, these biases have often been examined in separate lines of research, thereby precluding the recognition of shared principles. Here we argue that several-so far mostly unrelated-biases (e.g., bias blind spot, hostile media bias, egocentric/ethnocentric bias, outcome bias) can be traced back to the combination of a fundamental prior belief and humans' tendency toward belief-consistent information processing. What varies between different biases is essentially the specific belief that guides information processing. More importantly, we propose that different biases even share the same underlying belief and differ only in the specific outcome of information processing that is assessed (i.e., the dependent variable), thus tapping into different manifestations of the same latent information processing. In other words, we propose for discussion a model that suffices to explain several different biases. We thereby suggest a more parsimonious approach compared with current theoretical explanations of these biases. We also generate novel hypotheses that follow directly from the integrative nature of our perspective.

Structure learning principles of stereotype change

Why, when, and how do stereotypes change? This paper develops a computational account based on the principles of structure learning: stereotypes are governed by probabilistic beliefs about the assignment of individuals to groups. Two aspects of this account are particularly important. First, groups are flexibly constructed based on the distribution of traits across individuals; groups are not fixed, nor are they assumed to map on to categories we have to provide to the model. This allows the model to explain the phenomena of group discovery and subtyping, whereby deviant individuals are segregated from a group, thus protecting the group's stereotype. Second, groups are hierarchically structured, such that groups can be nested. This allows the model to explain the phenomenon of subgrouping, whereby a collection of deviant individuals is organized into a refinement of the superordinate group. The structure learning account also sheds light on several factors that determine stereotype change, including perceived group variability, individual typicality, cognitive load, and sample size.

On the role of exploitation and exploration strategies in the maintenance of cognitive biases: Beyond the pursuit of instrumental rewards

Why can initial biases persist in repeated choice tasks? Previous research has shown that frequent rewards can lure the decision maker into premature exploitation of a supposedly best option, which can result in the persistence of initial biases. Here, we demonstrate that even in the absence of rewards, initial biases can be perpetuated through a positive testing strategy. After eliciting a biased preference for one of two equally rewarding options, participants (N = 203) could sample freely from both options without the lure of any financial rewards. When participants were told to rule out alternatives in this phase, they explored the supposedly worse option and thereby managed to overcome their initial bias. When told to optimize their strategy, however, they exhibited a positive testing strategy resulting in the continued exploitation of the supposedly better option, a bias they maintained in an incentivized choice phase and later judgments. Across all participants, individual tendencies to exploit one option in earlier phases predicted biased behavior in subsequent phases. The findings highlight that not only the pursuit of instrumental rewards can lead to exploitation and the maintenance of initial biases. We discuss potential consequences for interventions.

Pseudocontingencies: Flexible contingency inferences from base rates

Humans are evidently able to learn contingencies from the co-occurrence of cues and outcomes. But how do humans judge contingencies when observations of cue and outcome are learned on different occasions? The pseudocontingency framework proposes that humans rely on base-rate correlations across contexts, that is, whether outcome base rates increase or decrease with cue base rates. Here, we elaborate on an alternative mechanism for pseudocontingencies that exploits base rate information within contexts. In two experiments, cue and outcome base rates varied across four contexts, but the correlation by base rates was kept constant at zero. In some contexts, cue and outcome base rates were aligned (e.g., cue and outcome base rates were both high). In other contexts, cue and outcome base rates were misaligned (e.g., cue base rate was high, but outcome base rate was low). Judged contingencies were more positive for contexts in which cue and outcome base rates were aligned than in contexts in which cue and outcome base rates were misaligned. Our findings indicate that people use the alignment of base rates to infer contingencies conditional on the context. As such, they lend support to the pseudocontingency framework, which predicts that decision makers rely on base rates to approximate contingencies. However, they challenge previous conceptions of pseudocontingencies as a uniform inference from correlated base rates. Instead, they suggest that people possess a repertoire of multiple contingency inferences that differ with regard to informational requirements and areas of applicability.

Fluent processing leads to positive stimulus evaluations even when base rates suggest negative evaluations

Fluency is the experienced ease of ongoing mental operations, which increases the subjective positivity of stimuli attributes. This may happen because fluency is inherently positive. Alternatively, people may learn the meaning of fluency from contingencies within judgment-contexts. We test pseudocontingencies (PCs) as a mechanism through which fluency's meaning is learned. PCs are inferred correlations between two attributes due to the observation of their jointly skewed base rates - people relate what is frequent in one attribute to what is frequent in the other. Using online seller evaluations as the dependent variable, we manipulated base rates of seller name-fluency and seller reputation, creating conditions where name-fluency aligned positively or negatively with reputation. However, participants evaluated high-fluency name sellers more positively across base-rate conditions, although we observed negative PCs between seller reputation and a fluency-neutral dimension in a follow-up study. We discuss the implications for the debate regarding fluency's positive vs. malleable nature.

Anonymized location data reveals trends in legal Cannabis use in communities with increased mental health risks at the start of the COVID-19 pandemic

BACKGROUND

The coronavirus disease 2019 (COVID-19) pandemic has led to increases in felt negative affect for many. This is concerning as individuals at increased risk for mental health issues are often more likely to use substances to cope with stressors.

OBJECTIVES

The aim of the current study is to examine whether communities reporting an increased risk for developing mental health issues showed differential patterns of legal cannabis use as the pandemic began. A secondary goal is to examine the feasibility of using anonymized location data to uncover community consumption patterns of potential concern.

METHODS

Anonymized location data from approximately 10% of devices in the United States provided a count of the number of visitors to 3,335 cannabis retail locations (medical and recreational) each day from December 1^st 2019 through April 2020. Visitor counts were merged with the average number of mentally unhealthy days (aMUDs) reported in the Federal Information Processing Standard (FIPS) county the retailer was located along with FIPS county population and poverty rate estimates. A Poisson spline regression predicting visitors by day, aMUDs, as well as their interaction was performed, entering population and poverty rate as covariates.

RESULTS

As the pandemic began communities reporting a greater aMUDs showed greater visitation to cannabis retailers.

CONCLUSIONS

These results suggest that the COVID-19 pandemic may have led to increased legal cannabis use in at risk communities. They also highlight the value anonymized location data can provide policymakers and practitioners in uncovering community level trends as they confront an increasingly uncertain landscape.

Causal illusions in the classroom: how the distribution of student outcomes can promote false instructional beliefs

Teachers sometimes believe in the efficacy of instructional practices that have little empirical support. These beliefs have proven difficult to efface despite strong challenges to their evidentiary basis. Teachers typically develop causal beliefs about the efficacy of instructional practices by inferring their effect on students’ academic performance. Here, we evaluate whether causal inferences about instructional practices are susceptible to an outcome density effect using a contingency learning task. In a series of six experiments, participants were ostensibly presented with students’ assessment outcomes, some of whom had supposedly received teaching via a novel technique and some of whom supposedly received ordinary instruction. The distributions of the assessment outcomes was manipulated to either have frequent positive outcomes (high outcome density condition) or infrequent positive outcomes (low outcome density condition). For both continuous and categorical assessment outcomes, participants in the high outcome density condition rated the novel instructional technique as effective, despite the fact that it either had no effect or had a negative effect on outcomes, while the participants in the low outcome density condition did not. These results suggest that when base rates of performance are high, participants may be particularly susceptible to drawing inaccurate inferences about the efficacy of instructional practices.

When Success Is Not Enough: The Symptom Base-Rate Can Influence Judgments of Effectiveness of a Successful Treatment

Patients’ beliefs about the effectiveness of their treatments are key to the success of any intervention. However, since these beliefs are usually formed by sequentially accumulating evidence in the form of the covariation between the treatment use and the symptoms, it is not always easy to detect when a treatment is actually working. In Experiments 1 and 2, we presented participants with a contingency learning task in which a fictitious treatment was actually effective to reduce the symptoms of fictitious patients. However, the base-rate of the symptoms was manipulated so that, for half of participants, the symptoms were very frequent before the treatment, whereas for the rest of participants, the symptoms were less frequently observed. Although the treatment was equally effective in all cases according to the objective contingency between the treatment and healings, the participants’ beliefs on the effectiveness of the treatment were influenced by the base-rate of the symptoms, so that those who observed frequent symptoms before the treatment tended to produce lower judgments of effectiveness. Experiment 3 showed that participants were probably basing their judgments on an estimate of effectiveness relative to the symptom base-rate, rather than on contingency in absolute terms. Data, materials, and R scripts to reproduce the figures are publicly available at the Open Science Framework: https://osf.io/emzbj/.

The Challenge of Diagnostic Inferences From Implicit Measures: The Case of Non-Evaluative Influences in the Evaluative Priming Paradigm

Implicit measures are diagnostic tools to assess attitudes and evaluations that people cannot or may not want to report. Diagnostic inferences from such tools are subject to asymmetries. We argue that (causal) conditional probabilities p(AM+|A+) of implicitly measured attitudes AM+ given the causal influence of existing attitudes A+ is typically higher than the reverse (diagnostic) conditional probability p(A+|AM+), due to non-evaluative influences on implicit measures. We substantiate this argument with evidence for non-evaluative influences on evaluative priming—specifically, similarity effects reflecting the higher similarity of positive than negative prime-target pairs; integrativity effects based on primes and targets’ potential to form meaningful semantic compounds; and congruity proportion effects that originate in individuals’ decisional strategies. We also cursorily discuss non-evaluative influences in the Implicit Association Test (IAT). These influences not only have implications for the evaluative priming paradigm in particular, but also highlight the intricacies of diagnostic inferences from implicit measures in general.

The Origin of Illusory Correlations

If the distributions of two binary variables are skewed, people erroneously perceive a correlation even if the variables are actually uncorrelated. Specifically, people perceive a correlation between the variables' infrequent (vs. frequent) levels. While such systematic Illusory Correlations (ICs) can account for important phenomena, including erroneous stereotypes linking minority groups with infrequent attributes, the theoretical explanation is still a matter of debate. As proposed in the distinctiveness-based account, ICs arise due to a memory advantage for infrequent events. Alternatively, it has been proposed that ICs reflect ad hoc inferences from univariate base rates. In two experiments, we tested the accounts' predictions using a standard IC paradigm followed by (1) a contingency judgment task, asking participants to directly report contingencies, and (2) a trial-by-trial source assignment task of group membership and attribute. Bayesian-hierarchical model analysis revealed that ICs in contingency judgments were positively correlated to ad hoc inferences, but uncorrelated to exemplar memory in participants' source assignments. Moreover, inducing a focus on infrequent events improved memory for distinct events, but did not result in stronger ICs. Thus, our results support the explanation of ICs as ad hoc inferences from base rates, rather than a memory advantage for distinctive item combinations.

Illusory correlations despite equated category frequencies: A test of the information loss account

Illusory correlations (IC) are the perception of covariation, where none exists. For example, people associate majorities with frequent behavior and minorities with infrequent behavior even in the absence of such an association. According to the information loss account, ICs result from greater fading of infrequent group-behavior combinations in memory. We conducted computer simulations based on this account which showed that ICs are expected under standard conditions with skewed category frequencies (i.e. 2:1 ratio for positive and negative descriptions), but not under conditions with equated category frequencies (i.e. 1:1 ratio for positive and negative descriptions). Contrary to these simulations, our behavioral experiments revealed an IC under both conditions, which did not decrease over time. Thus, information loss alone is not sufficient as an explanation for the formation of ICs. These results imply that negative items contribute to ICs not only due to their infrequency, but also due to their emotional salience.

Statistical numeracy as a moderator of (pseudo)contingency effects on decision behavior

Pseudocontingencies denote contingency estimates inferred from base rates rather than from cell frequencies. We examined the role of statistical numeracy for effects of such fallible but adaptive inferences on choice behavior. In Experiment 1, we provided information on single observations as well as on base rates and tracked participants' eye movements. In Experiment 2, we manipulated the availability of information on cell frequencies and base rates between conditions. Our results demonstrate that a focus on base rates rather than cell frequencies benefits pseudocontingency effects. Learners who are more proficient in (conditional) probability calculation prefer to rely on cell frequencies in order to judge contingencies, though, as was evident from their gaze behavior. If cell frequencies are available in summarized format, they may infer the true contingency between options and outcomes. Otherwise, however, even highly numerate learners are susceptible to pseudocontingency effects.

What Constitutes Strong Psychological Science? The (Neglected) Role of Diagnosticity and A Priori Theorizing

A Bayesian perspective on Ioannidis’s (2005) memorable statement that “Most Published Research Findings Are False” suggests a seemingly inescapable trade-off: It appears as if research hypotheses are based either on safe ground (high prior odds), yielding valid but unsurprising results, or on unexpected and novel ideas (low prior odds), inspiring risky and surprising findings that are inevitably often wrong. Indeed, research of two prominent types, sexy hypothesis testing and model testing, is often characterized by low priors (due to astounding hypotheses and conjunctive models) as well as low-likelihood ratios (due to nondiagnostic predictions of the yin-or-yang type). However, the trade-off is not inescapable: An alternative research approach, theory-driven cumulative science, aims at maximizing both prior odds and diagnostic hypothesis testing. The final discussion emphasizes the value of pluralistic science, within which exploratory phenomenon-driven research can play a similarly strong part as strict theory-testing science.

Anomalies in the detection of change: When changes in sample size are mistaken for changes in proportions

Detecting changes, in performance, sales, markets, risks, social relations, or public opinions, constitutes an important adaptive function. In a sequential paradigm devised to investigate detection of change, every trial provides a sample of binary outcomes (e.g., correct vs. incorrect student responses). Participants have to decide whether the proportion of a focal feature (e.g., correct responses) in the population from which the sample is drawn has decreased, remained constant, or increased. Strong and persistent anomalies in change detection arise when changes in proportional quantities vary orthogonally to changes in absolute sample size. Proportional increases are readily detected and nonchanges are erroneously perceived as increases when absolute sample size increases. Conversely, decreasing sample size facilitates the correct detection of proportional decreases and the erroneous perception of nonchanges as decreases. These anomalies are however confined to experienced samples of elementary raw events from which proportions have to be inferred inductively. They disappear when sample proportions are described as percentages in a normalized probability format. To explain these challenging findings, it is essential to understand the inductive-learning constraints imposed on decisions from experience.

Understanding antigay bias from a cognitive-affective-behavioral perspective

In general, United States citizens have become increasingly more accepting of lesbians and gay men over the past few decades. Despite this shift in public attitudes, antigay bias remains openly tolerated, accepted, practiced, and even defended by a substantial portion of the population. This article reviews why and how antigay bias persists using a cognitive-affective-behavioral perspective that touches on sociocognitive factors such as prejudice and stereotyping, as well as features unique to antigay bias, such as its concealable nature. The article concludes with a discussion of how understanding modern antigay bias through a cognitive-affective-behavioral lens can be applied to reduce discrimination against gays and lesbians.

Inferring correlations: from exemplars to categories

Research and theorizing suggest a processing advantage of category-level correlations over exemplar-level correlations. That research has also shown that category-level correlations serve as a proxy for inferring exemplar-level correlations. For example, an individual may learn that the demand for a product category, like cheese, in one store predicts the demand for this category in another. The individual could then draw the unwarranted conclusion that the demand for an exemplar, like cheddar, would also predict the demand for this exemplar in the other store. This notion is supported by previous experiments demonstrating that the subjective exemplar-level correlation follows the implication of the category-level correlation. However, in virtually all previous experiments suggesting a processing advantage for category-level over exemplar-level correlations, the stimulus correlation at the category level was substantial, whereas the correlation at the exemplar level was weak. Here, we tested the hypothesis that individuals process the level that is most informative, either the exemplar or the category level. We presented participants with a zero correlation at the category level, but varied the correlation at the exemplar level. Participants presented with a zero correlation across exemplar products correctly reproduced a zero correlation across product categories. When presented with a substantial correlation at the exemplar level, however, they erroneously reproduced a similar correlation at the category level. These findings therefore imply that there is no general processing advantage for correlations at higher aggregation levels. Instead, individuals seemingly attend to the level that holds the most regular information. Findings are discussed regarding the role of covariation strength in correlation detection and use.

Simpson's paradox in psychological science: a practical guide

The direction of an association at the population-level may be reversed within the subgroups comprising that population—a striking observation called Simpson's paradox. When facing this pattern, psychologists often view it as anomalous. Here, we argue that Simpson's paradox is more common than conventionally thought, and typically results in incorrect interpretations—potentially with harmful consequences. We support this claim by reviewing results from cognitive neuroscience, behavior genetics, clinical psychology, personality psychology, educational psychology, intelligence research, and simulation studies. We show that Simpson's paradox is most likely to occur when inferences are drawn across different levels of explanation (e.g., from populations to subgroups, or subgroups to individuals). We propose a set of statistical markers indicative of the paradox, and offer psychometric solutions for dealing with the paradox when encountered—including a toolbox in R for detecting Simpson's paradox. We show that explicit modeling of situations in which the paradox might occur not only prevents incorrect interpretations of data, but also results in a deeper understanding of what data tell us about the world.

Pseudocontingencies

The ability to extract statistical contingencies (e.g., between cause and effect, between response and feedback) is commonly presupposed as a basic module of adaptive behavior. In reality, however, stimulus input rarely contains the complete sets of correlated attributes required to assess the actual contingencies. Instead, cognitive inferences often rely on a base-rate-driven pseudocontingency rule, which links the more (or less) frequent level of one variable to the more (or less) frequent level of the other variable. Empirical evidence shows that logically unwarranted pseudocontingency inferences override genuine contingencies across many research paradigms. Although pseudocontingencies can be severely misleading, they also provide a useful proxy that accurately predicts existing contingencies most of the time.

Pseudocontingencies derived from categorically organized memory representations

Pseudocontingencies (PCs) allow for inferences about the contingency between two variables X and Y when the conditions for genuine contingency assessment are not met. Even when joint observations X i and Y i about the same reference objects i are not available or are detached in time or space, the correlation r(X i ,Y i ) is readily inferred from base rates. Inferred correlations are positive (negative) if X and Y base rates are skewed in the same (different) directions. Such PC inferences afford useful proxies for actually existing contingencies. While previous studies have focused on PCs due to environmental base rates, the present research highlights memory organization as a natural source of PC effects. When information about two attributes X and Y is represented in a hierarchically organized categorical memory code, as category-wise base rates p(X) and p(Y), the reconstruction of item-level information from category base rates will naturally produce PC effects. Three experiments support this contention. When the yes base rates of two respondents in four questionnaire subscales (categories) were correlated, recalled and predicted item-level responses were correlated in the same direction, even when the original responses to specific items within categories were correlated in the opposite direction.

Does personal intelligence exist? Evidence from a new ability-based measure

Personal intelligence has been defined as the ability to reason about personality and personality-relevant information and to use that information to guide one's actions and more generally, one's life. We constructed an initial version of an ability-based measure to test whether personal intelligence can be measured and whether it exists as a unitary intelligence. In 3 studies (N = 241, 308, and 385), we administered this Test of Personal Intelligence (TOPI), composed of 4 sections, to undergraduates along with criterion measures. Results suggested that a personal intelligence can be measured, that it might exist as a unified area of mental abilities, and that it represents psychological qualities that have intriguing predictive aspects.

The empirical content of theories in judgment and decision making: Shortcomings and remedies

According to Karl Popper, we can tell good theories from poor ones by assessing their empirical content (empirischer Gehalt), which basically reflects how much information they convey concerning the world. “The empirical content of a statement increases with its degree of falsifiability: the more a statement forbids, the more it says about the world of experience.” Two criteria to evaluate the empirical content of a theory are their level of universality (Allgemeinheit) and their degree of precision (Bestimmtheit). The former specifies how many situations it can be applied to. The latter refers to the specificity in prediction, that is, how many subclasses of realizations it allows. We conduct an analysis of the empirical content of theories in Judgment and Decision Making (JDM) and identify the challenges in theory formulation for different classes of models. Elaborating on classic Popperian ideas, we suggest some guidelines for publication of theoretical work.

Contingency inferences driven by base rates: Valid by sampling

Fiedler et al. (2009), reviewed evidence for the utilization of a contingency inference strategy termed pseudocontingencies (PCs). In PCs, the more frequent levels (and, by implication, the less frequent levels) are assumed to be associated. PCs have been obtained using a wide range of task settings and dependent measures. Yet, the readiness with which decision makers rely on PCs is poorly understood. A computer simulation explored two potential sources of subjective validity of PCs. First, PCs are shown to perform above chance level when the task is to infer the sign of moderate to strong population contingencies from a sample of observations. Second, contingency inferences based on PCs and inferences based on cell frequencies are shown to partially agree across samples. Intriguingly, this criterion and convergent validity are by-products of random sampling error, highlighting the inductive nature of contingency inferences.

Much Pain, Little Gain? Paradigm-Specific Models and Methods in Experimental Psychology

Paradigm-oriented research strategies in experimental psychology have strengths and limitations. On the one hand, experimental paradigms play a crucial epistemic and heuristic role in basic psychological research. On the other hand, empirical research is often limited to the observed effects in a certain paradigm, and theoretical models are frequently tied to the particular features of the given paradigm. A paradigm-driven research strategy therefore jeopardizes the pursuit of research questions and theoretical models that go beyond a specific paradigm. As one example of a more integrative approach, recent research on illusory and spurious correlations has attempted to overcome the limitations of paradigm-specific models in the context of biased contingency perception and social stereotyping. Last but not least, the use of statistical models for the analysis of elementary cognitive functions is a means toward a more integrative terminology and theoretical perspective across different experimental paradigms and research domains.

Making the illusory correlation effect appear and then disappear: The effects of increased learning

The acquisition of a negative evaluation of a fictitious minority social group in spite of the absence of any objective correlation between group membership and negative behaviours was described by Hamilton and Gifford (1976) as an instance of an illusory correlation. We studied the acquisition and attenuation through time of this correlation learning effect. In two experiments we asked for participants’ judgements of two fictitious groups using an online version of a group membership belief paradigm. We tested how judgements of the two groups changed as a function of the amount of training they received. Results suggest that the perception of the illusory correlation effect is initially absent, emerges with intermediate amounts of absolute experience, but diminishes and is eliminated with increased experience. This illusory correlation effect can be considered to reflect incomplete learning rather than a bias due to information loss in judgements or distinctiveness.

Pseudocontingencies can override genuine contingencies between multiple cues

The cognitive process of contingency assessment has traditionally been conceived as an inference from the joint frequencies (table cell entries) of correlated cues. Alternatively, pseudocontingency (PC) inferences are derived from the alignment of separate base rate trends (table marginals). The present research extends PCs to the simultaneous assessment of six contingencies between four personality cues. Consistently obtained PC effects, reflecting accurately assessed base rate trends, were unaffected by the actual cue intercorrelations, which were zero or consistent or inconsistent with the PCs. The functional value of PC inferences is discussed. Although PCs can be misleading, they afford approximations of actually existing correlations that suffice under many, although not under all, conditions.