Miroslav Sirota

Ecological rationality or nested sets? Individual differences in cognitive processing predict Bayesian reasoning

The presentation of a Bayesian inference problem in terms of natural frequencies rather than probabilities has been shown to enhance performance. The effect of individual differences in cognitive processing on Bayesian reasoning has rarely been studied, despite enabling us to test process-oriented variants of the two main accounts of the facilitative effect of natural frequencies: The ecological rationality account (ERA), which postulates an evolutionarily shaped ease of natural frequency automatic processing, and the nested sets account (NSA), which posits analytical processing of nested sets. In two experiments, we found that cognitive reflection abilities predicted normative performance equally well in tasks featuring whole and arbitrarily parsed objects (Experiment 1) and that cognitive abilities and thinking dispositions (analytical vs. intuitive) predicted performance with single-event probabilities, as well as natural frequencies (Experiment 2). Since these individual differences indicate that analytical processing improves Bayesian reasoning, our findings provide stronger support for the NSA than for the ERA.

Don’t wait to incubate: Immediate versus delayed incubation in divergent thinking

Previous evidence for the effectiveness of immediate incubation in divergent creative tasks has been weak, because earlier studies exhibited a range of methodological problems. This issue is theoretically important, as a demonstration of the effects of immediate incubation would strengthen the case for the involvement of unconscious work in incubation effects. For the present experiment, we used a creative divergent-thinking task (alternative uses) in which separate experimental groups had incubation periods that were either delayed or immediate and that consisted of either spatial or verbal tasks. Control groups were tested without incubation periods, and we carried out checks for intermittent conscious work on the target task during the incubation periods. The results showed significant incubation effects that were stronger for immediate than for delayed incubation. Performance was not different between the verbal and spatial incubation conditions, and we found no evidence for intermittent conscious working during the incubation periods. These results support a role for unconscious work in creative divergent thinking, particularly in the case of immediate incubation.

How to train your Bayesian: a problem-representation transfer rather than a format-representation shift explains training effects.

People improve their Bayesian reasoning most when they are trained to represent single-event probabilities as natural frequencies; nevertheless, the underlying mechanism of this representational training remains unclear. Several authors suggested that people learn to shift the initial format to natural frequencies, and improve their reasoning because natural frequencies align with an evolutionary designed frequency-coding mechanism—the format-representation shift hypothesis. Alternatively, people may acquire a generic problem representation in terms of nested sets that is then transferred to similar problems—the problem-representation transfer hypothesis. To disentangle the effect of the format shift from problem representation transfer, we devised two types of training with problems featuring a nonfrequency format and a concealed nested-sets structure. Participants learnt the adequate problem representation in both training manipulations, but in only one did they learn, in addition, to shift the format to frequencies. Substantial evidence (BF01 = 5, where BF = Bayes factor) indicates that both types of training improved reasoning in an immediate and a one-week follow-up posttest to the same extent. Such findings support the problem-representation transfer hypothesis because learning an adequate nested-sets problem representation accounts for the performance improvement, whereas the frequency format per se confers no additional benefit. We discuss the implications of these findings for two dominant accounts of statistical reasoning.

Now you Bayes, now you don’t: effects of set-problem and frequency-format mental representations on statistical reasoning

People appear to be Bayesian when statistical information is presented in terms of natural frequencies and non-Bayesian when presented in terms of single-event probabilities, unless the probabilities resemble natural frequencies, for example, as chances. The isomorphic format of chances, however, does not always facilitate performance to the extent that the format of natural frequencies does. Prior research has not addressed the underlying mechanism that accounts for this gap despite its theoretical significance. The mechanism explaining this external format gap could lie in the interpretation of the problem as a set-problem, which cues relevant problem model and arithmetic operations (the problem interpretation hypothesis) and/or in the interpretation of the format as frequencies, which may be easier to process (the format interpretation hypothesis). In two parallel experiments, we found support for the problem interpretation hypothesis only: set representations mediated solely the isomorphic format gap (Experiment 1: part A) and accounted for the transfer effect to natural frequencies (Experiment 1: part B); priming set representations improved performance with chances (Experiment 2). We discuss how the supported explanation corroborates the nested-sets rather than the ecological rationality account of statistical reasoning and how it helps explain individual differences in Bayesian reasoning.

Decisive evidence on a smaller-than-you-think phenomenon: revisiting the ''1-in-X'' effect on subjective medical probabilities

Accurate perception of medical probabilities communicated to patients is a cornerstone of informed decision making. People, however, are prone to biases in probability perception. Recently, Pighin and others extended the list of such biases with evidence that “1-in-X” ratios (e.g., “1 in 12”) led to greater perceived probability and worry about health outcomes than “N-in-X*N” ratios (e.g., “10 in 120”). Subsequently, the recommendation was to avoid using “1-in-X” ratios when communicating probabilistic information to patients. To warrant such a recommendation, we conducted 5 well-powered replications and synthesized the available data. We found that 3 out of the 5 replications yielded statistically nonsignificant findings. In addition, our results showed that the “1-in-X” effect was not moderated by numeracy, cognitive reflection, age, or gender. To quantify the evidence for the effect, we conducted a Bayes factor meta-analysis and a traditional meta-analysis of our 5 studies and those of Pighin and others (11 comparisons, N = 1131). The meta-analytical Bayes factor, which allowed assessment of the evidence for the null hypothesis, was very low, providing decisive evidence to support the existence of the “1-in-X” effect. The traditional meta-analysis showed that the overall effect was significant (Hedges’ g = 0.42, 95% CI 0.29–0.54). Overall, we provide decisive evidence for the existence of the “1-in-X” effect but suggest that it is smaller than previously estimated. Theoretical and practical implications are discussed.

Risk communication on shaky ground.

THE SCIENTIFIC COMMUNITY INCREASINGLY RECOGNIZES THE IMPORTANCE OF COMMUNICATING effectively and responsibly with the public (1), but questions remain about whether and how to “frame” scientifi c information about controversial issues such as climate change, evolution, and embryonic stem cell research (2, 3). Recent discussions about the biological determinants of behavior in voles provide an opportunity to refl ect on how scientists can frame information in ways that are both illuminating and responsible. By modulating the density and distribution of vasopressin receptors in specifi c regions of the brain, scientists can get ordinarily “promiscuous” montane voles to behave more like “monogamous” prairie voles (4). By the time this research was reported in the popular media, it had become a story about the discovery of a “gene for” “monogamy,” “fi delity,” “promiscuity,” or “divorce” in humans (5). Consider the major frames we identifi ed in the media coverage of this research: (i) “genetic determinism,” the idea that a single gene controls even complex social behaviors such as sexual monogamy (6); (ii) “triumph for reductionism,” the suggestion that soon we will understand love in terms that refer exclusively to physics and chemistry (7); (iii) “humans are like voles,” a parallel allowing wide-ranging extrapolation (8); (iv) “happiness drug,” the idea that applying lessons learned from this research to biotechnology efforts could save a relationship or marriage (9); and (v) “dangers of social manipulation,” which has led to stories about trust sprays of potential use to the military, department stores, politicians, and stalkers (10). These frames, albeit crude and oversimplifi ed, can help members of the public understand how research relates to broader social trends, issues, and debates. By paying close attention to the dominant frames used in highly publicized cases like this one, scientists can take advantage of these strengths while preemptively highlighting their potential weaknesses. For example, to correct a common source of misunderstanding in the “humans are like voles” frame, experts could emphasize that ordinary usage of terms such as “monogamy” can differ substantially from their technical applications in biology. (Spending 56% of the time with one’s spouse, 19% of the time alone, and 25% of the time copulating with strangers would not qualify as monogamous by ordinary human standards.) They could also suggest alternative frames that counteract weaknesses in existing ones. Instead of focusing on “happiness drugs,” for example, scientists could contextualize this research as a potential treatment for autism, thereby focusing attention on the more realistic relevance of this work. By recognizing the benefi ts of particular frames while preempting and mitigating common misperceptions, scientists can work with the media to develop frames that promote public interest in scientifi c advances without contributing to sensationalism and diminished scientifi c credibility. DANIEL J. MCKAUGHAN* AND KEVIN C. ELLIOTT Department of Philosophy, Boston College, Chestnut Hill, MA 02467, USA. Department of Philosophy, University of South Carolina, Columbia, SC 29208, USA.

Expectations for Antibiotics Increase Their Prescribing: Causal Evidence About Localized Impact.

Objective: Clinically irrelevant but psychologically important factors such as patients’ expectations for antibiotics encourage overprescribing. We aimed to (a) provide missing causal evidence of this effect, (b) identify whether the expectations distort the perceived probability of a bacterial infection either in a pre- or postdecisional distortions pathway, and (c) detect possible moderators of this effect. Method: Family physicians expressed their willingness to prescribe antibiotics (Experiment 1, n1 = 305) or their decision to prescribe (Experiment 2, n2 = 131) and assessed the probability of a bacterial infection in hypothetical patients with infections either with low or high expectations for antibiotics. Response order of prescribing/probability was manipulated in Experiment 1. Results: Overall, the expectations for antibiotics increased intention to prescribe (Experiment 1, F(1, 301) = 25.32, p < .001, &eegr;p2 = .08, regardless of the response order; Experiment 2, odds ratio [OR] = 2.31, and OR = 0.75, Vignettes 1 and 2, respectively). Expectations for antibiotics did not change the perceived probability of a bacterial infection (Experiment 1, F(1, 301) = 1.86, p = .173, &eegr;p2 = .01, regardless of the response order; Experiment 2, d = −0.03, and d = +0.25, Vignettes 1 and 2, respectively). Physicians’ experience was positively associated with prescribing, but it did not moderate the expectations effect on prescribing. Conclusions: Patients’ and their parents’ expectations increase antibiotics prescribing, but their effect is localized—it does not leak into the perceived probability of a bacterial infection. Interventions reducing the overprescribing of antibiotics should target also psychological factors.

The Role of Physicians’ First Impressions in the Diagnosis of Possible Cancers without Alarm Symptoms

Background. First impressions are thought to exert a disproportionate influence on subsequent judgments; however, their role in medical diagnosis has not been systematically studied. We aimed to elicit and measure the association between first impressions and subsequent diagnoses in common presentations with subtle indications of cancer. Methods. Ninety UK family physicians conducted interactive simulated consultations online, while on the phone with a researcher. They saw 6 patient cases, 3 of which could be cancers. Each cancer case included 2 consultations, whereby each patient consulted again with nonimproving and some new symptoms. After reading an introduction (patient description and presenting problem), physicians could request more information, which the researcher displayed online. In 2 of the possible cancers, physicians thought aloud. Two raters coded independently the physicians’ first utterances (after reading the introduction but before requesting more information) as either acknowledging the possibility of cancer or not. We measured the association of these first impressions with the final diagnoses and management decisions. Results. The raters coded 297 verbalizations with high interrater agreement (Kappa = 0.89). When the possibility of cancer was initially verbalized, the odds of subsequently diagnosing it were on average 5 times higher (odds ratio 4.90 [95% CI 2.72 to 8.84], P < 0.001), while the odds of appropriate referral doubled (OR 1.98 [1.10 to 3.57], P = 0.002). The number of cancer-related questions physicians asked mediated the relationship between first impressions and subsequent diagnosis, explaining 29% of the total effect. Conclusion. We measured a strong association between family physicians’ first diagnostic impressions and subsequent diagnoses and decisions. We suggest that interventions to influence and support the diagnostic process should target its early stage of hypothesis generation.

Incubation and suppression processes in creative problem solving

The present study investigated the role of thought suppression in incubation, using a delayed incubation paradigm. A total of 301 participants were tested over five conditions, viz., continuous work control, incubation with a mental rotations interpolated task, focussed suppression, unfocussed suppression and a conscious expression condition. Checks were made for intermittent work during the incubation condition. The target task was alternative uses for a brick. In the incubation and suppression conditions, participants worked for 4 minutes, then had a break during which suppression or the mental rotations interpolated task was carried out for 3 minutes before conscious work was resumed for a further 4 minutes on the alternative uses task. Results indicated that both incubation with an interpolated distractor task and incubation with suppression were effective in enhancing performance relative to controls. The intermittent work hypothesis (that effects of an incubation period are simply due to illicit conscious work on the target task during the incubation period) was not upheld. The effects of incubation/suppression persisted over the post-incubation working period and the results suggested that unfocussed suppression effects on subsequent fluency lasted longer than focussed suppression effects.

Interactivity mitigates the impact of working memory depletion on mental arithmetic performance

Doing long sums in the absence of complementary actions or artefacts is a multistep procedure that quickly taxes working memory; congesting the phonological loop further handicaps performance. In the experiment reported here, participants completed long sums either with hands down – the low interactivity condition – or by moving numbered tokens – the high interactivity condition – while they repeated “the” continuously, loading the phonological loop, or not. As expected, interactivity and articulatory suppression substantially affected performance; critically, the effect of articulatory suppression was stronger in the low than in the high interactivity condition. In addition, an independent measure of mathematics anxiety predicted the impact of articulatory suppression on performance only in the low (not high) interactivity condition. These findings suggest that interactivity augmented overall or systemic working memory resources and diminished the effect of mathematics anxiety, underscoring the importance of characterizing the properties of the system as it is configured by the dynamic agent-environment coupling