Rapid word learning under uncertainty via cross-situational statistics

There are an infinite number of possible word-to-word pairings in naturalistic learning environments. Previous proposals to solve this mapping problem have focused on linguistic, social, representational, and attentional constraints at a single moment. This article discusses a cross-situational learning strategy based on computing distributional statistics across words, across referents, and, most important, across the co-occurrences of words and referents at multiple moments. We briefly exposed adults to a set of trials that each contained multiple spoken words and multiple pictures of individual objects; no information about word-picture correspondences was given within a trial. Nonetheless, over trials, subjects learned the word-picture mappings through cross-trial statistical relations. Different learning conditions varied the degree of within-trial reference uncertainty, the number of trials, and the length of trials. Overall, the remarkable performance of learners in various learning conditions suggests that they calculate cross-trial statistics with sufficient fidelity and by doing so rapidly learn word-referent pairs even in highly ambiguous learning contexts.     

Between ignorance and truth: Partition dependence and learning in judgment under uncertainty

In 3 studies, participants viewed sequences of multiattribute objects (e.g., colored shapes) appearing with varying frequencies and judged the likelihood of the attributes of those objects. Judged probabilities reflected a compromise between (a) the frequency with which each attribute appeared and (b) the ignorance prior probability cued by the number of distinct values that the focal attribute could take on. Thus, judged probabilities were partition dependent, varying with the number of events into which the state space was subjectively divided. This bias was diminished among participants more confident in what they learned, was strong and insensitive to level of confidence when ignorance priors were especially salient, and required ignorance priors to be salient only when probabilities were elicited (not during encoding).     

Performance bias: Why judgments of learning are not affected by learning

Past research has shown a performance bias: People expect their future performance level on a task to match their current performance level, even when there are good reasons to expect future performance to differ from current performance. One explanation of this bias is that judgments are controlled by what learners can observe, and while current performance is usually observable, changes in performance (i.e., learning or forgetting) are not. This explanation makes a prediction that we tested here: If learning becomes observable, it should begin to affect judgments. In three experiments, after practicing a skill, participants estimated how they performed in the past and how they expected to perform in the future. In Experiments 1 and 2, participants knew they had been improving, as shown by their responses, yet they did not predict that they would improve in the future. This finding was particularly striking because (a) they did improve in the future and (b) as Experiment 3 showed, they did hold the conscious belief that past improvement predicted future improvement. In short, when learning and performance are both observable, judgments of learning seem to be guided by performance and not learning.     

Attention and expectation in human predictive learning: The role of uncertainty

Three localized, visual pattern stimuli were trained as predictive signals of auditory outcomes. One signal partially predicted an aversive noise in Experiment 1 and a neutral tone in Experiment 2, whereas the other signals consistently predicted either the occurrence or absence of the noise. The expectation of the noise was measured during each signal presentation, and only participants for whom this expectation demonstrated contingency knowledge showed differential attention to the signals. Importantly, when attention was measured by visual fixations, the contingency-aware group attended more to the partially predictive signal than to the consistent predictors in both experiments. This profile of visual attention supports the Pearce and Hall (1980) Pearce, J. M. and Hall, G. 1980. A model for Pavlovian learning: Variations in the effectiveness of conditioned but not unconditioned stimuli. Psychological Review, 87: 532–552. [Crossref], [PubMed], [Web of Science ®] , [Google Scholar] theory of the role of attention in associative learning.     

Pragmatic approach to decision making under uncertainty: The case of the disjunction effect

The disjunction effect (Tversky & Shafir, 1992 Tversky, A. and Shafir, E. 1992. The disjunction effect in choice under uncertainty. Psychological Science, 3: 305–309. [Crossref], [Web of Science ®] , [Google Scholar]) occurs when decision makers prefer option x (versus y) when knowing that event A occurs and also when knowing that event A does not occur, but they refuse x (or prefer y) when not knowing whether or not A occurs. This form of incoherence violates Savage's (1954 Savage, L. J. 1954. The foundations of statistics, New York: Wiley.  [Google Scholar]) sure-thing principle, one of the basic axioms of the rational theory of decision making. The phenomenon was attributed to a lack of clear reasons for accepting an option (x) when subjects are under uncertainty. Through a pragmatic analysis of the task and a consequent reformulation of it, we show that the effect does not depend on the presence of uncertainty, but on the introduction of non-relevant goals into the text problem, in both the well-known Gamble problem and the Hawaii problem.     

The absence of a shape bias in children's word learning

There is debate about whether preschool-age children interpret words as referring to kinds or to classes defined by shape similarity. The authors argue that the shape bias reported in previous studies is a task-induced artifact rather than a genuine word-learning strategy. In particular, children were forced to extend an object's novel label to one of several stand-alone, simple-shaped items, including a same-shape option from a different category and a different-shape option from the same superordinate category. Across 6 experiments, the authors found that the shape bias was eliminated (a) when the objects were more complex, (b) when they were presented in context, or (c) when children were no longer forced to choose. Moreover, children preferred the different-shape category alternatives when these were part of the same basic-level category as the target. The present experiments suggest that children seek out objects of the same kind when presented with a novel label, even if they are sometimes unable to identify the relevant kinds on their own.     

A bias to detail: how hand position modulates visual learning and visual memory

In this report, we examine whether and how altered aspects of perception and attention near the hands affect one’s learning of to-be-remembered visual material. We employed the contextual cuing paradigm of visual learning in two experiments. Participants searched for a target embedded within images of fractals and other complex geometrical patterns while either holding their hands near to or far from the stimuli. When visual features and structural patterns remained constant across to-be-learned images (Exp. 1), no difference emerged between hand postures in the observed rates of learning. However, when to-be-learned scenes maintained structural pattern information but changed in color (Exp. 2), participants exhibited substantially slower rates of learning when holding their hands near the material. This finding shows that learning near the hands is impaired in situations in which common information must be abstracted from visually unique images, suggesting a bias toward detail-oriented processing near the hands.     

Decision-making under uncertainty: biases and Bayesians

Animals (including humans) often face circumstances in which the best choice of action is not certain. Environmental cues may be ambiguous, and choices may be risky. This paper reviews the theoretical side of decision-making under uncertainty, particularly with regard to unknown risk (ambiguity). We use simple models to show that, irrespective of pay-offs, whether it is optimal to bias probability estimates depends upon how those estimates have been generated. In particular, if estimates have been calculated in a Bayesian framework with a sensible prior, it is best to use unbiased estimates. We review the extent of evidence for and against viewing animals (including humans) as Bayesian decision-makers. We pay particular attention to the Ellsberg Paradox, a classic result from experimental economics, in which human subjects appear to deviate from optimal decision-making by demonstrating an apparent aversion to ambiguity in a choice between two options with equal expected rewards. The paradox initially seems to be an example where decision-making estimates are biased relative to the Bayesian optimum. We discuss the extent to which the Bayesian paradigm might be applied to the evolution of decision-makers and how the Ellsberg Paradox may, with a deeper understanding, be resolved.     

Psychophysics of remembering: to bias or not to bias

Delayed matching to sample is typically a two-alternative forced-choice procedure with two sample stimuli. In this task the effects of varying the probability of reinforcers for correct choices and the resulting receiver operating characteristic are symmetrical. A version of the task where a sample is present on some trials and absent on others is analogous to a yes/no recognition task. We describe data from two experiments where an asymmetry in performance in the yes/no task could be attributed to a change in response bias with increasing retention-interval duration from a matching-law perspective, but not from a signal-detection perspective. Both approaches make explicit assumptions about response bias. The apparent inconsistency between the two approaches to the treatment of response bias is resolved in terms of a model proposed by K. G. White and J. T. Wixted (1999) which predicts asymmetrical matching-law functions and receiver operating characteristics without making any assumptions about response bias.     

Becoming us and them: Social learning and intergroup bias

In recent years, research has demonstrated that the basic features of prejudice and discrimination emerge early in children's development. These discoveries call into question the role of social learning in intergroup bias. Specifically, through what means do we learn to distinguish “us” from “them?” Here, we explore this question, focusing on three key issues: how children respond to biased information they receive from others, how children selectively seek out certain types of biased information, and how children communicate biased information to others. We close by discussing the implications of this research for interventions to reduce stereotyping, prejudice, and discrimination.     

Magnitude comparison revisited: An alternative approach to binary choice under uncertainty

Two generations of psychologists have been interested in understanding binary choice under uncertainty. In the 1970s and 1980s, researchers assumed that people rely on a two-stage magnitude comparison process to make these decisions (Banks, 1977; Moyer & Dumais, 1978). More recently, the focus has shifted to approaches that rely on probabilistic cues and simple heuristics (Gigerenzer & Goldstein, Psychological Review 103, 650-669, 1996). Here, we test competing predictions derived from these two very different approaches and conclude that the magnitude comparison process plays a central role in this task. In support of this conclusion, we present an experiment in which participants were timed as they decided which of two vehicles was more expensive. Pairs composed of one luxury vehicle (e.g., BMW 323i) and one nonluxury vehicle (e.g., Toyota 4Runner) were critical because the magnitude comparison approach correctly predicted that reaction times would decrease with subjective distance, whereas the heuristics approach incorrectly predicted that there would be no relation.     

Distributive judgments under uncertainty: Paccioli's game revisited

Many decision biases arise from the inability to ignore past events. The coherence of decisions is also compromised by the inability to fully use information related to the future. In Paccioli's game, a stake of money goes to the first player to score a certain number of wins. When the game is prematurely interrupted, they may divide the stake according to the proportions of wins relative to rounds played. Alternatively, they may assess the probability that a player would reach the criterion number of wins first if the game were continued. The first decision rule (ratio), which is past-oriented, leads to contradictions across games. The second rule (probability), which is future-oriented, does not. In seven studies, use of the ratio rule emerges across testing methods, in games of chance and games of skill, and independently of extraneous factors (such as random responding, lack of awareness, or proneness to other past-oriented biases).     

Speeded induction under uncertainty: The influence of multiple categories and feature conjunctions

When people are uncertain about the category membership of an item (e.g., Is it a dog or a dingo?), research shows that they tend to rely only on the dominant or most likely category when making inductions (e.g., How likely is it to befriend me?). An exception has been reported using speeded induction judgments where participants appeared to use information from multiple categories to make inductions (Verde, Murphy, &; Ross, 2005). In two speeded induction studies, we found that participants tended to rely on the frequency with which features co-occurred when making feature predictions, independently of category membership. This pattern held whether categories were considered implicitly (Experiment 1) or explicitly (Experiment 2) prior to feature induction. The results converge with other recent work suggesting that people often rely on feature conjunction information, rather than category boundaries, when making inductions under uncertainty.     

Attention and expectation in human predictive learning: the role of uncertainty

Three localized, visual pattern stimuli were trained as predictive signals of auditory outcomes. One signal partially predicted an aversive noise in Experiment 1 and a neutral tone in Experiment 2, whereas the other signals consistently predicted either the occurrence or absence of the noise. The expectation of the noise was measured during each signal presentation, and only participants for whom this expectation demonstrated contingency knowledge showed differential attention to the signals. Importantly, when attention was measured by visual fixations, the contingency-aware group attended more to the partially predictive signal than to the consistent predictors in both experiments. This profile of visual attention supports the Pearce and Hall (1980) theory of the role of attention in associative learning.     

Sequence learning under dual-task conditions: alternatives to a resource-based account

In two experiments with the serial reaction-time task, participants were presented with deterministic or probabilistic sequences under single- or dual-task conditions. Experiment 1 showed that learning of a probabilistic structure was not impaired over a first session by performing a counting task, but that such an interference arose over a second session, when the knowledge was tested under single-task conditions. In contrast, the effects of the secondary task arose earlier for participants exposed to deterministic sequences. This difference between deterministic and probabilistic sequences disappeared in Experiment 2, where the counting task was performed on tones associated to the locations. Comparisons between sessions indicated that the secondary task affected not only the expression but also the acquisition of sequence learning, and that greater interference was observed in those conditions that yielded more explicit knowledge. These results suggest that the effects of a dual task on the measures of implicit sequence learning may be partly due to the intrusion of explicit knowledge and partly due to the disruption of the sequence produced by the inclusion of random events.     

The jumping to conclusions bias associated with symptoms in schizophrenia: which factors influence this bias?

ABSTRACT

Jumping-to-conclusions (JTC) bias has been found to be associated with delusions. However, there is limited data about which other factors are related to JCT bias. The aim of the current study was to investigate the effects of psychotic, cognitive, impulsivity and anxiety symptoms on JTC bias. We compared individuals with paranoid schizophrenia, non-paranoid schizophrenia, generalized anxiety disorder and healthy individuals in terms of JTC bias. Beads task (90:10 and 60:40 versions), Tower of London task, BPRS, SAPS, SANS and BIS-11 were applied. Our results show that the factors related with JTC bias were different on each version of the Beads task. Education levels, anxiety and negative symptoms of individuals with schizophrenia were found to affect JTC bias on 90:10 version of task, whereas positive symptoms and impulsivity levels were related to JTC bias on 60:40 version. These results are important as they contribute to our understanding of this relationship.     

The role of causal models in multiple judgments under uncertainty

Two studies examined a novel prediction of the causal Bayes net approach to judgments under uncertainty, namely that causal knowledge affects the interpretation of statistical evidence obtained over multiple observations. Participants estimated the conditional probability of an uncertain event (breast cancer) given information about the base rate, hit rate (probability of a positive mammogram given cancer) and false positive rate (probability of a positive mammogram in the absence of cancer). Conditional probability estimates were made after observing one or two positive mammograms. Participants exhibited a causal stability effect: there was a smaller increase in estimates of the probability of cancer over multiple positive mammograms when a causal explanation of false positives was provided. This was the case when the judgments were made by different participants (Experiment 1) or by the same participants (Experiment 2). These results show that identical patterns of observed events can lead to different estimates of event probability depending on beliefs about the generative causes of the observations.