Confidence judgments in syllogistic reasoning: the role of consistency and response cardinality

In two experiments, we examined the resolution of confidence judgments in syllogistic reasoning and their heuristic bases. Based on the assumptions of Koriat's Self-Consistency Model of confidence, we expected the confidence judgments to be related to conclusion consensuality, reflecting the role of consistency as a heuristic cue to confidence. In Experiment 1, the participants evaluated 24 syllogisms with conclusions that varied with respect to validity and consensuality. In Experiment 2, the participants produced conclusions to 64 pairs of premises. The correlation between confidence and reasoning accuracy was low. In both experiments confidence was related to the consensuality of the responses. For consensually correct items, correlation between confidence and accuracy was positive; however, for consensually incorrect items it was negative. In Experiment 2, confidence was lower for syllogisms with higher response cardinality, or syllogisms that elicited a greater variety of conclusions.     

Monitoring the source monitoring

The hypothesis that the retrieval of correct source memory cues, those leading to a correct source attribution, increases confidence, whereas the retrieval of incorrect source memory cues, those leading to a source misattribution, decreases confidence was tested. Four predictions were derived from this hypothesis: (1) confidence should be higher for correct than incorrect source attribution except; (2) when no source cues are retrieved; (3) only the source misattributions inferred from the retrieval of incorrect source cues will be rated with low confidence; and (4) the number of source cues retrieved, either correct or incorrect, will affect the confidence in the source attributions. To test these predictions, participants read two narratives from two witnesses to a bank robbery, a customer and a teller. Then, participants completed a source monitoring test with four alternatives, customer, teller, both, or neither, and rated their confidence in their source attribution. Results supported the first three predictions, but they also suggested that the number of correct source monitoring cues retrieved did not play a role in the monitoring of the accuracy of the source attributions. Attributions made from the recovery of incorrect source cues could be tagged as dubious or uncertain, thus leading to lowered confidence irrespective of the number of incorrect source cues or whether another correct source cue was also recovered. This research has potential applications for eyewitness memory because it shows that confidence can be an indicator of the accuracy of a source attribution.     

Memory in motion: Movement dynamics reveal memory strength

Recognition memory is typically examined as a discrete end-state, describable by static variables, such as accuracy, response time, and confidence. In the present study, we combined real-time mouse-tracking with subsequent, overt confidence estimates to examine the dynamic nature of memory decisions. By examining participants' streaming x-, y- mouse coordinates during recognition decisions, we observed that movement trajectories revealed underlying response confidence. More confident decisions were associated with shorter decision times and more linear response trajectories. Less confident decisions were made slowly, with increased trajectory curvature. Statistical indices of curvature and decision times, including area-under-the-curve and time to maximum deviation, suggested that memory strength relates to response dynamics. Whether participants were correct or incorrect, old responses showed a stronger correspondence between mouse trajectories and confidence, relative to new responses. We suggest that people subjectively experience a correspondence between feelings of memory and feelings of confidence; that subjective experience reveals itself in real-time decision processes, as suggested by sequential sampling models of recognition decisions.     

Committing memory errors with high confidence: older adults do but children don't

We investigated lifespan differences of confidence calibration in episodic memory, particularly the susceptibility to high-confidence errors within samples of children, teenagers, younger adults, and older adults. Using an associative recognition memory paradigm, we drew a direct link between older adults' associative deficit and high-confidence errors. We predicted that only older adults would show high-confidence error even though their memory performance was at a similar level to that of children. Participants of all ages showed higher confidence following correct responses compared to incorrect responses, demonstrating the ability to calibrate subjective confidence in relation to memory accuracy. However, older adults were disproportionately more likely to indicate high confidence following erroneously remembered word pairs than participants of the other three age groups. Results are discussed in relation to the misrecollection account of high-confidence errors and ageing-related decline in hippocampus-dependent episodic memory functions.     

Enhanced perceived responsibility decreases metamemory but not memory accuracy in obsessive-compulsive disorder (OCD)

Mixed findings have been obtained in prior research with respect to the presence and severity of memory and metamemory deficits in obsessive-compulsive disorder (OCD). We tested the hypothesis that experimentally induced increments of subjective responsibility would lead to a disproportionately strong decline of memory confidence and enhanced response latencies in OCD while leaving memory accuracy unaffected. Twenty-eight OCD patients and 28 healthy controls were presented a computerized memory test framed with two different scenarios. In the neutral scenario, the participant was requested to imagine purchasing 15 items from a do-it-yourself store. In the recognition phase, the 15 needed items were presented along with 15 distractor items. The participant was asked to decide whether items were on his or her shopping list or not, graded by subjective confidence. In the responsibility scenario, the general experimental setup was analogous except that the participant now had to envision that he or she was a helper in a region recently struck by an earthquake, dispatched to provide 15 urgently needed goods from a nearby town. In line with prior work by our group, samples did not differ in either condition on memory accuracy in a subsequent recognition task. As hypothesized, OCD participants were less certain in their responses for the high responsibility condition than controls. Whereas patients and controls did not differ in their subjective estimates for memorized items, patients expressed stronger doubt that their earthquake mission was successful. The findings indicate that low memory confidence in OCD may only be elicited in situations where perceived responsibility is high and that patients may share higher performance standards ("good is not good enough") than controls when perceived responsibility is inflated.     

Errors committed with high confidence are hypercorrected.

The relation between people's confidence in the accuracy of an erroneous response and their later performance was investigated. Most models of human memory suggest that the higher a person's confidence, the stronger the item (in the context of the eliciting cue) that is retrieved from memory. In recall, stronger associates to a cue interfere with competing associates more than do weaker associates. This state of affairs implies that errors endorsed with high, rather than low, confidence should be more difficult to correct by learning the correct response feedback. In contrast to the authors' expectations, highly confident errors were the most likely to be corrected in a subsequent retest. Participants nearly always endorsed the correct response in cases in which both the correct response and the original erroneous response were generated at retest, suggesting that people possess a refined metacognitive ability to know what is correct and incorrect.     

The effects of the age of eyewitnesses on the accuracy and suggestibility of their testimony

Previous studies have compared the performance of young adult eyewitnesses with that of children or elderly eyewitnesses, but few studies have allowed direct comparison of the performance of all three age groups. The accuracy and suggestibility of accounts of a video recording of a kidnapping were investigated using an experimental eyewitness paradigm. Subjects were drawn from three age groups: children (aged 7–9 years); young adults (aged 16–18 years) and elderly subjects (aged 60–85 years). Subjects' accuracy in answering non-misleading questions and their susceptibility to misleading information was measured. Both the elderly and child subjects gave fewer correct answers and more incorrect answers to non-misleading questions than did young adults. The elderly subjects gave fewer correct responses but also fewer incorrect responses to non-misleading questions than did child subjects. Children were more suggestible than either elderly or young adults. No significant difference was found in the suggestibility of elderly and young adults. Contrary to the trace strength hypothesis no relationship was found between accuracy of recall and suggestibility. © 1997 John Wiley & Sons, Ltd.     

Effects of alternating set for speed or accuracy on response time,accuracy and confidence in a unidimensional discrimination task

Three approaches to a theoretical analysis of confidence judgments are considered: one linking confidence to the number of sensory observations, one based on a distinction between ‘state’ and ‘process’ factors, and a ‘balance of evidence’ hypothesis developed from an accumulator model of discrimination. An experiment is described in which observers were asked to decide which of two parallel lines was the longer, and to rate their confidence in each judgment. Each observer's set for speed or accuracy was manipulated over successive blocks of trials, and effects on response time, accuracy, and confidence were examined. Under an accuracy set, observers produced a higher percentage of correct responses, had longer response times, and made more confident judgments than under a set for speed. Within both speed and accuracy blocks, however, confidence ratings were inversely related to response time. The data on response accuracy, time, and confidence indicate certain deficiencies in either of the first two approaches, but were well accounted for by the third.     

Behaviors speak louder than explicit reports: Implicit metacognition in 2.5‐year‐old children

Recent research has shown that children as young as age 3.5 show behavioral responses to uncertainty although they are not able to report it explicitly. In this study, we tested the hypothesis that some form of metacognition is already available to guide children's decisions before the age of 3. Two groups of 2.5‐ and 3.5‐year‐old children were asked to complete a forced‐choice perceptual identification test and to explicitly rate their confidence in each decision. Moreover, participants had the opportunity to ask for a cue to help them decide if their response was correct. Our results revealed that all children asked for a cue more often after an incorrect response than after a correct response in the forced‐choice identification test, indicating a good ability to implicitly introspect on the results of their cognitive operations. On the contrary, none of these children displayed metacognitive sensitivity when making explicit confidence judgments, consistent with previous evidence of later development of explicit metacognition. Critically, our findings suggest that implicit metacognition exists much earlier than typically assumed, as early as 2.5 years of age.     

Initial Testing Does Not Necessarily Affect Eyewitness Recall Assessed by Specific Questioning

In eyewitness situations, open recall is followed by specific questioning about the witnessed event. We examined whether initial testing affects later recall of actions and specific details. After watching a video of a bank robbery, participants completed an initial testing phase that involved free recall, specific questions about the actions or details of the event or a control condition with no initial testing. In the final test, correct and incorrect answers, accuracy and response confidence for actions and details were analysed. Initial testing affected neither recall nor confidence. The participants were more accurate for actions than details. Response confidence was higher for correct than incorrect answers and higher for details than actions in correct answers. The results showed that specific questioning affects differently the recall of event actions and details and that remembering details increased confidence. Investigative interviewers can use this evidence when questioning information not reported in initial testing.Copyright © 2015 John Wiley & Sons, Ltd.     

Factors that influence search termination decisions in free recall: an examination of response type and confidence

In three experiments search termination decisions were examined as a function of response type (correct vs. incorrect) and confidence. It was found that the time between the last retrieved item and the decision to terminate search (exit latency) was related to the type of response and confidence in the last item retrieved. Participants were willing to search longer when the last retrieved item was a correct item vs. an incorrect item and when the confidence was high in the last retrieved item. It was also found that the number of errors retrieved during the recall period was related to search termination decisions such that the more errors retrieved, the more likely participants were to terminate the search. Finally, it was found that knowledge of overall search set size influenced the time needed to search for items, but did not influence search termination decisions.     

Phonetic symbolism and feeling of knowing

American subjects matched English antonym pairs with antonym pairs from Thai, Kannada, and Yoruba. For each match they also,indicated their degree of feeling of knowing (FOK). Although translation accuracy was in general better than chance, suggesting universal phonetic symbolism, it was found to vary considerably with the subject’s FOK ratings: Translations evoking stronger FOK were more likely to be correct than translations evoking weaker FOK, and matches which proved correct were assigned higher FOK ratings than matches which proved incorrect. Several implications of the finding that subjects can monitor their success in guessing the meaning of words from foreign languages are outlined.     

Can People Identify “Deceptive” or “Misleading” Items that Tend to Produce Mostly Wrong Answers?

In many domains, two‐alternative forced‐choice questions produce more correct responses than wrong responses across participants. However, some items, dubbed “deceptive” or “misleading”, produce mostly wrong answers. These items yield poor calibration and poor resolution because the dominant, erroneous response tends to be endorsed with great confidence, even greater than that of the correct response. In addition, for deceptive items, group discussion amplifies rather than mitigates error while enhancing confidence in the erroneous response. Can participants identify deceptive items when they are warned about their existence? It is argued that people's ability to discriminate between deceptive and non‐deceptive items is poor when the erroneous responses are based on the same process assumed to underlie correct responses. Indeed, participants failed to discriminate between deceptive and non‐deceptive perceptual items when they were warned that some of the items (Experiment 1) or exactly half of the items (Experiment 2) were deceptive. A similar failure was observed for general‐knowledge questions (Experiment 3) except when participants were informed about the correct answer (Experiment 4). Possibly, for these tasks, people cannot escape the dangers lurking in deceptive items. In contrast, the results suggest that participants can identify deceptive problems for which the wrong answer stems from reliance on a fast, intuitive process that differs from the analytic mode that is likely to yield correct answers (Experiment 5). The practical and theoretical implications of the results were discussed. Copyright © 2017 John Wiley & Sons, Ltd.     

The persistence of the fluency–confidence association in problem solving

Confidence in answers is known to be sensitive to the fluency with which answers come to mind. One aspect of fluency is response latency. Latency is often a valid cue for accuracy, showing an inverse relationship with both accuracy rates and confidence. The present study examined the independent latency–confidence association in problem-solving tasks. The tasks were ecologically valid situations in which latency showed no validity, moderate validity, and high validity as a predictor of accuracy. In Experiment 1, misleading problems, which often elicit initial wrong solutions, were answered in open-ended and multiple-choice test formats. Under the open-ended test format, latency was absolutely not valid in predicting accuracy: Quickly and slowly provided solutions had a similar chance of being correct. Under the multiple-choice test format, latency predicted accuracy better. In Experiment 2, nonmisleading problems were used; here, latency was highly valid in predicting accuracy. A breakdown into correct and incorrect solutions allowed examination of the independent latency–confidence relationship when latency necessarily had no validity in predicting accuracy. In all conditions, regardless of latency’s validity in predicting accuracy, confidence was persistently sensitive to latency: The participants were more confident in solutions provided quickly than in those that involved lengthy thinking. The study suggests that the reliability of the latency–confidence association in problem solving depends on the strength of the inverse relationship between latency and accuracy in the particular task.     

No pain no gain: The positive impact of punishment on the strategic regulation of accuracy

Previous studies have shown that punishing people through a large penalty for volunteering incorrect information typically leads them to withhold more information (metacognitive response bias), but it does not appear to influence their ability to distinguish between their own correct and incorrect answers (metacognitive accuracy discrimination). The goal of the current study was to demonstrate that punishing people for volunteering incorrect information—versus rewarding volunteering correct information—produces more effective metacognitive accuracy discrimination. All participants completed three different general-knowledge tests: a reward test (high points for correct volunteered answers), a baseline test (equal points/penalties for volunteered correct/incorrect answers) and a punishment test (high penalty for incorrect volunteered answers). Participants were significantly better at distinguishing between their own correct and incorrect answers on the punishment than reward test, which has implications for situations requiring effective accuracy monitoring.     

Responding to Feedback after Multiple-choice Answers: The Influence of Response Confidence

A model is proposed that describes how people choose multiple-choice answers and deal with subsequent feedback. Central to the model is the construct of likelihood, a signal that: (1) indexes the match between task demands and what is known, (2) gives a basis for decisions, and (3) is estimated by response confidence ratings. In two experiments, subjects answered an item, rated confidence, and studied feedback (for 100 items), and then answered the items again. All predictions of the model were confirmed. Specifically, for right/wrong feedback messages, feedback study intervals were longer after correct than incorrect choices (for correct choices, these intervals decreased as confidence increased). Also, there was no relation between the probability of correcting errors and level of confidence. For feedback messages that consisted of an item with correct alternative marked, there was a positive relation between probability of correcting errors and level of confidence.     

Chronometric evidence for memory retrieval in causal conditional reasoning: The case of the association strength effect

According to many models of conditional reasoning, correct responses to the uncertain forms affirmation of consequent (AC) and denial of antecedent (DA) rely on the retrieval of an alternative antecedent from semantic memory. The main prediction issuing from this hypothesis of online retrieval is that the associative strength of the relevant information in long-term memory affects the latency of its retrieval and then of the correct response of uncertainty to AC and DA. This prediction was tested in 64 adults who performed a syllogism evaluation task with familiar causal conditional premises. As predicted, correct responses of uncertainty to AC and DA were not only more frequent but also faster when the antecedent term was weakly rather than strongly associated with the consequent. In line with the semantic framework, we also observed that this strength effect in response times was reversed for incorrect responses of certainty to AC.     

Rats' knowledge of incorrect responses in a matching-to-sample task

The idea that rats possess some knowledge about having been incorrect when having previously made an incorrect response is proposed. A matching-to-sample task with a second choice was used to assess the hypothesis. Only if the correct lever was chosen on the second choice was the animal rewarded. Further, the second choice was offered whether the animal was initially correct or incorrect in its first choice, preventing it from "knowing" if it had been correct in its initial decision. If rats possessed some knowledge whether they had been correct or not, then the number of correct second choice responses after an incorrect first one should have been greater than chance. The current experiment provided some evidence that the rats did "know" when its choices were incorrect.     

社会信念的自信判断:自我一致性模型的验证

采用重复测验的自信判断范式,检验自我一致性模型。项目一致性与项目共识性的分析表明:被试从大量共享的信念总体中抽取样本进行正误判断,自信判断则基于各个样本信念的一致性,并反映新样本做相同判断的可能性;建立在代表性样本上的判断有更高的一致性或共识性,信念建构的反应时更短且自信更高;项目一致性与项目共识性存在交互关系:高一致性的判断也是高共识性的判断。结果验证了自我一致性模型在中国文化背景下的存在。     

Some deleterious consequences of the act of recollection

Two experiments investigated the impact of responding to recognition test items that do not include a correct alternative. In Experiment 1, subjects who were given exclusively incorrect response alternatives were less likely than control subjects to favor the correct alternatives on a second recognition test. Analysis of subjects’ responses indicated that commitments, rather than distractor familiarity, was the main source of this effect. In Experiment 2, an impairing effect of committing to an incorrect alternative was observed even when the initial distractors were excluded from the final test. Thus, this decreased performance cannot simply be attributed to a bias toward remaining consistent. One interpretation of these results is that committing to a distractor causes subjects to remember a false detail that can interfere with their later ability to access the original information. Other potential theoretical and applied implications of these results are explored.