Symposia

Effects of Positive and Negative Context on Linguistic Probabilities
Mandeep K. Dhami , Middlesex University, United Kingdom
When people use high and low probability terms such as probable and improbable, they intend for them to have discriminable meanings. However, the context in which such terms are applied can affect how they are interpreted and how well they are discriminated. In Experiment 1, we examined the effect of context (i.e., positive vs. negative event) on the numeric interpretations of high and low probability terms. In Experiment 2, we additionally examined the effect of intensifying a term (e.g., very likely). Both experiments also examined the effect of these variables on judgments made on the basis of statements containing the terms. We found that compared to the positive context, the negative context significantly reduced discrimination for both the interpretation of probability terms and judgments. We discuss the potential reasons for why negative contexts may reduce the discriminability of probability terms.

Communicating Probabilistic Information – The Role of Format on Understanding and Perceived Communicator Credibility
Sarah Jenkins , University College London, United Kingdom
Much previous research has explored the effect of communication format on understanding, mainly in relation to verbal probability expressions [VPEs, e.g., ‘unlikely’]. However, very little has investigated the effect of format on the perceived credibility of a communicator – a vital component for the effective communication of probabilistic information. In two studies communicating natural hazard risks, we compare VPEs, numerical expressions (e.g., ‘20% likelihood’) and mixed expressions in two orders (verbal-numerical, e.g., ‘unlikely [20% likelihood]’ and numerical-verbal format, e.g., ‘20% likelihood [unlikely]’). We find considerable differences in the way these formats are understood, with use of verbal and verbal-numerical formats leading to an ‘extremity effect’. We show that this understanding has downstream effects for the perceived credibility of a communicator. We conclude by proposing a pragmatic account of communication format, in which individuals draw inferences from the communicator’s choice of format.

Is People’s Use of Attribute Frame Information Adaptive when Receiving Probabilistic Communications?
Adam J. L. Harris , University College London, United Kingdom
The informational leakage account of attribute framing effects proposes that a speaker’s choice of frame provides informational value, such that different frames are not informationally equivalent. Across three studies communicating food risks, we investigated the adaptiveness of a listener’s use of frame information by manipulating the degree to which the speaker ostensibly had a choice over how the information was framed. Within-participants framing effects were observed across all studies, and these effects were not moderated by the speaker’s degree of choice. If framing effects are driven by the informational value contained in frame choice, people do not appear to be sensitive to situations where that choice is removed.

Arithmetic Computation with Probability Words and Numbers
David R. Mandel , Defence Research and Development, Canada, Canada
Probability information is often communicated verbally (e.g., “likely”) rather than numerically (e.g., “p=.75”). However, people learn to do arithmetic with decimals but not with verbal probabilities. We hypothesized—and found in two experiments (Ns = 213 and 201) that manipulated communication format (numeric, verbal) between-subjects—that the accuracy and coherence (i.e., respecting normative constraints) of averaging and multiplying probabilities is poorer when individuals receive verbal rather than numeric probability information. In Experiment 2, translating probabilities from the verbal format to the numeric format improved accuracy, whereas translating from the numeric format to the verbal format reduced accuracy. Whereas incoherence was related to (low) numeracy and verbal-reasoning ability in the numeric condition, it was not related to either cognitive performance measure in the verbal condition. Communicating probabilities numerically improved arithmetic computation and, evidently, many individuals do not have a schema for performing arithmetic operations on verbal probabilities.

Estimation of Subjective Probability Distributions with Ratio Judgments and Scaling (RJS) Methodology
David V. Budescu , Fordham University, USA
We propose a new methodology for estimating subjective probabilities of continuous variables. Under the new method (a) the range of the variable is divided into C exhaustive and mutually exclusive intervals, (b) each judge is asked to compare and provide ratio judgments regarding the relative likelihood for each of the C(C-1)/2 pairs of intervals, and (c) a statistical algorithm is used to find the distribution that best fits the collection of judgments. A simulation study and two experiments manipulating the number of intervals, C, and their spacing for various distributions show that these estimates are highly consistent, more accurate, and less sensitive to the partition procedure than direct estimates. We recommend the adoption of the new method and illustrate its use in real life contexts.

Back to top

The Effects of Repetition on Belief: The Role of Prior Knowledge and Development
Lisa K. Fazio, Vanderbilt University, USA
Repetition increases belief in false statements. This illusory truth effect occurs with many different types of statements (e.g., trivia facts, news headlines, advertisements), and even occurs when the false statement contradicts participants’ prior knowledge.  I will present a series of studies demonstrating that the effects of repetition are near universal – occurring for even very implausible statements and occurring across development. Across two studies, we measured the effect of repetition on belief for statements across the full range of plausibility (extremely implausible to extremely plausible).  Regardless of whether plausibility was measured using general knowledge norms or individual knowledge, our results suggest repetition increases belief in all statements. We also find that repetition increased belief in false statements even for young children. Five-year-olds, 10-year-olds and adults were all more likely to judge repeated statements as true. Implications for theoretical explanations of the illusory truth effect will be discussed.

Doubt Deficit: The Neuropsychology of Credulity
Erik Asp, Hamline University, USA
In the early 1990s Dr. Daniel Gilbert contrasted two psychological models of belief and doubt: The Cartesian model (belief is subsequent and separate from comprehension) and the Spinozan model (belief and comprehension are the same process).  Typically, cognitive resource depletion has been used as the standard method to adjudicate between the two models.  Here, neuroscientific data will be brought to bear on the issue.   The diverse constellation of deficits and symptomology following prefrontal cortex damage will be examined, and several empirical studies conducted in lesion patients will be described.  Broadly, our results suggest that damage to the prefrontal cortex tends to increase credulity generally.  Of the lesion patients studied, no one showed a dissociation between comprehension and belief.  These neuropsychological data argue for a Spinozan perspective: belief is inextricably linked to comprehension, and the prefrontal cortex mediates retroactive doubt.  Finally, a rudimentary neural circuitry model of belief and doubt will be offered using recent discoveries in fear conditioning processes as a guide.

Are We Naturally Gullible or Naturally Skeptical?
Ruth Mayo, Hebrew University, Israel
The main claim for the strength of the influence of misinformation is that the process of comprehension entails believing as a primary process, while rejection is only a secondary process, prone to failure. I will present research demonstrating that rejection can be successful and primary. Specifically, I will report studies demonstrating that together with a gullible mindset, humans also have a skeptical mindset. The mindsets alter according to context and individual differences regarding trust. Critically, we find that in a skeptical mindset, one's initial response is to reject information, hence eliminating congruent effects such as priming, confirmatory biases, accessibility effects and routine reasoning. I will end with a discussion suggesting that the skeptical mindset offers new insights both regarding cognitive antecedents as well as possible remedies for the post-truth era.

Developmental Origins of the Ability to Differentiate between Real and Not-Real
Jacqueline D. Woolley, University of Texas at Austin, USA
Children spend considerable time in worlds in which reality and non-reality are intertwined. Storybooks and television present events that merge real and fantastical elements, peers pretend one thing is another, and parents go to great lengths to promote belief in fantastical beings. How do children decide what to believe? I argue that young children have sophisticated tools for figuring this out and use a range of cues and strategies to make these decisions. I will present a series of empirical studies that show that, with age, children increasingly use physical evidence, cues in peoples’ everyday conversations, and context to make reality judgments about novel entities. I will also present evidence that children use existing knowledge and beliefs about both the world and the people in it to draw conclusions about what’s real. I conclude that the process of determining reality status can reveal much about how we think.

Perceived Truth as a Function of the Number and Spacing of Repetitions
Sarah J. Barber, Georgia State University, USA
Repeated information is perceived as more truthful than new information. This is known as the illusory truth effect. It occurs because repetition increases processing fluency.  Because fluency and truth are often correlated, people use processing fluency as a marker for truthfulness. Although this is a robust effect, almost all prior studies examining it have used three or fewer repetitions, and no study has examined whether repetition spacing matters.  In a series of experiments, we examined how the number and spacing of repetitions affected the perceived validity of trivia statements. We found that perceived validity increased as the number of repetitions increased (up to 27 repetitions). We also found that perceived validity was higher for spaced repetitions than for massed repetitions, and tended to increase as the lag between the repetitions increased.  These findings add to our theoretical understanding and have applications for advertising, politics, and the propagation of “fake news.”

Back to top

The Mental Representation of Race
Mahzarin R. Banaji, Harvard University, USA
Although traditional textbooks on cognitive psychology hardly make any mention of the construct of race (leaving it, I suppose, to other areas of psychology to mop up) I will argue that understanding the representation of race is fundamental to understanding human perception and cognition. The construct of race reveals core principles of attention and perception, learning and memory, and decision making and judgment.   It makes the study of human cognition come alive for students of cognitive psychology as few issues are able. Psychology will be better served if understanding the human mind can include, as a matter of course, the richness of human social categories, their mental representations and transformation over time.

How to Predict Bad Policing: Theory and Evidence
Phillip Atiba Goff, Yale University, USA & Center for Policing Equity, USA;
with Jillian K. Swencionis, Yale University, USA & Center for Policing Equity, USA;
and Hilary Rau, Center for Policing Equity, USA
What situations predict terrible police outcomes? We unite a situationist approach from psychology with routine activity theory from criminology to theorize the factors that predict police officers’ discrimination and violence. In this framework, when officers who are vulnerable to situational risk factors for bias encounter situations that facilitate bias and are not well regulated by norms or policies, they are more likely to engage in discrimination and violence. As an example, we describe evidence from three U.S. cities showing that White police officers who support social stratification may use physical force to maintain hierarchies. Because police are empowered to use force to maintain social order, and because White officers hold a dominant racial identity, we hypothesized White officers with high Social Dominance Orientation (SDO) would be more likely to use force than low-SDO White officers. This work suggests a need to continue investigating situations as potential opp ortunities for reform.

Visual dehumanization of Blacks under economic stress:  ERP and fMRI evidence and implications for biased behavior
David M. Amodio, New York University, USA & University of Amsterdam, The Netherlands
When the economy declines, racial discrimination typically increases. Previously, we found that perceived economic scarcity leads White Americans people to view Blacks as darker and more “stereotypically Black,” which then predicted discriminatory decisions. Here, we demonstrate that scarcity even affects early face processing, selectively impairing configural encoding of Black faces—a form of “perceptual dehumanization.” In Study 1, the framing of resources as scarce (vs. neutral) delayed the N170 ERP to Black (relative to White) faces, and this effect predicted anti-Black money allocation decisions. In Study 2, fMRI revealed selective decreases in fusiform and striatal activity to Black faces under scarcity, which together predicted race-biased decisions. Hence, scarcity appears to impede the encoding of Black faces as face-like, which facilitates racial discrimination. This pattern supports a “visual dehumanization” account of scarcity effects on discrimination and suggests a new mechanism through which economic stress can exacerbate racial inequality.

Moving Beyond a Hypodescent Framework for Ambiguous Face Categorization
Sarah E. Gaither, Duke University, USA
Social categorization of some faces is easier than others. In particular, categorizing racially ambiguous faces can be more difficult and cognitively taxing compared to unambiguous faces. Many theorists have argued that people’s judgments in racially ambiguous categorization are driven by the one-drop rule, a heuristic whereby one drop of “Black blood” identifies a mixed-race individual as Black (Davis, 1991) also known as “hypodescent” (Ho, Sidanius, Levin, & Banaji, 2011). However, my recent meta-analysis (k = 55; Young, Sanchez, Pauker, & Gaither, in press) shows the majority of this work focuses on White and western samples and biracial Black/White and male stimuli, meaning we do not know how generalizable these findings are. Here, I discuss this meta-analysis and two studies testing the role that participant race (White, Black, Asian, Biracial; N = 515) and cultural group membership (Asian American vs. Taiwanese children; N = 139) play in shaping distinct perceptions of racially ambiguous faces.

Back to top

The Insight Memory Advantage
Amory H. Danek, Universität Heidelberg, Germany ;
with Jennifer Wiley, University of Illinois at Chicago, USA
The idea that experiencing an insight could contribute to better memory is intuitively appealing and supported by prior research. However, the role of solution correctness as well as the role of solvers' confidence has remained unclear so far. The present study used magic tricks as a problem solving task to test the hypothesis that solution correctness, the strength of the Aha! experience, and feelings of confidence would each independently predict better recall of solutions after one week. As expected, solutions associated with Aha! experiences were remembered better than those without. Correctness and confidence independently predicted better solution memory. None of the two-way interactions between the predictors was significant. The lack of an interaction between ratings of Aha! and correctness suggests that regardless of correctness, the feelings underlying the Aha! experience are associated with stronger memory traces. This leads to the counterintuitive conclusion that at least part of the insight memory advantage is not due to actually having solved a problem correctly, but can also occur for incorrect solutions if accompanied by an Aha! experience.

The Dark Side of Eureka
Ruben Laukkonen, Vrije Universiteit, Netherlands;
with Benjamin T. Kaveladze, University of California, Santa Barbara, USA;
John Protzko, University of California, Santa Barbara, USA;
Jason M. Tangen, University of Queensland, Australia;
and Jonathan W. Schooler, University of California, Santa Clara, USA
Some ideas feel mundane, but others seem immediately profound. I propose that feelings of insight make ideas feel more valuable to aid quick decision-making, akin to a heuristic. Since shortcuts can incur errors, we predicted that facts would appear truer if artificially accompanied by aha moments. In a preregistered experiment, we found that participants (n = 300) gave higher truth ratings for statements accompanied by solved anagrams, and the effect was pronounced when participants reported an aha experience. We also replicated the effect showing that aha moments can shift core beliefs such as, "life has purpose" or "free will is an illusion". Although feelings of insight usually accompany correct ideas (Salvi et al., 2016), we found that aha moments can also be overgeneralized and bias how true a belief or fact appears. I discuss the potential dark sides of aha moments, with potential clinical importance.

Forgetting as a Mechanism for Overcoming Fixation in Creative Problem Solving
Benjamin Storm, University of California, Santa Cruz, USA
Research on creative problem solving has shown that prior knowledge and existing ideas and solutions can impede the generation of new ideas and solutions.   This phenomenon, known as mental fixation, has been observed in many problem-solving contexts, including in studies using the Remote Associates Test.   In the Remote Associates Test, participants are presented with three cue words and asked to try to come up with a fourth word that is related to each of the three words.   The task, however, can be made more difficult by exposing participants to unhelpful associates to the three cue words before having them attempt to generate the fourth word.   The current talk will describe research on the mechanisms by which people appear to be able to overcome the effects of mental fixation, focusing in particular on research using the Remote Associates Test, and on the potential roles of inhibition and forgetting.

Individual Differences that Moderate Fixation, Creative Solutions, and the Insight Experience
Jennifer Wiley, University of Illinois at Chicago, USA
with Tim George, Union College, USA
and Rebecca Koppel, University of Illinois at Chicago, USA
For analytic problem solving, the ability to focus and resist distraction is generally beneficial. Similarly, expertise generally aids in the solution process. In contrast, for creative problem solving, sometimes attentional control and relying on past experience are less helpful. Rather, reaching an insightful solution is thought to require overcoming an impasse, fixation, or mental set imposed by our interpretation of a problem, which may be exacerbated by the problem solving context, the activation of inappropriate prior knowledge, or focusing too much on an incorrect solution. This talk will discuss findings from research exploring the role of individual differences in predicting who is most likely to experience impasse, overcome fixation, generate creative ideas, and experience feelings of insight.

Jumping About: The Role of Task-Switching in Facilitating Creative Problem-Solving
Paul Seli , Duke University, USA;
with Nicholaus P. Brosowsky, Duke University, USA;
Madeleine E. Gross, University of California, Santa Barbara, USA;
and Jonathan W. Schooler, University of California, Santa Barbara, USA
Research on creativity has long touted the benefits of an incubation interval. One interpretation of such an interval is that it enables task-switching. And, indeed, some evidence is supportive of the value of switching between tasks in enhancing creative performance. Such benefits may be due to at least two possible sources: First, task-switching may enable people to break set in order to identify solutions that were otherwise unavailable. Second, task-switching may encourage the alternation between idea generation and evaluation. In this talk, we consider the available evidence for the benefits of task-switching, for both mental-set breaking and idea-generation/evaluation alternation, and describe new studies that examine these potential sources of benefit of task-switching on creativity.

Oculometric Signature of Switch into Awareness?
Carola Salvi, University of Texas at Austin, USA
For the Gestalt theorists, restructuring is an essential component of insight problem-solving, contributing to the 'Aha!' experience and similar to the perceptual switch experienced when reinterpreting ambiguous figures. Because these perceptual and conceptual 'representational changes' rely on similar processes, they should present similar behavioral responses. Previous research showed that pupil diameter increases during the perceptual switch, indexing norepinephrine functioning mediated by the locus coeruleus (LC-NA). Similarly, our data on problem-solving shows that pupil diameter increases before an insight, suggesting the involvement of LC-NA, and it is a possible indicator of the switch into awareness of unconscious processes humans depend upon for insight. LC-NA might be involved during the interruption of ongoing functional networks at the emergence of a new idea, causing a switch of attention toward novel ideas and thus the Aha! feeling. This result is in line with the idea that insight is a highly accurate off-on discontinuous process.

Back to top

The Talk of the Town: A Network Approach to Characterizing Bilingual Conversational Topics in Montreal
Mehrgol Tiv , McGill University, Canada;
with Jason Gullifer, McGill University, Canada;
Ruo Feng, McGill University, Canada;
and Debra Titone, McGill University, Canada
Recent work within bilingualism has sought new methods to characterize how people use language across different communicative contexts. We approach this issue using a novel application of Network Science to map the conversational topics that Montréal bilinguals discuss across communicative contexts, in their dominant vs. non-dominant language. Our results demonstrate that all communicative contexts display a unique pattern in which conversational topics are discussed. We also demonstrate that the dominant language has greater network size, strength, and density than the non-dominant language, suggesting that more topics are used in a wider variety of contexts in this language. Lastly, using community detection to thematically group the topics in each language, we find evidence of greater specificity in the non-dominant language than the dominant language. We contend that Network Science is a valuable tool for representing complex information, such as individual differences in bilingual language use, in a rich and granular manner.

Using Network Science to Assess Speech Patterns in Novel Word Learning in Children with Typical and Atypical Language Development
Sara Benham, University of Texas at Dallas, USA
Preschoolers with developmental language disorder (DLD) produce novel words with low accuracy and high variability. I show, across three studies, how network science approaches provide novel insights into the structure of these children's speech errors. In study 1, we examined syllable connectivity patterns in nonwords, showing that children with DLD were more variable than their peers in the syllables they produced and in the links between syllables. In study 2, we probed the malleability of these networks by associating a visual referent to the novel words. Children with DLD stabilized on a reduced set of nodes and edges when words included a referent, but did not become more accurate. In study 3 (ongoing), we investigate the structure of sound and syllable networks in typical 2-year-olds and preschoolers with DLD to determine whether the speech deficits in DLD are developmental in nature or relate uniquely to their language disorder.

Capturing the Aging Lexicon Using Network Science Techniques: Implications for Dementia and Aphasia
Nichol Castro , University of Washington, USA
Understanding the impact of cognitive aging on language processes remains a thriving area of cognitive science research. One factor that has received a new vigor of interest in this multi-faceted area is the role of mental lexicon structure (i.e., the way in which words are organized in memory) on language processes, like spreading activation or search strategies. The tools of network science have provided a suite of computational and mathematical tools to quantitatively define large, complex systems (e.g., the mental lexicon) and the role of structure on process. This talk will provide examples of capturing the aging lexicon structure through analysis of phonological associations and verbal fluency data using techniques from network science, and will discuss the implications of considering the dynamical change of lexicon structure across adulthood for the purpose of clinical translation, particularly in the context of identifying and treating significant word retrieval impairments.

What do Phonological Networks Do? A Comparison of Simulations to Understand the Process of Spoken Word Recognition
Michael S. Vitevitch, University of Kansas, USA;
with Gavin J. Mullin, University of Kansas, USA
Cognitive Psychology has traditionally examined how information is represented, and the processes that operate on those representations. Contemporary computational models of spoken word recognition adequately capture the processes involved in recognizing words, but fail to account for the influence on processing of the large-scale structure that exists among lexical representations. We compare the ability of a contemporary model of spoken word recognition (jTRACE; Strauss et al, 2007) to the ability of a network model with a spreading activation-like process (spreadr; Siew, 2019) to account for the findings from several previously published behavioral studies of language processing. The results suggest that viable models of spoken word recognition must account for the influence on processing of the large-scale structure that exists among lexical representations.

Back to top

  Symposium VIII: Verbal Working Memory: Domain-General or Domain-Specific?

Introduction
Nazbanou Nozari, Carnegie Mellon University, USA
The symposium Chair welcomes the audience, and gives a brief overview of the topic of the symposium (domain-generality or domain-specificity of verbal working memory) and its relevance to the fields of memory, language, and attention. She then introduces the three speakers, briefly explaining the angle that each talk covers with regard to the overall goal of the symposium.

Verbal Working Memory: Domain General Sequencing Shaped by Domain Specific Sequences
Maryellen C. MacDonald , University of Wisconsin-Madison, USA
Viewing verbal working memory maintenance as emergent from utterance planning in language production yields testable predictions for how long term linguistic knowledge such as word frequency and co-occurrence affects short term maintenance: The same factors shaping serial ordering in everyday speaking should also affect ordering in verbal working memory tasks, though parallels are sometimes obscured by different task demands. This view appears to be a highly domain-specific approach to verbal working memory. However, these claims must contend with evidence that serial ordering biases in language production are themselves emergent from serial ordering processes in action planning more generally. Together, these claims become a mix of domain generality and specificity: a) language production draws on general serial ordering routines, strongly shaped by task goals and the domain-specific statistics of prior language use, and b) verbal working memory maintenance is emergent from these hybrid specific/general language production processes.

Principles of Resource Division in Working Memory: Domain-General or Domain-Specific?
Nazbanou Nozari, Carnegie Mellon University, USA
with Christopher R. Hepner, Carnegie Mellon University, USA
A key characteristic of working memory is its limited capacity often attributed to “limited resources”. The nature of such resources has been studied extensively in the visual domain, with evidence supporting models with flexibly and continuously divisible resources. It remains unclear, however, whether similar mechanisms mediate the division of resources in phonological working memory, given the non-negligible differences between the nature of the representations in the two domains. Adapting continuous reproduction paradigms from visual to phonological domain, I will show that the principles of resource division are indeed similar in visual perception, phonological perception and phonological production. Moreover, I will demonstrate the similarity between the effect of attention on resource division in visual and verbal working memory. Collectively, these results point to domain-general principles of flexible resource allocation in working memory operating on domain-specific representations.

Domain-Specific Phonological Working Memory: Evidence from Neuropsychology and Neuroimaging
Randi C. Martin, Rice University, USA;
with Qiuhai Yue, Vanderbilt University, USA
One domain-general account of working memory (WM) is the embedded processes approach, which argues that WM consists of the activated portion of long-term memory, including a limited number of items in the focus of attention.   Neuropsychological evidence argues instead for dedicated phonological storage, as brain damaged patients show reduced phonological WM capacity that is independent of: 1) phonological processing ability, 2) WM in other domains (e.g., semantic, graphemic), and 3) general attentional abilities. Recent neuroimaging studies provide converging evidence. Voxel-based lesion-symptom mapping reveals an inferior parietal region (supramarginal gyrus; SMG) supporting phonological WM that is separate from the region involved in graphemic WM and in phoneme processing (superior temporal gyrus; STG). An fMRI study of healthy individuals demonstrates that, using response-similarity analysis (RSA), phonological (but not semantic) codes can be decoded in the SMG but not in the STG during the delay period of a phonological WM task.

Domain Generality and Specificity of Verbal Working Memory Training
Thomas S. Redick , Purdue University, USA
Research on working memory training has proliferated over the past two decades. Multiple studies convincingly show that individuals of all ages and abilities greatly improve on the working memory tasks themselves after practicing for multiple days. However, there are conflicting interpretations about what causes the observed changes on the trained working memory tasks, and whether there is transfer to unpracticed tasks and processes. I review the research (including some from our lab) that has examined how and why verbal working memory training and transfer occurs. Our general conclusion is that such transfer is narrow, and likely reflects strategies that trainees have adapted to the specific verbal memoranda and the tasks used. I discuss the implications for accounts of verbal working memory and reasons why the system would be tuned this way.

Question & Answer
Nazbanou Nozari, Carnegie Mellon University, USA
The final 30 minutes of the symposium will be dedicated to questions from the audience, as well as questions from the presenters from one another. Because of the nature of the topic, which requires listening to all four talks before drawing a conclusion about domain-generality or specificity of verbal working memory, we have allocated more time for questions at the very end, as opposed to at the end of each talk. We have also intentionally merged "panel discussion" with open Q & A, to encourage more interaction between presenters and the audience. I have designated the Chair as the "author" for this abstract, but all four presenters will participate.

Back to top