Infusing behavior science into large language models for activity coaching.

Large language models (LLMs) have shown promise for task-oriented dialogue across a range of domains. The use of LLMs in health and fitness coaching is under-explored. Behavior science frameworks such as COM-B, which conceptualizes behavior change in terms of capability (C), Opportunity (O) and Motivation (M), can be used to architect coaching interventions in a way that promotes sustained change. Here we aim to incorporate behavior science principles into an LLM using two knowledge infusion techniques: coach message priming (where exemplar coach responses are provided as context to the LLM), and dialogue re-ranking (where the COM-B category of the LLM output is matched to the inferred user need). Simulated conversations were conducted between the primed or unprimed LLM and a member of the research team, and then evaluated by 8 human raters. Ratings for the primed conversations were significantly higher in terms of empathy and actionability. The same raters also compared a single response generated by the unprimed, primed and re-ranked models, finding a significant uplift in actionability and empathy from the re-ranking technique. This is a proof of concept of how behavior science frameworks can be infused into automated conversational agents for a more principled coaching experience.

Attractors: Incidental values that influence forecasts of change.

This article examines whether forecasts of change are influenced by attractors, salient values in the direction of the considered change. When an attractor is relatively distal from (vs. proximal to) the base value from which change originates, it encourages forecasts of greater change. Participants showed this pattern when predicting which of two airfare changes was imminent (Study 1) and by how much gas prices (Study 2) or a stock's price (Study 3) would change. Attractors have this influence because they alter the way people translate even equivalent subjective interpretations of prospective changes into objective forecasts of change. In the context of a distal (vs. a proximal) attractor, forecasters thought more objective change was necessary to reflect the same subjective characterization of that change (Study 4). Having participants precommit to a subjective interpretation of an objective amount of change reduced a subsequently introduced attractor's influence on forecasting (Study 5). Following almost five decades of research showing many ways arbitrary values anchor judgments, we discuss how attractors reflect the first evidence that such values can also influence adjustment. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Inside the STEM pipeline: Changes in students' biomedical career plans across the college years.

Researchers often invoke the metaphor of a pipeline when studying participation in careers in science, technology, engineering, and mathematics (STEM), focusing on the important issue of students who "leak" from the pipeline, but largely ignoring students who persist in STEM. Using interview, survey, and institutional data over 6 years, we examined the experiences of 921 students who persisted in biomedical fields through college graduation and planned to pursue biomedical careers. Despite remaining in the biomedical pipeline, almost half of these students changed their career plans, which was almost twice the number of students who abandoned biomedical career paths altogether. Women changed plans more often and were more likely than men to change to a career requiring fewer years of post-graduate education. Results highlight the importance of studying within-pipeline patterns rather than focusing only on why students leave STEM fields.

College Students' Reasons for Leaving Biomedical Fields: Disenchantment with Biomedicine or Attraction to Other Fields?

This study examined whether students who left biomedical fields of study during college did so primarily because they became disenchanted with those fields or because they felt attracted to alternative fields of study. We identified 1193 students intending to pursue biomedical fields of study early in college, collected data about their beliefs and performance throughout college, and interviewed them near graduation about their future plans. Descriptively, we examined the topics students discussed as affecting their attrition decisions. Predictive research aims were to determine how academic performance, interest, and demographic factors predicted students' likelihood of overall attrition and likelihood of reporting distinct reasons for attrition. Among the 192 students who left biomedical fields, 62.5% described leaving only in terms of feeling disenchanted, whereas 37.4% expressed that they left at least in part due to feeling attracted towards non-biomedical fields. Most students who left biomedical fields expressed changing plans for reasons related to interest; this was especially prevalent among students who reported leaving due to attraction towards non-biomedical fields. Predictive analyses showed that interest in biology and grades at the end of an introductory biology course predicted the likelihood of overall attrition and likelihood of leaving due to feeling disenchantment, whereas underrepresented ethnic minority status predicted these outcomes positively. Interest and course grades also predicted the likelihood of students leaving due to feeling attraction towards other fields, but interest was a stronger predictor relative to grades. Results highlight distinct types of attrition that may have implications for policies to promote STEM retention.

The Benefits of Combining Value for the Self and Others in Utility-Value Interventions.

Utility-value interventions, in which students complete writing assignments about the personal usefulness of course material, show great promise for promoting interest and performance in introductory college science courses, as well as persistence in STEM fields. As researchers move toward scaling up this intervention, it's important to understand which features are key to its effectiveness. For example, prior studies have used different types of utility-value assignments (i.e., self-focused essays and other-focused letters) and different assignment structures (i.e., over time, researchers provided a variety of tasks or choices between tasks), without comparing them. It is not known whether these assignment features are incidental details or key aspects of the intervention that impact its effectiveness. In the current study, we systematically compared different utility-value assignments, as well as ways of combining them, in a randomized controlled trial in an introductory college biology course (N = 590). Specifically, we compared different versions of the intervention in terms of their relative effectiveness for promoting course performance and the motivational mechanisms through which they operated. The intervention was most effective when students had opportunities to write about utility for both the self and others. Grades were higher in conditions in which students were either assigned a variety of self-focused and other-focused assignments or given the choice between the two. Among students with low performance expectations, grades were higher when students were assigned a specific combination: a self-focused assignment followed by other-focused assignments. Results suggest that different versions of the intervention may work through different mechanisms.

Choose Your Own Intervention: Using Choice to Enhance the Effectiveness of a Utility-Value Intervention.

Utility-value interventions, in which students are asked to make connections between course material and their lives, are useful for improving students' academic outcomes in science courses. These interventions are thought to be successful in part because the intervention activities afford students autonomy while they complete them, but no research has explored directly whether interventions that include more support for autonomy are more effective. In this study, the degree of choice incorporated in a utility-value intervention was systematically varied in order to test this possibility. We assigned college biology students (n = 406) to a high-choice utility-value intervention condition (choose between two formats- essay or letter- for each of 3 writing assignments), one of two low-choice intervention conditions (complete either an essay and then a letter, or vice versa, and choose a format for the third assignment), or a control condition (summarize course material 3 times). Students in the high-choice condition reported significantly higher perceived utility value and interest for biology course content compared to students in the low-choice conditions. There were also significant, but small, indirect effects of choice on students' final course grades and enrollment in the next course in the biology sequence via perceived utility value and interest. Results suggest that social-psychological interventions which include more choice are likely to be more effective than those which include less choice.

How Do Perceptions of Importance Support From a Reading Intervention Affect Students' Motivation, Engagement, and Comprehension?

Interventions can enhance students' motivation for reading, but few researchers have assessed the effects of the specific motivation-enhancing practices that comprise these interventions. Even fewer have evaluated how students' perceptions of different intervention practices impact their later motivation and academic outcomes. In this study, we utilized data from a study of Concept-Oriented Reading Instruction, a program designed to enhance seventh-grade students' reading comprehension and motivation. We examined the effects of students perceiving one practice from this intervention, emphasizing the importance of reading, which was designed to enhance their task values for reading (Eccles-Parsons et al, 1983). Unexpectedly, structural equation modeling analyses showed that students' perceptions of importance support predicted their later competence-related beliefs, but not their task values. Students' competence-related beliefs predicted their reading comprehension and behavioral engagement, whereas students' task values predicted reading engagement. However, there were no significant indirect effects of perceiving importance support on students' reading outcomes.

What if Reading is Easy but Unimportant? How Students' Patterns of Affirming and Undermining Motivation for Reading Information Texts Predict Different Reading Outcomes.

Many affirming and undermining motivational constructs affect students as they read information texts, but few researchers have explored how these motivations are patterned within students. In this study we used cluster analysis to classify middle school students (n = 1,134) based on their patterns of self-efficacy, perceived difficulty, value, and devalue for reading school information texts. We then compared how the patterns predicted students' language arts grades, science information text comprehension, and dedication to reading school information texts. We found and validated a four-cluster solution. One cluster included a pattern of high affirming and low undermining motivations, and another included low affirming and high undermining motivations. Students with these patterns earned the highest and lowest scores, respectively, on all outcomes. A third pattern showed high self-efficacy/low difficulty with low value/high devalue, and a fourth showed moderate levels of all four motivational constructs. Students with the high efficacy and devalue pattern showed high information text comprehension but relatively low dedication. Students with the moderate pattern showed high dedication but low initial information text comprehension. Students with these two patterns earned similar grades. We discuss the implications of our findings for motivation theories and for school instruction that involves information text reading.

With Eyes Wide Open: How and Why Awareness of the Psychological Immune System Is Compatible With Its Efficacy.

Life is full of negative events that threaten our self-worth, and we deploy a wide range of potent psychological strategies-such as dissonance reduction, motivated reasoning, downward comparison, self-serving attribution, and outgroup derogation-to defend our egos. People are highly adept at using these psychological immune system strategies while remaining blind to the fact that they have done so. In fact, prominent voices in the field have suggested that this lack of awareness is a necessary condition for the psychological immune system's efficacy; how else could someone continue to believe a self-serving attribution while being aware that the attribution was generated precisely because it favored him or her? In this article, I outline the argument underlying why awareness might be a threat to the efficacy of the psychological immune system and then closely review the empirical literature for evidence supporting this claim. On the contrary, the data indicate people can and do use these strategies with awareness, intention, and efficacy. I subsequently consider three ways people may achieve the apparent paradox of being aware of their own biased mental processes while also believing the conclusions that result from them. The third of these is a novel conceptual approach to the illusion of objectivity, which highlights the potential for dissociation between the objectivity of our mental processes and of our mental products. Finally, I outline the implications of this work for future theoretical and applied research.

When Knowledge Knows No Bounds: Self-Perceived Expertise Predicts Claims of Impossible Knowledge.

People overestimate their knowledge, at times claiming knowledge of concepts, events, and people that do not exist and cannot be known, a phenomenon called overclaiming. What underlies assertions of such impossible knowledge? We found that people overclaim to the extent that they perceive their personal expertise favorably. Studies 1a and 1b showed that self-perceived financial knowledge positively predicts claiming knowledge of nonexistent financial concepts, independent of actual knowledge. Study 2 demonstrated that self-perceived knowledge within specific domains (e.g., biology) is associated specifically with overclaiming within those domains. In Study 3, warning participants that some of the concepts they saw were fictitious did not reduce the relationship between self-perceived knowledge and overclaiming, which suggests that this relationship is not driven by impression management. In Study 4, boosting self-perceived expertise in geography prompted assertions of familiarity with nonexistent places, which supports a causal role for self-perceived expertise in claiming impossible knowledge.

On motivational readiness.

The construct of motivational readiness is introduced and explored. Motivational readiness is the willingness or inclination, whether or not ultimately realized, to act in the service of a desire. Building on prior relevant conceptions that include, among others, animal learning models (Hull, 1943; Spence, 1956; Tolman, 1955) and personality approaches (e.g., Atkinson, 1964; Lewin, 1935), a general theory of motivational readiness is presented. Major parameters of this theory include the magnitude of a Want state (i.e., individual's desire of some sort) and the Expectancy of being able to satisfy it. The Want is assumed to be the essential driver of readiness: Whereas some degree of readiness may exist in the absence of Expectancy, all readiness is abolished in the absence of desire (Want). The concept of incentive is conceptualized in terms of a Match between the contents of the Want and perceived situational affordances. Whereas in classic models incentive was portrayed as a first-order determinant of motivational readiness, it is depicted here as a second-order factor that affects readiness via its impact on the Want and/or the Expectancy. A heterogeneous body of evidence for the present theory is reviewed, converging from different domains of psychological research. The theory's relation to its predecessors and its unique implications for new research hypotheses are also discussed.

The performance heuristic: a misguided reliance on past success when predicting prospects for improvement.

In estimating whether they are likely to improve on a performance task, people lean on a performance heuristic. That is, people rely on their previous performance success as a positive cue when estimating their prospects for performance improvement. Participants whose initial performance was better--either at a darts game (Study 1) or an anagram task (Study 2)-bet more money (Study 1) or estimated a higher subjective likelihood (Study 2) that their subsequent performance would show a specified amount of improvement. Reliance on the heuristic was unwise, for initial performance did not positively predict (and, in fact, negatively predicted) performance improvement. Study 2 suggests that the performance heuristic emerges because forecasters engage in attribute substitution, naturally focusing on their demonstrated performance instead of whether they have already maxed out their potential for improvement on the task. Self-assessments of their initial performance mediated the performance heuristic, but focusing participants on how much performance potential lay before them disrupted it (Study 2). Study 3 showed that the performance heuristic is a general-purpose heuristic that is used not merely to predict one's own prospects for improvement, but the prospects for other improvement (e.g., mutual funds' rate of return) as well.

Reconsolidation from negative emotional pictures: is successful retrieval required?

Finn and Roediger (Psychological science 22:781-786, 2011) found that when a negative emotional picture was presented immediately after a successful retrieval, later test performance was enhanced as compared to when a neutral picture or a blank screen had been shown. This finding implicates the period immediately following retrieval as playing an important role in determining later retention via reconsolidation. In two new experiments, we investigated whether successful retrieval was required to show the enhancing effect of negative emotion on later recall. In both experiments, the participants studied Swahili-English vocabulary pairs, took an intervening cued-recall test, and were given a final cued-recall test on all items. In Experiment 1, we tested a distinctiveness explanation of the effect. The results showed that neither presentation of a negative picture just prior to successful retrieval nor presentation of a positive picture after successful retrieval produced the enhancing effect that was seen when negative pictures were presented after successful retrieval. In Experiment 2, we tested whether the enhancing effect would occur when a negative picture followed an unsuccessful retrieval attempt with feedback, and a larger enhancement effect occurred after errors of commission than after errors of omission. These results indicate that effort in retrieving is critical to the enhancing effect shown with negative pictures; whether the target is produced by the participant or given by an external source following a commission error does not matter. We interpret these results as support for semantic enrichment as a key element in producing the enhancing effect of negative pictures that are presented after a retrieval attempt.

Buyer's remorse or missed opportunity? Differential regrets for material and experiential purchases.

Previous research has established that experiential purchases tend to yield greater enduring satisfaction than material purchases. The present work suggests that this difference in satisfaction is paralleled by a tendency for material and experiential purchases to differ in the types of regrets they elicit. In 5 studies, we find that people's material purchase decisions are more likely to generate regrets of action (buyer's remorse) and their experiential purchase decisions are more likely to lead to regrets of inaction (missed opportunities). These results were not attributable to differences in the desirability of or satisfaction provided by the two purchase types. Demonstrating the robustness of this effect, we found that focusing participants on the material versus experiential properties of the very same purchase was enough to shift its dominant type of regret. This pattern of regret is driven by the tendency for experiences to be seen as more singular--less interchangeable--than material purchases; interchangeable goods tend to yield regrets of action, whereas singular goods tend to yield regrets of inaction.