PsyArxiv
Not a member yet
49120 research outputs found
Sort by
(In)sensitivity to surface-level heuristics: A case from Turkish verbal attractors
No full textLinguistic illusion literature debates what information accesses memory representations. Prior work tests whether structural, semantic, or discourse cues guide subject-verb dependencies; however, it remains unclear whether native speakers rely on surface level heuristics, such as phonological information during dependency resolution. Traditionally, accidental phonological resemblance to plural ending (e.g., the /s/ in cruise) does not induce erroneous agreement in English, whereas resemblance correlating with controllerhood amplifies attraction across varies languages. Contradicting this generalization, Slioussar (2018) proposed that accidental phonological resemblance can mediate memory search for Russian subjects. Given the theoretical importance of this proposal and the lack of comparable effects in other languages such as Czech, we propose re-interpret previous findings under the light of a recently growing literature of association with being a possible controller. We test whether phonological overlap or association with controllerhood elicits erroneous agreement in Turkish. Turkish provides a critical test: both verbal and nominal elements can surface as subjects and the plural morpheme -lAr marks number in both of them, but only nominal plural -lAr controls verbal agreement. Two speeded acceptability studies show no attraction from plural-marked verbs (N = 80; N = 95) but robust attraction from genitive plural nouns. We report a first-of-its-kind dissociation under minimal manipulation: verbal attractors that can surface as subjects yet cannot control agreement do not induce attraction, whereas genitive plural nouns—which can be subjects and control in other environments—do. This pattern constrains retrieval processes by tying attraction to abstract controller features rather than surface phonology
Are LLMs better expert judges? Rethinking content validity assessment in the age of AI
No full textIn this article, we demonstrate a novel application of large language models (LLMs) as expert judges for item-level content relevance evaluation. Eleven advanced LLMs were included, each treated as a separate expert panel based on multiple procedurally independent judgments generated via repeated API queries. Their performance was compared with ratings provided by human judges, including psychology students and academic experts. Internal agreement within each panel was assessed using Krippendorff’s alpha and Kendall’s coefficient of concordance. Beyond agreement, theoretically predefined control items representing content-nonrelevant material were used to determine the accuracy of each panel’s ratings. Results indicated that several LLM-based panels (Gemini 3 Pro, GPT-5.2 Pro, Claude Sonnet 4.5, and DeepSeek-V3.2) combined near-perfect internal agreement with high accuracy in identifying nonrelevant items, outperforming human panels. Even when individual model instances were treated as independent judges, agreement remained high and all control items were correctly identified. These findings demonstrate that selected LLMs can function as highly consistent content judges, particularly in detecting content-nonrelevant items that are conceptually distant from the measured construct. However, given the lack of empirical evidence on LLMs as expert judges, their ratings should currently be viewed as complementary to human expertise, and further research is required to clarify the conditions under which they can be reliably incorporated into procedures for evaluating test item content relevance
Fluctuations in emotional valence and arousal are differentially linked to cognitive performance and subjective mental effort
No full textBackground: Emotional states vary dynamically, such that their momentary levels and variability are linked to mental health. However, it remains unclear how these dynamics relate to cognitive performance, both objectively and subjectively. We examined whether momentary levels of valence and arousal, and their temporal variability, are associated with objective working memory performance and subjective mental effort.
Methods: Two studies were conducted. Experiment 1 (online cohort, n = 289) validated four state items (happiness, calmness, energy, fatigue) as markers of valence and arousal using a semantic association task and exploratory factor analysis. Experiment 2 was a five-day ecological momentary assessment study (n = 59) with three sessions per day, in which participants reported these items and completed a working memory task, followed by a subjective report of mental effort, and a control motor speed task. Linear mixed models tested associations between emotional dynamics and behavior.
Results: Across both experiments, happiness and calmness loaded highly on a valence factor, whereas energy and fatigue loaded highly on an arousal factor. In Experiment 2, higher momentary arousal, but not momentary valence, was associated with lower perceived mental effort. Greater between-day variability in valence was associated with lower working memory accuracy, driven by increased intrusion errors, but with lower perceived mental effort.
Conclusions: Valence and arousal dynamics are differentially associated with cognition. Daily fluctuations in valence are associated with lower working memory performance, and lower subjective mental effort, possibly reflecting reduced cognitive control. Emotional variability, particularly cross-day changes, may reflect an important target for research on motivation, depression, and self-regulation
How Analytic Approaches Shape Deception Detection Results: A Comparison of Raw Percentage and Signal Detection Metrics
No full textHuman deception detection can be both poor in an absolute sense and, statistically, substantially better than chance. This empirical paradox is documented and explored in the present article by comparing raw percentages and signal detection metrics in a reanalysis of fourteen prior deception detection experiments (total N = 2,349 respondents; 32,776 truth-lie judgments from 5 different countries). We show that different analytic approaches applied to the same data can yield inconsistent or mixed findings. Measures of raw percent-correct accuracy and sensitivity are nearly perfectly correlated yet are open to very different descriptive interpretations. In contrast, raw and signal detection estimates of bias diverge, indicating competing interpretations of human judgment error. We advocate for avoiding simple face-value interpretations of both approaches and embracing multiple analytic approaches simultaneously. A new R package, liaR, is presented to facilitate raw and signal detection calculations in parallel
Harmful Healing: Iatrogenic Fragility in Psychological Safety Interventions
No full textTrigger warnings, content warnings, and related psychological safety interventions have proliferated across educational, digital, and clinical
environments with the explicit goal of protecting vulnerable individuals from emotional harm. Meta-analytic evidence demonstrates these
interventions fail to reduce distress upon content exposure while reliably increasing anticipatory anxiety beforehand. This paper proposes
that safety interventions are not merely ineffective but potentially iatrogenic, causing the psychological fragility they claim to prevent.
I introduce “iatrogenic fragility” as a theoretical framework explaining how well-intentioned protective measures may construct vulnera-
bility through three interacting mechanisms. First, nocebo priming: warnings function as verbal suggestions of negative outcomes, activating
anticipatory anxiety through neurobiological pathways documented in clinical nocebo research. Second, institutionalized avoidance: warn-
ings legitimize safety-seeking behaviors that prevent the inhibitory learning necessary for resilience, functioning as institutional-scale safety
signals that maintain rather than extinguish anxiety. Third, identity consolidation: warnings reinforce trauma-centered identity by implic-
itly communicating that recipients require protection, increasing the centrality of past adversity to self-concept in ways that are directly
countertherapeutic for trauma recovery.
These mechanisms interact to produce a self-reinforcing cycle: safety interventions increase fragility, which generates demand for ex-
panded protections, which further increases fragility. The framework synthesizes findings from clinical psychology, health psychology, and
social cognition that have not previously been integrated. By identifying the causal pathways through which protective intentions produce
harmful outcomes, the iatrogenic fragility framework generates testable predictions and suggests that resilience-focused alternatives may
better serve the populations these interventions were designed to help
Fast uncertainty quantification in EZ cognitive models
No full textClassical approaches to uncertainty quantification in cognitive modeling rely on computationally expensive Monte Carlo methods or resampling of raw trial data, which creates a barrier for real-time analysis and large-scale studies. We present a computationally efficient bootstrap method that operates directly on summary statistics, exploiting the synthetic likelihood structure of a small class of cognitive models that includes the simple diffusion model and the circular diffusion model in addition to signal detection and multinomial processing trees. The method relies on a numerical transformation-of-variables technique in which known sampling distributions of summary statistics of behavior are parametrically bootstrapped, after which the resampled statistics are transformed to parameter estimates with a known analytical system. This approach does not require additional assumptions beyond those already made by the models themselves, but achieves over 1000-fold speed improvements over already efficient fully Bayesian methods. The proposed method makes real-time uncertainty quantification accessible and enables new applications in adaptive testing, meta-analyses, and exploratory data analysis
Reciprocal Effects Between Self-Esteem and Work Experiences: A Reanalysis of Two Longitudinal Studies
No full textDo people’s work experiences (i.e., work conditions and outcomes) influence their self-esteem, and does people’s self-esteem influence their work experiences? In this preregistered research, data used by Kuster et al. (2013) were reanalyzed using new types of statistical models. Whereas the previous study used the traditional cross-lagged panel model, we used the random intercept cross-lagged panel model and the dynamic panel model, which allow for better control of unmeasured time-invariant confounders, enhancing the validity of causal conclusions. Data came from two longitudinal studies with five assessments over eight months (N = 663) and three assessments over two years (N = 600). Thirteen of the 36 cross-lagged paths tested were significant, and all significant effects were in the expected direction: Self-esteem predicted increases in positive outcomes (i.e., coworker justice) and decreases in negative outcomes (e.g., effort-reward imbalance). Positive work variables (e.g., job satisfaction) predicted increases in self-esteem, and negative work variables (e.g., time pressure) predicted decreases in self-esteem. The pattern of findings aligns with theoretical perspectives suggesting reciprocal effects between self-esteem and work experiences. Moderator analyses indicated that effects held across gender. The findings advance the understanding of dynamic self-esteem–work relations and inform interventions that could benefit employees and organizations
Deaf Signers Exhibit White-Matter Alterations in Right-Hemispheric Language Pathways
No full textLanguage processing relies on a distributed network of white-matter pathways connecting frontal, temporal, and parietal regions of cortex. While the arcuate fasciculus in the left hemisphere connecting the posterior inferior frontal gyrus to the posterior temporal cortex is considered universal across languages and modalities (i.e., spoken vs. signed), little is known about how other language-related pathways in both hemispheres adapt in deaf individuals who use a sign language. Here, we used diffusion-weighted MRI to characterize the macro- and microstructural properties of fiber pathways in the core and extended language network in 24 early deaf signers and a control group of 24 hearing non-signers. Automated fiber quantification revealed no macrostructural group differences in left-hemispheric pathways, including the canonical left arcuate fasciculus. In contrast, deaf signers manifested increased streamline counts in the right posterior arcuate fasciculus connecting posterior temporal and parietal regions. Moreover, microstructural group differences were observed in several right-hemispheric tracts, including the right arcuate and right superior longitudinal fasciculus, as well as bilateral uncinate fasciculus. These data suggest that most language-relevant left-hemispheric pathways develop in a similar fashion regardless of the modality of language use, while several right-hemispheric tracts exhibit structural alterations potentially reflecting modality-specific processing demands associated with sign language. In sum, our findings highlight the crucial role that right-hemispheric structures play in processing a visuo-spatial language and, at the same time, indicate that the typical development of left-hemispheric language-relevant pathways does not depend on acquiring and using a spoken language
The association between sleep spindles and cognitive performance in euthymic bipolar disorder
No full textIntro: Many people with bipolar disorder (BD) experience persistent cognitive deficits. Sleep spindles have been linked to cognitive ability in healthy populations and psychotic disorders. While there is preliminary evidence for altered spindle activity in BD, research directly examining the association between sleep spindle parameters and cognitive performance in this population is lacking. Therefore, our primary objective was to examine the association between fast spindle density and episodic memory performance in euthymic individuals with BD, hypothesising a positive relationship. As exploratory analyses, we looked at associations between fast and slow spindle density and subjective sleep quality with other cognitive domains. We also conducted a sensitivity analysis, separating all analyses by lithium intake.
Methods: Thirty-four euthymic participants with BD underwent comprehensive cognitive assessments and were assessed for three consecutive nights using mobile sleep-EEG headbands. Sleep spindles were detected using validated, adapted, fully automated algorithms, subdivided into slow (≤13 Hz) and fast (>13 Hz) spindles, and characterised by density.
Results: Contrary to our hypothesis, fast spindle density was not associated with episodic memory performance (r=-.004, p=.491, 95% CI [-.353, .499]) but did show a significant positive association with working memory in exploratory analyses (r=.423, p=.014, 95% CI [-.167, .722]). When removing participants taking lithium from analyses, several positive significant associations emerged between fast spindle density and cognitive performance.
Conclusions: These findings suggest domain-specific relationships between sleep spindle activity and cognition in BD, with fast spindles potentially being associated with working memory. Preliminary evidence for lithium-related modulation highlights the importance of considering pharmacological factors. However, the analyses were underpowered, and large-scale studies are needed to deepen our understanding of sleep spindle-cognition relationships in BD.
Keywords: bipolar disorder; cognition; episodic memory; working memory; sleep spindles; sleep-EEG; lithiu
Inside Authoritarianism: Heterogeneous RWA Expressions and Their Dynamic Links to COVID-19 Fear and Prevention Beliefs
No full textThe COVID-19 pandemic provides an opportunity to examine how authoritarianism shapes risk perception and preventive norms under sustained uncertainty. Drawing on seven waves of U.S. panel data collected during mid-2020, this study applies dynamic network modeling to investigate reciprocal relations between item-level right-wing authoritarian (RWA) expressions, COVID-19 fear, and prevention beliefs. Results reveal that authoritarianism is best understood as a heterogeneous set of expressed attitudes rather than a unitary trait. Enforcement-oriented authoritarian aggression consistently predicts subsequent reductions in COVID-19 fear, including concern about both personal infection and close others’ illness, suggesting a threat-regulatory function. At the same time, these same expressions predict later declines in support for preventive norms such as restricting outings and social distancing. By contrast, submission-related authoritarian expressions show weaker and more context-dependent associations with fear and prevention beliefs, occasionally aligning with prevention endorsement at the between-subjects level. These findings help reconcile mixed evidence in pandemic research by showing that authoritarian expressions can both respond to perceived threat and subsequently reshape risk perception and compliance-related beliefs. The results underscore the importance of treating authoritarianism as internally differentiated when assessing its role in risk-related cognition and public-health responses
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇