Deniz S. Ones

This research addresses an often-overlooked opportunity for police reform: the predictive value of prehire misbehavior data for reducing posthire police misconduct. While most reform efforts focus on officers' actions after they are hired, our findings highlight the impact of rigorous screening before hire. We examined how specific prior employment and nonwork misbehaviors related to general occupational instability, trouble in previous law enforcement roles, prior temper problems and violence, and irresponsible behaviors predicted future misconduct among 6,075 police officers tracked over 5 years. Notably, some prehire behaviors significantly elevated risks of citizen complaints and misconduct-related lawsuits, with hazard ratios up to 14.59. Contrary to common assumptions, candidates with prior law enforcement experience showed increased liability, including excessive use of force, suggesting that this background does not inherently reduce misconduct risk. After identifying the strongest predictors of police misconduct, we also examined their relation to termination for cause and assessed how agency decision makers respond differently to prehire versus posthire misbehaviors. By integrating findings on the predictive value of specific prehire misbehaviors, we offer targeted, evidence-based guidance and actionable recommendations for police agencies and policymakers. This work provides a scientifically grounded foundation for effective and consistent police screening decisions, offering a framework for establishing long-overdue national police hiring standards. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

El desempeño laboral es el criterio con el que se evalúan todas las intervenciones organizativas. Este artículo sintetiza décadas de investigación. El desempeño laboral tiene una clara estructura jerárquica: un factor general amplio, tres factores principales (desempeño en las tareas, comportamientos de ciudadanía organizativa y evitación de comportamientos laborales contraproducentes) en el segundo nivel, y aspectos específicos en el nivel inferior. El factor general representa una covariación psicológica genuina, más que un artefacto metodológico, que persiste incluso cuando se eliminan el halo y el error de medición mediante rigurosas correcciones psicométricas. El modelo jerárquico da cabida a dimensiones emergentes del desempeño, como el desempeño creativo y adaptativo. Los principios críticos de medición están bien establecidos: las autoevaluaciones adolecen de la evaluación externa que define el constructo del desempeño y los coeficientes de fiabilidad entre evaluadores deben utilizarse en las correcciones por atenuación si se utilizan calificaciones para evaluar el desempeño. Se examinan las complejidades temporales, incluida la distinción entre desempeño máximo y típico, la variabilidad intrapersonal y las trayectorias de desarrollo a través de las etapas profesionales. También se presta atención a los retos contemporáneos: el trabajo a distancia crea dificultades de medición al reducir la posibilidad de observar el comportamiento, mientras que el aumento de la IA plantea cuestiones sobre los límites del desempeño individual cuando la colaboración entre humanos e IA hace que la atribución del comportamiento sea ambigua. Conforme el trabajo evoluciona, el modelo jerárquico proporciona una estructura duradera para entender como los comportamientos individuales se agregan a la contribución de la organización. La estructura jerárquica es un fundamento esencial para una evaluación válida y una eficaz toma de decisiones organizativas. Lan-jarduna da antolakundeetako esku-hartze guztiak ebaluatzeko erabiltzen den irizpidea. Artikulu honetan, hainbat hamarkadatako ikerketak laburbiltzen dira. Lan-jardunaren egitura erabat hierarkikoa da: lehenik, faktore orokor zabal bat du; bigarrenik, hiru talde-faktore nagusi (zereginak betetzea, antolakunde-arloko herritartasunaren portaerak eta laneko portaera kaltegarririk eza), eta, hirugarrenik, behe-beheko mailan, alderdi espezifikoak. Faktore orokorrak benetako kobariazio psikologikoa irudikatzen du ¿hau da, ez da tresna metodologiko huts bat¿, eta bere horretan irauten du, baita halo-efektua eta neurketa-errorea zuzenketa psikometriko zorrotzen bidez ezabatzen direnean ere. Eredu hierarkikoak jardunaren alderdi emergenteak barne hartzen ditu, hala nola sormenezko jarduna eta egokitzapenezko jarduna. Neurketaren printzipio kritikoak ongi ezarrita daude: autoebaluazioetan ez dago jardunaren konstruktua definituko duen kanpo-ebaluaziorik, eta, hortaz, jarduna ebaluatzeko kalifikazioak baliatzen badira, ebaluatzaileen arteko fidagarritasun-koefizienteak erabili behar dira ahultze-zuzenketetan. Orobat, denborazko konplexutasunak aztertu ditugu, baita gehieneko jardunaren eta ohiko jardunaren arteko aldea, pertsonen barne-aldakortasuna eta karreraren etapak garatzeko ibilbideak ere. Gaur egungo erronkei ere heldu diegu: urrutiko lanak neurketak egitea zailtzen du, portaeraren behagarritasuna murrizten duelako; bestalde, adimen artifizialaren gorakadak banako jardunaren mugei buruzko zalantzak planteatzen ditu, zeren eta gizakien eta adimen artifizialaren arteko lankidetzaren ondorioz portaeraren atribuzioa anbiguoa baita. Lana garatzen doan heinean, eredu hierarkikoak esparru iraunkor bat eskaintzen du ulertzeko nola eragiten duten banakoen portaerek antolakunde-ekarpenean. Egitura hierarkikoa ezinbesteko oinarria da, ebaluazioa baliozkoa izan dadin eta antolakundeek erabaki eraginkorrak har ditzaten. Job performance is the criterion against which all organizational interventions are evaluated. This paper synthesizes decades of research. Job performance has a clear hierarchical structure: a large general factor, three major group factors (task performance, organizational citizenship behaviours, and avoidance of counterproductive work behaviours) at the second level, and specific facets at the lowest level. The general factor represents genuine psychological covariation rather than methodological artifact, persisting even when halo and measurement error are removed through rigorous psychometric corrections. The hierarchical model accommodates emerging performance dimensions such as creative and adaptive performance. Critical measurement principles are well established: self-ratings lack the external evaluation that defines the performance construct and interrater reliability coefficients should be used in corrections for attenuation if ratings are used to assess performance. We examine temporal complexities including the maximal-typical performance distinction, within-person variability, and developmental trajectories across career stages. We also attend to contemporary challenges: remote work creates measurement difficulties by reducing behavioural observability, while AI augmentation raises questions about the boundaries of individual performance when human-AI collaboration makes behavioural attribution ambiguous. As work evolves, the hierarchical model provides an enduring framework for understanding how individual behaviours aggregate into organizational contribution. The hierarchical structure is a foundation essential for valid assessment and effective organizational decision-making.

Quiet quitting remains an undertheorized and inconsistently defined construct in the scholarly literature. There is little understanding of or agreement about its construct domain, and no unifying conceptual framework to distinguish it from related constructs. Popular accounts often equate quiet quitting with employee work disengagement. Drawing on a synthesis of academic and practitioner literatures, we conceptualized quiet quitting as a form of psychological withdrawal that extends beyond in‐role task disengagement to include disengagement from colleagues and the organization more broadly. Its key components include a weakening of social connections at work, diminishing job engagement, a generalized disinclination toward work, and the desire to set boundaries between work and personal life. To reflect this multifaceted conceptualization, we developed the Multidimensional Quiet Quitting (MQQ) scale, comprising four interrelated dimensions: (1) lack of job‐specific engagement, (2) lack of generalized work engagement, (3) lack of social connections at work, and (4) boundary setting. Using a content validation sample of working adults ( N = 114), we evaluated the scale's content validity. We then examined its latent structure and internal consistency using experience sampling data from 124 professionals over 10 weeks, followed by confirmatory factor analysis in an independent replication sample ( N = 290). Finally, we mapped the nomological network of quiet quitting by examining associations with theoretically relevant antecedents and outcomes. Results indicated that organizational newcomers and younger employees were slightly more prone to quiet quitting, while job autonomy, work responsibility, marriage, and childcare responsibilities served as protective factors. Quiet quitters also tended to be less extraverted, conscientious, emotionally stable, and agreeable. Quiet quitting and disengagement were each associated with concurrent and subsequent mental health and work outcomes. Quiet quitting components incrementally predicted stress, job satisfaction, burnout, and work effectiveness beyond traditional work disengagement, an overlapping construct. By specifying and measuring quiet quitting as a multidimensional construct, this study provides much‐needed construct clarity and demonstrates empirical effects associated with the phenomenon. Implications for research and practice are discussed.

We engage with the arguments presented by Woods and Patterson (2023) [Woods, S. A., & Patterson, F. (2023). A critical review of the use of cognitive ability testing for selection into graduate and higher professional occupations. Journal of Occupational and Organizational Psychology , 97 , 253–272.], regarding the use of cognitive ability tests in employee selection and academic admissions. We examine assumptions about cognitive ability test validity, fairness and bias and address arguments against the use of cognitive ability tests in five ways. First, we highlight that Woods and Patterson's (2023) propositions overly depend on a single study by Sackett et al. (2022) Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta‐analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range. Journal of Applied Psychology , 107 (11), 2040–2068.], which has faced significant criticisms. Second, we point out that a test with adverse impact can be used if it is job‐related and an unbiased predictor of performance. Third, we question the alleged ‘higher’ objectivity and fairness of alternative non‐cognitive selection procedures. Fourth, we critique overgeneralizations of findings from employment contexts to educational settings and challenge the atheoretical basis of rejecting cognitive ability tests in complex work settings. Fifth, we point out that recent calls for reducing the role of cognitive ability testing in selection do not consider the side effects and costs of those presumed alternatives. We advocate for responsible cognitive ability testing by combining it with valid non‐cognitive predictors to balance validity and adverse impact, using validated tests by trained professionals, ensuring cultural relevance and effectively communicating test processes.

This study investigated the use of deep learning to automate the classification of environmental initiatives in organizational research. We focused on transformer-based models, including BERT and large language models (LLMs), as scalable alternatives to traditional content analysis methods. Two experiments were conducted: the first compared deep learning models with conventional machine learning approaches, while the second evaluated the performance of various LLM prompting configurations. BERT outperformed traditional models, achieving a weighted F1-score of 0.74, while fine-tuned GPT-3.5-turbo surpassed all models with an F1-score of 0.77. In contrast, base versions of GPT models underperformed relative to simpler classifiers. These findings underscored the benefits of fine-tuning for domain-specific tasks and demonstrated that deep learning models can automate text classification without manual feature engineering. The study also identified challenges, including lower performance in rare categories and hallucination issues in some LLM configurations. Overall, this research advanced methodological practices in psychological and organizational science by showcasing how deep learning can strengthen computer-aided content analysis. The results further pointed to the value of hybrid approaches that combine automated methods with expert validation to improve reliability.

Content analysis is a common and flexible technique to quantify and make sense of qualitative data in psychological research. However, the practical implementation of content analysis is extremely labor-intensive and subject to human coder errors. Applying natural language processing (NLP) techniques can help address these limitations. We explain and illustrate these techniques to psychological researchers. For this purpose, we first present a study exploring the creation of psychometrically meaningful predictions of human content codes. Using an existing database of human content codes, we build an NLP algorithm to validly predict those codes, at generally acceptable standards. We then conduct a Monte-Carlo simulation to model how four dataset characteristics (i.e., sample size, unlabeled proportion of cases, classification base rate, and human coder reliability) influence content classification performance. The simulation indicated that the influence of sample size and unlabeled proportion on model classification performance tended to be curvilinear. In addition, base rate and human coder reliability had a strong effect on classification performance. Finally, using these results, we offer practical recommendations to psychologists on the necessary dataset characteristics to achieve valid prediction of content codes to guide researchers on the use of NLP models to replace human coders in content analysis research. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

We critically evaluate Dupré and Wille's (2024) proposal for using assessments for organizational personality development through the lens of empirical evidence on adult personality change. We present an overview of research on personality stability and malleability throughout adulthood examining rank‐order stability, mean‐level changes, and the impact of life events and interventions. Empirical evidence reveals that while personality exhibits some plasticity in young adulthood, significant changes become increasingly rare beyond age 30. For older employees, personality remains highly stable, making age an important consideration in workforce development. Life experiences and intentional interventions have been shown to prompt modest personality changes, with emotional stability being the most malleable trait. We quantify these changes, noting shifts of up to two‐thirds of a standard deviation in emotional stability through targeted interventions, with more limited effects on other Big Five traits. We also provide insights for organizational assessment practices, including the need for tailored personality (re‐)assessment intervals and age‐based norm composition for better utilization of personality information. With Cybernetic Trait Complexes Theory, we introduce a framework for aligning personality traits with situational cues in work environments. This approach emphasizes trait activation rather than personality modification, allowing organizations to harness employees' strengths by strategically designing environments that naturally encourage beneficial trait expression. This shifts the focus from personality change to strategic activation of beneficial traits through environmental design. We describe how organizations can leverage employees' existing personality trait complexes while fostering incremental behavioral adaptations, offering a pragmatic alternative to traditional employee development approaches. By aligning individuals with environments that activate their traits, organizations can enhance both personal and organizational outcomes, contributing to broader societal benefits as well.

Frank L. Schmidt was born on a dairy farm outside Louisville, Kentucky, on April 29, 1944, the oldest of six children to Swiss German parents with a grade-school education. At his first faculty job at Michigan State University, he met John (Jack) Hunter, with whom he began a prolific, impactful collaboration that lasted until Hunter's 2002 death. Together, they invented the methods of psychometric meta-analysis. He believed that the goal of science is to establish universal principles. Schmidt and Hunter's pioneering development of validity generalization (VG) methods showed that statistical artifacts were responsible for study-to-study differences in cognitive ability tests' validities. Schmidt's influential articles included research on selection, bias, utility, job performance, employee engagement, smoking cessation, psychopathology, and corporate social responsibility. But psychometric meta-analysis was his most far-reaching contribution. Schmidt coauthored four widely cited and used books on the technique. Meta-analysis transformed hundreds of fields, where it became the bedrock of scientific knowledge. Schmidt received many prestigious awards for his significant contributions. Schmidt was a paradigm-shifting scientist, a father of modern meta-analytic techniques, and an ardent and intellectually honest researcher of individual differences. He leaves behind a legacy that will continue to shape the future of psychology and management, but also more broadly, science in general. He offered an elegant and quantitative way of knowing. His legacy will live in those whose intellects continue to be shaped by the ideas that he introduced. (PsycInfo Database Record (c) 2023 APA, all rights reserved).