Michael A. Campion’s research while affiliated with Purdue University West Lafayette and other places

We identify the central role of toxic team climate in understanding dysfunctional work teams. We develop and test a model identifying team-level psychopathic personality and teamwork knowledge as proximal contributors to the creation of toxic team climates, which mediates the effects to create dysfunctional team outcomes. We also tested an alternative model that treats teamwork knowledge as a moderator of the relationship between team psychopathy and toxic team climate. Using a four-wave longitudinal design, we evaluated these models in a sample of 508 business students comprising 107 student teams (35 MBA and 72 undergraduate) completing semester-long team projects. Results indicate that (1) teams with relatively higher levels of psychopathy and lower levels of teamwork knowledge were more likely to have toxic team climates, (2) teams with more toxic climates had higher levels of social loafing and interpersonal deviance, and lower team performance and satisfaction, and (3) the effects of psychopathy on dysfunctional outcomes were mediated by toxic team climate, but the effects of teamwork knowledge were not. In addition, teamwork knowledge moderated the relationship between team psychopathy and toxic team climate by reducing its effects. The study yields implications for the management of toxic work teams.

As more organizations move to remote hiring assessments, important questions emerge as to the effects on scores, racioethnic, and gender subgroup differences, and candidate reactions. We compare scores of candidates assessed remotely under proctored conditions ( N = 902) versus onsite ( N = 891) in an actual selection context in the same organization, in the same time period, and on the same cognitive ability tests, case exercises, and structured interviews. Controlling for job, there were no differences for cognitive ability tests or case exercises in the remote environment, but higher scores for structured interviews, leading to a slightly higher total score for all assessments combined and a 5% increase in the overall passing rate. Within groups, Hispanic or Latino candidates performed better on the remote cognitive ability test compared with Hispanic or Latino candidates onsite, while Asian candidates performed better remotely for the case exercise. All subgroups performed better on the remote structured interview compared with their onsite counterparts. No between‐group differences emerged by racioethnicity, but women outperformed men on the remote cognitive ability test compared to onsite. Candidate fairness reactions did not differ by test environment for any assessments or subgroups. We conclude that: (1) remote proctored assessments will not create lower overall passing rates (i.e., fewer candidates for hire); (2) differences in remote assessment scores may depend on the type of assessment, with the greatest positive differences for structured interviews; (3) remote assessments do not disadvantage racioethnic minority candidates or candidates overall; and (4) remote assessments do not reduce candidate fairness reactions.

Human resources (HR) managers struggle to manage the adverse impact‐validity tradeoff where some of the most predictive and affordable hiring procedures, such as mental ability tests, often result in lower hiring rates for racioethnic minority subgroups of candidates, thus creating legal risks due to anti‐discrimination laws. In this study, we examine whether employer‐sponsored practice testing will reduce subgroup differences in test performance by offering an equalizing preparation opportunity framework based on the tenant of access , including access to information and opportunities to perform. In a large diverse sample in an operational selection context ( N practice test = 29,626; N actual test = 18,408; N both = 5078), we found that candidates who took the practice test scored higher on the actual tests than those who did not. All candidates benefitted by receiving an accurate estimate of passing the actual test and increasing their likelihood of applying. Further, racioethnic minorities realized greater score gains than racial non‐minorities, thereby reducing subgroup mean differences and subsequent adverse impact. The results were supportive for all major racioethnic minority subgroups (Asians, Blacks or African Americans, and Hispanics), and effect sizes were meaningfully large. Finally, we examined differences in other preparation tactics (e.g., gaining additional experience, using study guides) and found that racioethnic minorities were likely to use some tactics more than racioethnic non‐minorities, but not the most predictive tactics. We conclude that HR managers should consider equalizing preparation opportunities, specifically practice testing, to help address the adverse impact‐validity tradeoff by reducing impact without abandoning valid employment tests.

This study integrates identity and situational strength theories into a realistic accuracy model framework to develop and test hypotheses regarding how hiring managers perceive social media (SM) information, and how those perceptions translate into ratings. In Study 1, 310 h professionals rated which SM content they perceive as most useful in assessing applicants on various work‐related constructs. The provided a foundation for Study 2, where 151 h professionals participated in an experimental hiring simulation to understand whether these perceptions of SM information translate into rating behavior. Results indicate that manipulating SM content influenced HR professionals’ ratings of applicants. Non‐job‐related SM content influenced ratings of work‐related constructs more strongly than job‐related content. Professional SM was rated more favorably than personal SM.

The purpose of this research is to demonstrate how using natural language processing (NLP) on narrative application data can improve prediction and reduce racial subgroup differences in scores used for selection decisions compared to mental ability test scores and numeric application data. We posit there is uncaptured and job-related constructs that can be gleaned from applicant text data using NLP. We test our hypotheses in an operational context across four samples (total N = 1,828) to predict selection into Officer Training School in the U.S. Air Force. Boards of three senior officers make selection decisions using a highly structured rating process based on mental ability tests, numeric application information (e.g., number of past jobs, college grades), and narrative application information (e.g., past job duties, achievements, interests, statements of objectives). Results showed that NLP scores of the narrative application generally (a) predict Board scores when combined with test scores and numeric application information at a level of correlation equivalent to the correlation between human raters (.60), (b) add incremental prediction of Board scores beyond mental ability tests and numeric application information, and (c) reduce subgroup differences between racial minorities and nonracial minorities in Board scores compared to mental ability tests and numeric application information. Moreover, NLP scores predict (a) job (training) performance, (b) job (training) performance beyond mental ability tests and numeric application information, and (c) even job (training) performance beyond Board scores. Scoring of narrative application data using NLP shows promise in addressing the validity-adverse impact dilemma in selection.

Machine learning (ML) may be the biggest innovative force in personnel selection since the invention of employment tests. As such, the purpose of this special issue was to draw out research from applied settings to supplement the work that appeared in academic journals. In this overview article, we aim to complement the special issue in five ways: (1) provide a brief tutorial on some ML concepts and illustrate the potential applications in selection, along with their strengths and weaknesses; (2) summarize findings of the four articles in the special issue and provide an independent appraisal of the strength of the evidence; (3) identify some of the less‐obvious lessons learned and other insights that researchers new to ML might not clearly recognize from reading the special issue; (4) present best practices at this stage of the knowledge in selection; and (5) propose recommendations for future needed research based on the articles in the special issue and the current state of the science.