Gender composition predicts gender bias: A meta-reanalysis of hiring discrimination audit experiments

Socioeconomic inequality in mental health disorders: A cross-sectional study from the Tehran University of Medical Sciences employees' cohort study

Understanding socioeconomic inequalities in health helps identify vulnerable groups and guide targeted interventions. Mental health disorders significantly affect well-being and productivity. This study assessed the prevalence and socioeconomic inequalities in depression, anxiety, and stress among employees of the Tehran University of Medical Sciences. This cross-sectional study analyzed data from the Tehran University of Medical Sciences Employees' Cohort (TEC) baseline phase, comprising 4,442 individuals. The Depression, Anxiety, and Stress Scale-42 (DASS-42) was utilized to measure mental health disorders. Education level and wealth index were considered as socioeconomic indicators. The Slope Index of Inequality (SII) and the Relative Index of Inequality (RII) were employed to estimate socioeconomic inequality. The age-adjusted prevalence of depression, anxiety, and stress was 8.7%, 8.6%, and 11.5%, respectively. The relative wealth-related inequality analysis revealed that, after adjusting for age, sex, marital status, and education level, the prevalence of depression, anxiety, and stress in the lowest wealth index was 2.54, 2.89, and 1.65 times higher than in the highest wealth index, respectively. Additionally, the relative education-related inequality analysis indicated that, adjusted for age, sex, marital status, and wealth index, individuals with primary education or no formal education had 2.58, 2.99, and 2.14 times higher prevalence of depression, anxiety, and stress compared to those with a doctoral degree, respectively. Significant disparities in the prevalence of mental health disorders were found across educational and wealth index levels. Targeted interventions and policies should aim to achieve and sustain long-term benefits for vulnerable and disadvantaged groups.

Gender differences in paid work over time: Developments and challenges in comparative research

This paper examines gender differences in paid work over time and illustrates the pitfalls encountered by any comparative research that only considers either labor force participation rates or average working hours. To do so, we analyze harmonized survey data from Europe and the United States from 1992 to 2022 (N = 43,283,172) and show that more progress was made in closing gender gaps in labor force participation rates than in working hours. In most countries, women's labor force participation rates increased considerably, but their average working hours decreased, whereas both men's labor force participation rates and average working hours decreased or stagnated (but nonetheless still remained much higher than women's). We show and argue that these countervailing trends in working hours and labor force participation rates make it difficult to paint a coherent picture of cross-national differences in women's and men's paid work and of changes over time. In response, we propose "work volume" as a supplementary or alternative measure for any type of comparative research. Work volume records zero working hours for nonemployed individuals and thus allows straightforward comparisons between women's and men's (or any other groups') involvement in paid work. Using the proposed work volume measure, we show that gender gaps in paid work decreased over time, but that even in 2022, men's involvement in paid work remained considerably higher than women's-with gender gaps being lowest in the Scandinavian and the former Communist countries.

Shifting landscapes of gender equity in oncology journals: a decade of authorship trends

BACKGROUND

Gender disparities persist in academic oncology, particularly in authorship and senior academic roles. This study evaluates trends in authorship gender representation over the past decade across top oncology journals, focusing on regional, journal-specific, and citation-based disparities.

METHODS

A cross-sectional analysis was conducted on 29,005 articles published between 2014 and 2023 in the top 20 oncology journals, identified through the Web of Science database. Author gender was determined using the NamSor tool. Temporal trends were analyzed using linear regression, and multivariate logistic regression identified factors contributing to gender disparities. Regional and citation analyses explored geographic variations and citation count differences.

RESULTS

Among analyzed articles, 41.81% of first authors and 29.93% of last authors were female. Female first authorship showed a significant upward trend (P < 0.01), with gender parity projected by 2034, while parity for last authors is expected by 2055. Regional differences were notable, with North America and Europe leading in female representation. Certain journals, such as CA: A Cancer Journal for Clinicians and Molecular Cancer, exhibited higher female authorship proportions, while Journal of Clinical Oncology had the lowest. Citation analysis revealed female-authored articles received significantly fewer citations than male-authored ones (P < 0.01).

CONCLUSIONS

Although female authorship in oncology journals has increased over the past decade, disparities remain, particularly in senior roles and citation impact. Addressing these issues requires targeted strategies, including mentorship programs, greater female representation in editorial boards, and institutional policies promoting gender equity.

Workplace Injury and Death: A National Overview of Changing Trends by Sex, United States 1998-2022

Women represent a substantial portion of the US workforce. However, injury and fatality rates for female workers have, historically, remained lower than rates for male workers. Fatal occupational data from the Census of Fatal Occupational Injuries (CFOI) and nonfatal injury data from the National Electronic Injury Surveillance System-Occupational Supplement (NEISS-Work) for the years 1998-2022 were examined to produce rate ratios of male to female fatal and nonfatal occupational injury rates for all workers in the United States. Auto-regressive linear models were developed to analyze rate ratios by sex for fatal and nonfatal occupational injuries by age group, injury event, and select industries to determine if female occupational fatal and nonfatal injury rates were following trends comparable to male rates. Over the 25-year study period, male injury and fatality rates were consistently higher than females. Occupational fatality rates for males were more than nine times higher than female rates, and for nonfatal occupational injuries, male rates were 1.4 times higher than female rates. These analyses indicate that the differences in nonfatal injury rates by sex may be attenuating, however, the large gap by sex in workplace fatalities has remained unchanged. Occupational safety and health research with a more specific focus on these sex differences is needed to gain a clearer understanding of how sex differences affect hiring, job training, task assignment and completion, and injury risk, to identify areas where prevention efforts could be most successful.

Unfit or disliked: How descriptive and prescriptive gender stereotypes lead to discrimination against women

Decades of research attest to the role of gender stereotypes in the emergence of gender-based discrimination. Placing a focus on recent studies, we provide evidence that gender stereotypes continue to negatively affect women's career outcomes in jobs and fields that are seen as male in gender-type. We identify two pathways through which gender stereotypes bring about discrimination: Whereas descriptive gender stereotypes lead to gender discrimination through negative performance expectations produced by lack-of-fit perceptions, prescriptive gender stereotypes lead to gender discrimination through social penalties elicited by perceived stereotype violation. We end by discussing how characteristics of women and those evaluating them may amplify or ameliorate discriminatory behavior, and by considering how organizations and policymakers can leverage research to promote gender equality.

Perceived warmth and competence predict callback rates in meta-analyzed North American labor market experiments

Extensive literature probes labor market discrimination through correspondence studies in which researchers send pairs of resumes to employers, which are closely matched except for social signals such as gender or ethnicity. Upon perceiving these signals, individuals quickly activate associated stereotypes. The Stereotype Content Model (SCM; Fiske 2002) categorizes these stereotypes into two dimensions: warmth and competence. Our research integrates findings from correspondence studies with theories of social psychology, asking: Can discrimination between social groups, measured through employer callback disparities, be predicted by warmth and competence perceptions of social signals? We collect callback rates from 21 published correspondence studies, varying for 592 social signals. On those social signals, we collected warmth and competence perceptions from an independent group of online raters. We found that social perception predicts callback disparities for studies varying race and gender, which are indirectly signaled by names on these resumes. Yet, for studies adjusting other categories like sexuality and disability, the influence of social perception on callbacks is inconsistent. For instance, a more favorable perception of signals like parenthood does not consistently lead to increased callbacks, underscoring the necessity for further research. Our research offers pivotal strategies to address labor market discrimination in practice. Leveraging the warmth and competence framework allows for the predictive identification of bias against specific groups without extensive correspondence studies. By distilling hiring discrimination into these two dimensions, we not only facilitate the development of decision support systems for hiring managers but also equip computer scientists with a foundational framework for debiasing Large Language Models and other methods that are increasingly employed in hiring processes.

Bias of AI-generated content: an examination of news produced by large language models

Large language models (LLMs) have the potential to transform our lives and work through the content they generate, known as AI-Generated Content (AIGC). To harness this transformation, we need to understand the limitations of LLMs. Here, we investigate the bias of AIGC produced by seven representative LLMs, including ChatGPT and LLaMA. We collect news articles from The New York Times and Reuters, both known for their dedication to provide unbiased news. We then apply each examined LLM to generate news content with headlines of these news articles as prompts, and evaluate the gender and racial biases of the AIGC produced by the LLM by comparing the AIGC and the original news articles. We further analyze the gender bias of each LLM under biased prompts by adding gender-biased messages to prompts constructed from these news headlines. Our study reveals that the AIGC produced by each examined LLM demonstrates substantial gender and racial biases. Moreover, the AIGC generated by each LLM exhibits notable discrimination against females and individuals of the Black race. Among the LLMs, the AIGC generated by ChatGPT demonstrates the lowest level of bias, and ChatGPT is the sole model capable of declining content generation when provided with biased prompts.

Gender equity in hiring: examining the effectiveness of a personality-based algorithm

Introduction

Gender biases in hiring decisions remain an issue in the workplace. Also, current gender balancing techniques are scientifically poorly supported and lead to undesirable results, sometimes even contributing to activating stereotypes. While hiring algorithms could bring a solution, they are still often regarded as tools amplifying human prejudices. In this sense, talent specialists tend to prefer recommendations from experts, while candidates question the fairness of such tools, in particular, due to a lack of information and control over the standardized assessment. However, there is evidence that building algorithms based on data that is gender-blind, like personality - which has been shown to be mostly similar between genders, and is also predictive of performance, could help in reducing gender biases in hiring. The goal of this study was, therefore, to test the adverse impact of a personality-based algorithm across a large array of occupations.

Method

The study analyzed 208 predictive models designed for 18 employers. These models were tested on a global sample of 273,293 potential candidates for each respective role.

Results

Mean weighted impact ratios of 0.91 (Female-Male) and 0.90 (Male-Female) were observed. We found similar results when analyzing impact ratios for 21 different job categories.

Discussion

Our results suggest that personality-based algorithms could help organizations screen candidates in the early stages of the selection process while mitigating the risks of gender discrimination.