scholarly journals Double-blind reviewing and gender biases at EvoLang conferences: An update

Author(s):  
Christine Cuskley ◽  
Seán G Roberts ◽  
Stephen Politzer-Ahles ◽  
Tessa Verhoef

Abstract A previous study of reviewing at the Evolution of Language conferences found effects that suggested that gender bias against female authors was alleviated under double-blind review at EvoLang 11. We update this analysis in two specific ways. First, we add data from the most recent EvoLang 12 conference, providing a comprehensive picture of the conference over five iterations. Like EvoLang 11, EvoLang 12 used double-blind review, but EvoLang 12 showed no significant difference in review scores between genders. We discuss potential explanations for why there was a strong effect in EvoLang 11, which is largely absent in EvoLang 12. These include testing whether readability differs between genders, though we find no evidence to support this. Although gender differences seem to have declined for EvoLang 12, we suggest that double-blind review provides a more equitable evaluation process.

PeerJ ◽  
2019 ◽  
Vol 7 ◽  
pp. e6702 ◽  
Author(s):  
Amelia R. Cox ◽  
Robert Montgomerie

To date, the majority of authors on scientific publications have been men. While much of this gender bias can be explained by historic sexism and discrimination, there is concern that women may still be disadvantaged by the peer review process if reviewers’ biases lead them to reject publications with female authors more often. One potential solution to this perceived gender bias in the reviewing process is for journals to adopt double-blind reviews whereby neither the authors nor the reviewers are aware of each other’s identity and gender. To test the efficacy of double-blind reviews in one behavioral ecology journal (Behavioral Ecology, BE), we assigned gender to every authorship of every paper published for 2010–2018 in that journal compared to four other journals with single-blind reviews but similar subject matter and impact factors. While female authorships comprised only 35% of the total in all journals, the double-blind journal (BE) did not have more female authorships than its single-blind counterparts. Interestingly, the incidence of female authorship is higher at behavioral ecology journals (BE and Behavioral Ecology and Sociobiology) than in the ornithology journals (Auk, Condor, Ibis) for papers on all topics as well as those on birds. These analyses suggest that double-blind review does not currently increase the incidence of female authorship in the journals studied here. We conclude, at least for these journals, that double-blind review no longer benefits female authors and we discuss the pros and cons of the double-blind reviewing process based on our findings.


2018 ◽  
Author(s):  
Amelia R Cox ◽  
Robert Montgomerie

To date, the majority of authors on scientific publications have been men. While much of this gender bias can be explained by historic sexism and discrimination, there is concern that women may still be disadvantaged by the peer review process if reviewers' unconscious biases lead them to reject publications with female authors more often. One potential solution to this perceived gender bias in the reviewing process is for journals to adopt double-blind reviews whereby neither the authors nor the reviewers are aware of each other's identities and genders. To test the efficacy of double-blind reviews, we assigned gender to every authorship of every paper published in 5 different journals with different peer review processes (double-blind vs. single blind) and subject matter (birds vs. behavioral ecology) from 2010-2018 (n = 4865 papers). While female authorships comprised only 35% of the total, the double-blind journal Behavioral Ecology did not have more female authorships than its single-blind counterparts. Interestingly, the incidence of female authorship is higher at behavioral ecology journals (Behavioral Ecology and Behavioral Ecology and Sociobiology) than in the ornithology journals (Auk, Condor, Ibis), for papers on all topics as well as those on birds. These analyses suggest that double-blind review does not currently increase the incidence of female authorship in the journals studied here. We conclude, at least for these journals, that double-blind review does not benefit female authors and may, in the long run, be detrimental.


Author(s):  
Manjul Gupta ◽  
Carlos M. Parra ◽  
Denis Dennehy

AbstractOne realm of AI, recommender systems have attracted significant research attention due to concerns about its devastating effects to society’s most vulnerable and marginalised communities. Both media press and academic literature provide compelling evidence that AI-based recommendations help to perpetuate and exacerbate racial and gender biases. Yet, there is limited knowledge about the extent to which individuals might question AI-based recommendations when perceived as biased. To address this gap in knowledge, we investigate the effects of espoused national cultural values on AI questionability, by examining how individuals might question AI-based recommendations due to perceived racial or gender bias. Data collected from 387 survey respondents in the United States indicate that individuals with espoused national cultural values associated to collectivism, masculinity and uncertainty avoidance are more likely to question biased AI-based recommendations. This study advances understanding of how cultural values affect AI questionability due to perceived bias and it contributes to current academic discourse about the need to hold AI accountable.


2019 ◽  
Author(s):  
Iris C.I. Chao ◽  
Efrem Violato ◽  
Brendan Concannon ◽  
Charlotte McCartan ◽  
Katarzyna Nicpon ◽  
...  

Abstract Background: Several forms of bias, including ethnic and gender bias, are thought to impact evaluations on Clinical Performance Assessments (CPAs). Unfairness may influence student learning attitudes if a loss of trust causes a lack of engagement in learning. Understanding the biases occurring in CPAs can lead to well-designed examiner training to ensure equality and fairness. The purpose of this systematic review is to determine the current evidence in the literature for ethnic and/or gender bias by examiners evaluating pre-licensure healthcare students in CPAs using standardized patients (SPs). Methods: Literature was systematically searched in CINAHL, PubMed and Medline from inception to February 2019, and no date range was set. Studies related to the investigation of ethnic and/or gender biases occurring in CPAs using SPs for examining health professions students were selected. A systematic review was conducted to assess the methodological quality and strength of evidence of relevant research and to identify if any potential ethnic and/or gender bias occurred in CPAs. The Guidelines for Critical Review were used to appraise the selected studies. Results: Nine studies published from 2003 to 2017 were retrieved for review. Three studies met all the Guidelines for Critical Review quality criteria, indicating stronger evidence of their outcomes, two of the studies reported ethnic and/or gender bias existing in the CPAs. Overall, four studies found ethnic and/or gender bias in CPAs, but all study results had small effect sizes. Conclusions: No systematic and consistent bias was found across the studies; nonetheless, the possibility of ethnic or gender bias by some examiners cannot be ignored. To minimize potential examiner bias, the investigation of Frame of Reference training, multiple examiners per station, and combination assessments in CPAs is recommended.


2008 ◽  
Vol 6 (7) ◽  
pp. 354-354 ◽  
Author(s):  
Katrin Hammerschmidt ◽  
Klaus Reinhardt ◽  
Jens Rolff

BMJ ◽  
2019 ◽  
pp. l6721
Author(s):  
John A Emelifeonwu ◽  
James E Hazelwood ◽  
Oscar Nolan ◽  
Emma Sharland ◽  
Anna O’Donald ◽  
...  

AbstractObjectivesTo compare the proportional representation of healthcare workers in receipt of New Year honours (NYHs) with workers in other industries and to determine whether the NYH system has gender or geographical biases.DesignObservational study of the UK honours system with a comparative analysis of proportional representation of the UK workforce and subgroup analyses of gender and geographical representations.ParticipantsRecipients of NYHs from 2009 to 2018.Main outcome measuresAbsolute risk of receiving an NYH based on industry, gender, or region of the UK. Relative risk of receiving an NYH for services to healthcare compared with other industries.Results10 989 NYHs were bestowed from 2009 to 2018, 47% of which were awarded to women. 832 awards (7.6%) were for services to healthcare. People working in sport and in the arts and media were more likely to receive NYHs than those working in healthcare (relative risks of 22.01 (95% confidence interval 19.91 to 24.34) and 5.84 (5.31 to 6.44), respectively). There was no significant difference between the rate of receiving honours for healthcare and for science and technology (P=0.22). 34% (3741) of awards were issued to people living in London and in the southeast of England, and only 496 of 1447 (34%) higher order awards (knighthoods, damehoods, companions of honour, and commanders of the order of the British empire) were received by women.ConclusionsIn relation to the size of its workforce, a career in healthcare is not as “honourable” as careers in certain other industries. Geographical and gender biases might exist in the honours system.


2018 ◽  
Vol 42 (3) ◽  
pp. 343-354 ◽  
Author(s):  
Mike Thelwall

Purpose The purpose of this paper is to investigate whether machine learning induces gender biases in the sense of results that are more accurate for male authors or for female authors. It also investigates whether training separate male and female variants could improve the accuracy of machine learning for sentiment analysis. Design/methodology/approach This paper uses ratings-balanced sets of reviews of restaurants and hotels (3 sets) to train algorithms with and without gender selection. Findings Accuracy is higher on female-authored reviews than on male-authored reviews for all data sets, so applications of sentiment analysis using mixed gender data sets will over represent the opinions of women. Training on same gender data improves performance less than having additional data from both genders. Practical implications End users of sentiment analysis should be aware that its small gender biases can affect the conclusions drawn from it and apply correction factors when necessary. Users of systems that incorporate sentiment analysis should be aware that performance will vary by author gender. Developers do not need to create gender-specific algorithms unless they have more training data than their system can cope with. Originality/value This is the first demonstration of gender bias in machine learning sentiment analysis.


2019 ◽  
Vol 20 (3) ◽  
pp. 203-225 ◽  
Author(s):  
Sarah Cleeland Knight

Abstract A perennial critique of international relations is that the field focuses disproportionately on the United States and Europe and contains a gender bias in terms of ignoring issues of particular concern to women. The field is also infamous for how difficult it is for female scholars to publish and have their publications cited. This study evaluates these claims of bias in the area of undergraduate international relations teaching by analyzing an original dataset of 48 introduction to international relations syllabi from ten countries. The study analyzes the authors of required readings and the theories and empirical topics taught, and finds that the geographic and gender biases are both firmly in place. The first finding is that courses assign readings predominantly from US-resident, US-trained, male authors, even those courses taught outside the United States and those taught by female faculty. A second finding is that assigned readings focus overwhelmingly on the United States more than any other country or region, and only 1 percent of readings focus specifically on gender-related issues.


2019 ◽  
Vol 95 (5) ◽  
pp. 117-147
Author(s):  
Henry L. Friedman

ABSTRACT This study examines whether investor-level preferences for director characteristics influence portfolio choices, using data on the U.S. holdings of non-U.S. funds. Consistent with bias-based preferences influencing portfolio allocations, funds from countries with greater gender inequality invest less and hold smaller stakes in firms with more female directors. Since variation in funds' home country gender biases are plausibly unrelated to the selection and performance of female directors in U.S. firms, the empirical strategy mitigates endogeneity concerns arising from estimates based on associations between market performance and gender demographics. The study contributes by linking investments to measured gender biases and by providing evidence, through additional analysis, of potential channels through which gender bias may affect portfolio choice. JEL Classifications: G11; J16; M10.


2020 ◽  
Vol 31 (2) ◽  
pp. 206-231
Author(s):  
Sina Otten ◽  
Dorothea Alewell

We analyze the effects of deviation from gender stereotypes on job satisfaction for male and female employees in general and for employees in leadership positions. Based on social role theory, backlash mechanisms owing to the violation of gender norms and role incongruity theory, we expect that deviating from gender stereotypes negatively affects job satisfaction. We test our hypotheses by hierarchically applying multiple linear regressions to German employee data. Results show a stable negative effect of deviation from gender stereotypes on job satisfaction for women only. Our findings are consistent with recent studies that confirm traditional gender structures on the labor market and expand our knowledge about backlash effects, since they indicate that deviation from gender norms not only affects objective career indicators but also subjective ones. As job satisfaction is a predictor of organizational success, we discuss ways for organizations to reduce the harmful effects of persistent traditional gender stereotypes in workplaces.


Sign in / Sign up

Export Citation Format

Share Document