Emergency Medicine

Gender coding in job advertisements for academic, non-academic, and leadership positions in emergency medicine

a b s t r a c t

Objectives: Gender disparities continue to exist in emergency medicine (EM) despite increasing percentages of women in medical school and residencies. Prior studies in other male dominated industries have shown using masculine or feminine-coded language in job advertisements affects the proportion of male versus female appli- cants who choose to apply for those jobs. The goal of this study was to determine if gender-coding exists in EM job advertisements, and to see if there were differences between academic vs. non-academic jobs or administra- tive vs. non-administrative jobs.

Methods: This was a Cross sectional study of EM jobs advertised in the United States on 13 academic and non- academic medical job databases from September 2020-February 2021. Using a gender decoder program based on prior research by Gaucher et al. on gendered wording in job advertisements, we analyzed each job to deter- mine if the job advertisement was overall highly masculine, masculine, highly feminine, feminine, or neutral. Each job was categorized as academic, non-academic, administrative, or non-administrative. Data were analyzed using descriptive statistics and chi-square analysis.

Results: Seventy-four EM job advertisements were posted during the study period. Forty-four (59.4%) of these coded out as masculine or strongly masculine, 18 (24.3%) coded out as feminine or strongly feminine, and 12 (16.2%) were neutral. Only one job advertisement contained no gender-coded words. There were no differences in the gender-coding of academic, non-academic, or administrative jobs.

Conclusion: Job advertisements for EM physicians tend to contain more masculine-coded language. Almost all job advertisements for emergency medicine physicians in this study contained at least one gender-coded word. Further studies could explore whether changing the language of job advertisements in EM has an impact on the proportion of women who choose to apply to EM jobs.

(C) 2022

  1. Introduction

Gender disparities continue to exist in many specialties in medicine. According to a report from the American Association of Medical Colleges (AAMC) entitled the State of Women in Medicine in 2018-19, the per- centage of women graduating from medical school has steadily in- creased over the past 40 years [1]. In 2018, women made up 48% of medical school graduates and 46% of all residents [1]. Women, however, have a high rate of attrition after medical school, and have differential representation in a variety of specialties [2]. As a result, there are still

* Corresponding author.

E-mail addresses: Kelly.O’[email protected] (K. O’Brien), [email protected] (V. Petra), [email protected] (D. Lal), [email protected] (K. Kwai),

[email protected] (M. McDonald), [email protected] (R. Jeanmonod).

large gaps in the percentage of women in many fields of medicine in- cluding emergency medicine (EM). In 2018, only 36% of EM residents were female [1]. In the physician specialty data report published by the AAMC for 2019, 71.7% of EM attending physicians are male while only 28.3% are female [3].

This is important because diversity, inclusion, and equity in health- care can improve patient care. Patients are often more comfortable interacting with physicians who are more similar to them in culture, race, and ethnicity [4,5]. Female patients, who have higher ED utilization as compared to men, have more trust in female physicians and rank them higher on measures of showing concern, time spent with them, and overall, as compared to male physicians [6,7]. By increasing diver- sity of the workforce to be more representative of the patient popula- tion, patients are more likely to have a better experience and better compliance with medical therapy [3]. Additionally, female physicians

https://doi.org/10.1016/j.ajem.2022.02.023

0735-6757/(C) 2022

have improved patient related outcomes in Medicare populations [8], and diversity and inclusion have been shown to be financially beneficial for companies, improving their reputations and assisting healthcare or- ganizations recruit and retain top physicians [9,10].

Several studies have acknowledged this gender gap, and have made recommendations for increasing the recruitment and retention of women in medicine. These studies advocate for increasing awareness of gender bias in the workplace [11,12]. Although these recommenda- tions promote recognition of overt discrimination, unconscious bias still affects recruitment of women into EM [13]. Institutional climate and lack of diversity in selection committees can introduce unconscious bias leading to fewer women being recruited into jobs or being ad- vanced in Leadership roles [13-15].

An understudied area in which unconscious bias may play a role in female recruitment into EM is language used in advertisements and promotional materials for professional positions. Gender-coded lan- guage, which is the usage of words that are typically associated with men or women, may arise in advertising as a covert and subtle way of perpetuating gender inequality within a given organization, or it may arise in a more organic way through social role theory, where histori- cally male-dominated fields become associated with traits commonly assigned to the prevailing gender within that field [16]. Men have tradi- tionally been associated with traits such as leadership, strength, and dominance whereas women are perceived as communal, nurturing, and interpersonally oriented [16,17]. Prior research on language and gender wording has shown that job descriptions containing more masculine-coded words make readers believe there is a higher male representation in that role [18]. One study by Gaucher, et al. looked at gender-coding of words in job advertisements and found that there were a higher percentage of masculine words in advertisements for jobs that are traditionally male dominated [16]. This study also demon- strated that women found job advertisements with feminine-coding more appealing and promoting higher levels of belongingness com- pared to masculine-coded advertisements [16]. Continuing to use gen- dered words in job advertisements in male dominated fields may discourage women from applying to these jobs and strengthen the be- lief that men are more fit for those jobs [16,17].

The objective of this study is to determine if there is consistent gender-coded language in advertisements for jobs in EM. Additionally, we sought to determine if there is a difference in gender-coding between academic vs non-academic or administrative vs non- administrative job advertisements.

  1. Methods
    1. Study design

This was a cross sectional study of EM physician job advertisements in the United States posted from September 2020 through February 2021. The study was reviewed by the IRB and found to be exempt.

    1. Database selection

Job databases were chosen that were affiliated with professional or- ganizations in the specialty of EM. Jobs in databases that were discover- able without password protection were also included. Databases were specifically chosen to represent both academic and non-academic jobs. Databases used in this study include careerconnect.aamc.org, careers. saem.org, cordjobboard.com, careers.acponline.org, practicematch.com, academic-med.com, kbic.com, merritthawkins.com, careers.jamanetwork. com, doccafe.com, healthcareers.com, practicelink.com, and physicianhire. com. Databases that were primarily affiliated with corporate medical groups (CMGs) were not included, as these websites contain dozens of jobs all of which are advertised from a template, with identical descrip- tions. We did not want to include upwards of 100 jobs in our analysis which were really representative of a single job advertisement.

    1. Job inclusion and exclusion criteria

All EM physician job advertisements in the United States listed on the included databases from September 2020 to February 2021 were in- cluded. Although databases affiliated with CMGs were not included, CMG jobs were NOT excluded if they appeared on the above listed data- bases. Job advertisements were excluded if they were recruiting for ad- vanced practitioners, nurse practitioners, or physician assistants.

    1. Research protocol

All included databases were searched for EM physician jobs. Search terms (for sites that were not exclusively EM sites) included “emer- gency medicine,” “physician,” and “academic.” Job advertisements were reviewed in the order in which they were listed on the website. All jobs were entered into a single Excel database, and reviewed for re- peat entries, as many jobs are advertised on multiple websites.

There is an online gender decoder that is based on Gaucher’s study of gendered language in job advertisements (https://gender-decoder. katmatfield.com/). This decoder recognizes gendered words and clas- sifies text as “strongly masculine,” “masculine,” “neutral,” “feminine,” and “strongly feminine.” Job advertisements were cut and pasted verba- tim from the websites into this decoder. A priori, we determined that words which included feminine roots but did not have the same mean- ing (for instance, “committee,” which has the root “commit,” or “Con- necticut,” which has the root “connect”) and words such as “children” or “community” when used as proper nouns (for instance, “Community Hospital”) would not accurately reflect the gender-coding of the adver- tisement. These words were manually removed prior to advertisement analysis. The gender decoder assigned each job advertisement as overall highly masculine, masculine, highly feminine, feminine or neutral. The decoder accomplishes this by recognizing the total number of gendered words and comparing them. Therefore, an advertisement with an equal number of masculine and feminine words would code out as neutral. An advertisement with an imbalance (for instance, one masculine and zero feminine words) would code out as masculine. When there are in excess of 4 or more gender-coded words, the modifier “highly” is applied. Since neutral advertisements may be balanced or may simply contain no gender-coded words, we also recorded which individual words coded out as masculine or feminine in each advertisement.

Each job was categorized as academic or non-academic and admin- istrative or non-administrative. Academic jobs were defined as core fac- ulty positions at primary or secondary residency training sites or medical school affiliates where teaching and academic productivity are core components of the position. Administrative positions were de- fined as director, assistant director, associate director, chair, vice chair, or chief positions. The location of the job was recorded in terms of the region of the country that the job was in (Northeast, Southeast, South- west, Midwest or West). We also recorded if there was a subspecialty or fellowship requirement for the job.

    1. Investigator training

Each member of the research team was provided written instruc- tions for data collection, and these instructions were reviewed in per- son. These included the step-by-step approach on how to search for jobs, how to use the gender decoder, and how to enter data from the job advertisements into the data collection spreadsheet.

    1. Data handling and analysis

Data were entered into an Excel spreadsheet and were analyzed using descriptive statistics, Chi square, and Fisher Exact, using MedCalc ((C)1993-2013, Ostend, Belgium) and VassarStats.net ((C) Richard Lowry 1998-2021). All interpretations and analyses were performed in dupli- cate by two study personnel, with identical results.

  1. Results

We found 74 unique job advertisements for EM physicians that were listed on the study websites during the study period. Jobs were fairly equally divided into academic and nonacademic positions, and they represented all areas of the country, although there were only 6 jobs ad- vertised in the West. Further details regarding characteristics of listed jobs can be found in Table 1.

Gender-coding was common in EM job advertisements, with almost 60% coded as masculine or highly masculine. Twenty-five percent of jobs coded as feminine or highly feminine. Sixteen percent of jobs

Distribution of gender-coding in EM job advertisements

40

35

30

Number of jobs

25

20

15

10

5

coded as neutral on the whole. Breakdown of gender-coding for all an- alyzed jobs is shown in Fig. 1.

Although our N is small, there were no statistically significant differ- ences in the presence of gender-coded language in job advertisements

0

Strongly masculine

Masculine Neutral Feminine Strongly feminine

Gender-coding

when comparing different geographical regions (p = 0.45), academic vs non-academic jobs (p = 0.75), or administrative vs. non- administrative positions (p = 0.59).

Although 16.2% of jobs coded as neutral (n = 12), these jobs largely contained a balanced number of masculine- and feminine-coded words, rather than an absence of any coded words. There was only a single job advertisement that did not contain any gender-coded words. Ninety-six percent (n = 71) of advertisements contained at least one masculine- coded word, and 82.4% (n = 61) contained at least one feminine- coded word. The most commonly used masculine-coded words were derivatives of the words “leader,” “compete,” and “challenge.” The most commonly used feminine-coded words were derivatives of the words “commit,” “collaborate,” and “compassion.” Other commonly used gender-coded words are shown in Table 2.

  1. Discussion

Overall, our study found that EM job advertisements contain more masculine-coded language compared to feminine-coded language, which is consistent with studies in other male-dominated fields [16]. The degree to which this contributes to the disparities between male and female representation in EM is less clear.

There is an obvious drop-off in female representation in medicine at every milestone, from medical school to residency to clinical and aca- demic positions and rank, to leadership positions. Personal factors for the women themselves likely contribute to this phenomenon. In re- search on systems justification in political and business spheres, people tend to defend existing inequality in representation as “natural and preferable,” incorporating the status quo into their belief systems, and using the status quo to justify the inequality [19,20]. Women feel inter- nal and external pressure to make life choices consistent with current societal standards regarding nurturing, family, and priorities. Over time, this becomes part of their belief systems and is used to justify in- equalities. As inequality is perpetuated, sex- and gender-related

Fig. 1. Distribution of gender-coding in EM job advertisements.

stereotypes and biases are strengthened. This may affect not only the likelihood of hiring entities to consider women for positions, but also the likelihood of women considering specific positions as appropriate for themselves.

Beyond personal factors, institutional (or systemic) factors likely play a role in Sex disparities in medical fields. In social dominance theory, stable group-based dominance (for example, hierarchical structures around caste, religion, sex, race, or gender identity) is main- tained through reinforced ideology (in education, policy, and media), conscious and unconscious discrimination, and systemic bias that deters deviation from the model [21]. The dominant group is able to maintain power and prestige with little expended effort or personal intent, as the structure of the system itself is designed to insure perpetuation of power. Therefore, even if women do not personally endorse the status quo, it is challenging for them to move into traditionally male domi- nated arenas due to system level bias, whether that is at the level of ed- ucation or the level of job entry.

Advertisements for EM physician jobs reflect the way institutions choose to present themselves to recruit potential applicants. Although these advertisements may not represent conscious and intentional bias on the part of the institutions looking for physicians, they may dem- onstrate unconscious, subtle, or systemic bias that ultimately has poten- tial to shape the diversity of their workforce. The studies by Gaucher et al. demonstrated that job descriptions with masculine language make women feel less of a sense of belongingness and appeal which may dissuade women from applying for a job [16]. An institution cannot hire a candidate who chooses not to apply on account of feelings of ex- clusion. Therefore, the language of recruitment can be a tool that per- petuates the status quo. Use of gender-neutral language with concrete descriptors of jobs is the optimal way to maximize recruitment reach while improving qualified diverse candidacy.

Table 1

Job characteristics

Job type (%)

Total (n = 74)

Table 2

Common gender-coded words used in EM physician job advertise- ments

Academic 32 (43.2)

Clinical 28 (37.8)

Academic administration 10 (13.5)

Administration 4 (5.4)

Region (%)

Southeast

20 (27)

Confident

Share

Norththeast

19 (25.7)

Active

Responsibility

Midwest

18 (24.3)

Independent

Support

Southwest

11 (14.9)

Driven

Enthusiasm

West

6 (8.1)

Individual

Inclusive

Subspecialty required (%)

Decisive

Connect

No

70 (94.6)

Courage

Trust

Yes

4 (5.4)

Force

Together

Masculine-coded Feminine-coded

Leader- Commit-

Compet- Collaborat-

Challeng- Compassion-

Interestingly, there was no statistical difference between the gender-coding in job advertisements between academic vs non- academic and administrative vs non-administrative jobs. Given the small number of administrative jobs in this study, we may not have had the power to determine a difference if one exists. However, aca- demic jobs (like clinical jobs) were also mostly male-coded, in spite of the fact that women now make up 41% of full-time academic positions in EM in 2018 [1]. Therefore, the difference in sex distribution in aca- demic and non-academic jobs is not reflected in the gender-coding of advertisements for those kinds of jobs.

Whether gender-coded language in EM jobs influences the actual applicant pool has not been studied. It would be helpful to assess feel- ings of belongingness and job interest as a factor of job advertisement language. Additionally, since sexism in medicine is structural and long-standing, the role of biased advertising in EM physician jobs may be small relative to structural bias at upstream levels in society and ed- ucation that discourage women from entering EM. Nevertheless, re- moving gender-coding in job advertisements in EM has the potential to lead to broader application from females which may ultimately in- crease the percentage of females in EM.

  1. Limitations

There are several limitations of this study. First, this is a cross sec- tional study of jobs advertised during a specific period of time, and may not represent all EM jobs. Additionally, many of our searched websites were specific to EM or specific to academic jobs, and this may introduce bias. There is no way for us to know who writes the re- viewed job advertisements (whether a physician group, a person in marketing, or some other hiring agency) and this likely plays a role in advertisement wording. We specifically excluded websites associated with individual CMGs, and a large number of EM jobs are through CMGs. We did not want to overwhelm our dataset with mass templated advertisements from CMGs. However, this means that there is a large proportion of EM jobs that we did not assess for gender-coded language. Nevertheless, in looking at CMG websites, dozens of jobs are advertised with identical wording, and including each of them would dramatically dilute our findings as a whole. Our study was performed during the COVID pandemic, and this may have played a significant factor in terms of influencing the availability and advertising of EM jobs. Al- though our study reflected a preponderance of masculine-coded lan- guage in EM job advertisements, we cannot know whether this is part of the cause of EM jobs being dominated by men, or if it is a result of a historically strongly male medical field. Finally, although all our investi- gators were trained and all data was manually reviewed by 2 investiga- tors, there is the potential for errors in data abstraction.

  1. Conclusions

Job advertisements for EM physicians tend to contain more masculine-coded language. Almost all job advertisements for EM physi- cians in this study contained at least one gender-coded word. There was no difference between gender-coding of job advertisements when looking at different regions, subspecialty requirement, academic vs non-academic or administrative vs. non-administrative positions.

Financial support

None.

Author statement

Study concept and design: RJ, JW, KO, VP. Acquisition of the data: DL, MM, KK, KO, VP, RJ. Analysis and interpretation of the data: KO, RJ, VP, CJ. Drafting of the manuscript: KO, RJ, CJ.

Critical revision of the manuscript for important intellectual content: JW, RJ, KO, CJ.

Statistical expertise: RJ. Obtained funding: N/A.

Administrative, technical, or material support: KO. Study supervision: VP, KO, RJ.

None of the study authors have any conflict of interest to report.

CRediT authorship contribution statement

Kelly O’Brien: Writing – review & editing, Writing – original draft, Project administration, Investigation, Formal analysis, Data curation, Conceptualization. Veronica Petra: Project administration, Investiga- tion, Data curation, Conceptualization. Divya Lal: Investigation, Data curation. Kim Kwai: Investigation, Data curation. Marian McDonald: Investigation, Data curation. Judy Wallace: Writing – review & editing, Visualization, Resources, Conceptualization. Chloe Jeanmonod: Con- ceptualization, Visualization, Writing – review & editing. Rebecca Jeanmonod: Conceptualization, Data curation, Formal analysis, Investi- gation, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing.

Declaration of Competing Interest

There are no conflicts of interest for any of the investigators.

References

  1. Lautenberger DM and Dandar VM. The state of women in academic medicine 2018-2019: Exploring pathways to equity updates. AAMC. https://www.aamc.org/ data-reports/data/2018-2019-state-women-academic-medicine-exploring- pathways-equity. Accessed 12/2/2021.
  2. Paturel A. Why women leave medicine. AAMC; October 1, 2019.https://www.aamc. org/news-insights/why-women-leave-medicine Accessed 12/2/2021.
  3. Active Physicians by Sex and Specialty, 2019. AAMC. https://www.aamc.org/data- reports/workforce/interactive-data/active-physicians-sex-and-specialty-2019. Accessed 12/1/2021.
  4. Cooper LA, Powe NR. Disparities in patient experiences, health care processes, and outcomes: the role of patient-provider racial, ethnic, and language concordance. The Commonwealth Fund; July 1, 2004. https://www.commonwealthfund.org/ publications/fund-reports/2004/jul/disparities-patient-experiences-health-care- processes-and Accessed 12/1/2021.
  5. Poma PA. Race/ethnicity concordance between patients and physicians. J Natl Med Assoc. 2017;109(1):6-8.
  6. Derose KP, Hays RD, McCaffrey DF, Baker DW. Does physician gender affect satisfac- tion of men and women visiting the emergency department? J Gen Intern Med. 2001;16(4):218-26.
  7. Trends in the utilization of emergency department services, 2009-2018. US Department of Health and Human Services. Office of the assistant secretary for planning and evaluation; March 1, 2021. https://aspe.hhs.gov/reports/trends- utilization-emergency-department-services-2009-2018 Accessed 11/16/2021.
  8. Tsugawa Y, Jena AB, Figueroa JF, Orav EJ, Blumenthal DM, Jha AK. Comparison of hos- pital mortality and readmission rates for Medicare patients treated by male vs fe- male physicians. JAMA Intern Med. 2017;177(2):206-13.
  9. Parker RB, Stack SJ, Schneider SM. ACEP diversity summit 2016 attendees. Why di- versity and inclusion are critical to the American College of Emergency Physicians’ future success. Ann Emerg Med. 2017;69(6):714-7.
  10. Bicket J, Wara D, Atkinson BF, et al. Increasing women’s leadership in academic med- icine: report of the AAMC project implementation committee. Acad Med. 2002;77 (10):1043-61.
  11. Choo EK, Kass D, Westergaard M, et al. The development of best practice recommen- dations to support the hiring, recruitment, and advancement of women physicians in emergency medicine. Acad Emerg Med. 2016;23(11):1203-9.
  12. Morton MJ, Bristol MB, Atherton PH, Schwab CW, Sonnad SS. Improving the recruit- ment and hiring process for women faculty. J Am Coll Surg. 2008;206(6):1210-8.
  13. Proceedings of the diversity and inclusion innovation forum: unconscious bias in ac- ademic medicine. How the prejudices we don’t know we have affect medical educa- tion, medical careers, and patient health. AAMC; 2017..
  14. Carr PL, Gunn CM, Kaplan SA, Raj A, Freund KM. Inadequate progress for women in academic medicine: findings from the national faculty study. J Womens Health (Larchmt). 2015;24(3):190-9.
  15. Berlin G, Darino L, Greenfield M, Starikova I. Women in the healthcare industry. McKinsey & Company; Jun 7, 2019. https://www.mckinsey.com/industries/healthcare- systems-and-services/our-insights/women-in-the-healthcare-industry Accessed 12/1/ 2021.
  16. Gaucher D, Friesen J, Kay AC. Evidence that gendered wording in job advertisements exists and sustains gender inequality. J Personal Soc Psych. 2011;101(1):109-28.
  17. Deshmukh A. Bias in job descriptions: your first step to creating a more diverse work- force. Myacom; November 12, 2019. https://www.mya.com/blog/unconscious-bias- in-job-descriptions/ Accessed 12/1/2021.
  18. Sczesny S, Formanowicz M, Moser F. Can gender-fair language reduce gender stereotyping and discrimination? Front Psychol. 2016;7. https://www.frontiersin. org/articles/10.3389/fpsyg.2016.00025/full. [Accessed 12/1/2021].
  19. Kay AC, Gaucher D, Peach JM, et al. Inequality, discrimination, and the power of the status quo: direct evidence for a motivation to see the way things are as the way they should be. J Personal Soc Psych. 2009;97:421-34.
  20. Kay AC, Zanna MP. A contextual analysis of the system justification motive and its societal consequences. In: Jost T, Kay AC, Thoridottir H, editors. Social and psycho- logical bases of ideology and system justification; 2009. p. 158-81. New York.
  21. Pratto F, Steward AL. Social dominance theory. In: Christie DJ, editor. The encyclope- dia of peace psychology. Blackwell Publishing; 2011. https://onlinelibrary.wiley. com/doi/full/10.1002/9780470672532.wbepp253. stewart AL. Social dominance theory. Accessed 12/1/2021.

Leave a Reply

Your email address will not be published. Required fields are marked *