Moving beyond the binary - Exploring the potentials, challenges and consequences of measuring sex and gender alternatively in surveys - II |
|
| Session Organisers |
Professor Stephanie Steinmetz (University of Lausanne) Dr Mirjam Fischer (University of Cologne) Dr Léïla Eisner (University of Zurich) Dr Verena Ortmanns (GESIS - Leibniz Institute for the Social Sciences) |
| Time | Friday 16 July, 16:45 - 18:00 |
Western society's awareness that diverse gender identity exists beyond the male-female binary has grown precipitously recently. With the increasing recognition of diverse gender identity comes the question how to 'move beyond the binary' in surveys. A specific challenge is to maintain the validity and the reliability of survey measures while also representing gender diverse (e.g., trans or non-binary individuals) populations. Traditional measurements of sex and gender identity in surveys might fail to represent this diversity. Still, new measurements also raise specific issues as to whether they are understood consistently by the general population. In continuation of the first session, the four contributions in this session discuss specific challenges and provide new tools and measurement strategies to practitioners, survey providers, and researchers interested in assessing gender identity and sexual orientation in surveys.
Keywords: sex and gender indentity, LGBT, self-categorizing; gender scales
Dr Mirjam Fischer (University of Cologne) - Presenting Author
Ms Lisa de Vries (University of Bielefeld)
Mr David Kasprowski (German Institute for Economic Research)
Since 2018, Germany legally recognizes a third gender option. For the collection of survey data - especially longitudinal data – a dilemma presents itself: how can the collection of survey data respond to dynamics in social reality while safeguarding some degree of comparability of their data (across time, surveys and countries)?
Similarly, survey institutes increasingly make an effort to better accommodate same-sex couples and families in existing repeat-social surveys. This may involve more changes than simply adding answer categories to ensure that the answer options for marital status, questions about the partnership and about parenthood are exhaustive. Accommodating more diverse family and couple types in social survey also requires attention to gendered language and the assumption that couples are by default made up by a woman and a man, which appear both explicitly and implicitly in the wording of questions. A fundamental revision of these questions is required if gender and sexual diversity is to be adequately represented in social surveys. Again, consistency and comparability of these data must be weighed against validity.
We present the strategies implemented in the Socio-Economic Panel (SOEP) in response to the addition of a sample of gay, lesbian, bisexual and transgender people from 2018. Issues regarding measurement, comparability and weighting will be discussed.
Ms Carina Schönmoser (Leibniz Institute for Educational Trajectories (LIfBi)) - Presenting Author
Ms Elena Chincarini (Leibniz Institute for Educational Trajectories (LIfBi))
The respondents’ sex represents a standard question of every social survey. On the one hand, it is indispensable for filtering through the questionnaire and for adequately addressing respondents in the question formulations (at least for the German language and some others). On the other hand, sex distribution is a fundamental part of every sample description. It is often used for hypothesis formatting and testing and is of great importance as a control variable. In Germany and some other countries, a third sex category ("diverse" in the German case) has been implemented by law, alongside the female and male categories. This introduction poses new challenges for the collection of gender information in survey studies: which gender categories will be included in surveys? How are non-binary persons addressed in question wording (for language contexts with gender-specific personal pronouns and adjectives)? How can data protection for non-binary persons in scientific data be ensured? What is the acceptance of this adaptation among survey respondents?
In the process of redesigning the starting cohorts of the German National Educational Panel Study (NEPS), beginning in 2022, we have addressed some of the issues raised above about sex in surveys, by developing a new survey instrument, which has been tested via a cognitive pretest and an online pretest with adults and children aged 10 to 14 years. The focus was not only on the adjustment of the gender categories and therefore going beyond the dichotomy of female-male. It also lies on respondents´ understanding and acceptance of a separate query for sex and gender, which is particularly tricky in the German language since they actually cannot be distinguished by terms.
Initial results show that there are cultural differences in responses to the gender questions. Migrants were more likely to feel uncomfortable answering the gender questions, but some German participants also showed discomfort. Information about the gender of a partner, on the other hand, was given without further ado. Children and adolescents between the ages of 10 and 14 were not shy in answering the questions. Regarding the categories, the term "diverse" showed fewer problems of understanding than the term "intersexual", which was partly misinterpreted as homosexual. Questions about the preferred personal pronouns displayed a lack of understanding by the majority of all respondent groups.
Dr Rachel Morgan (Bureau of Justice Statistics, US Department of Justice)
Dr Jennifer Truman (Bureau of Justice Statistics, US Department of Justice) - Presenting Author
In 2016, survey items measuring gender identity were added to the National Crime Victimization Survey (NCVS), a self-report survey administered by the U.S. Census Bureau for the Bureau of Justice Statistics in the U.S. Department of Justice. NCVS respondents are asked about their experiences with nonfatal personal and property crimes that occurred within the prior 6 months. In recent years, interviews were conducted with approximately 250,000 persons age 12 or older living in the United States. The NCVS is one of two measures of criminal victimization in the U.S., along with the Federal Bureau of Investigation’s (FBI) Uniform Crime Reporting (UCR) program. While the UCR collects information on crimes reported to police, the NCVS captures the ‘dark figure of crime’ or crime that is not reported to police. Gender identity has been identified in empirical research as a correlate of victimization, and the U.S. Violence Against Women Reauthorization Act of 2013 explicitly bars discrimination based on actual or perceived gender identity status, ensuring that transgender persons have access to key services. The inclusion of this measure in the NCVS will provide important national-level estimates on victimization of transgender people, and allow researchers to understand victimization risk and access to victim services. The NCVS collects information on sex of the respondent during the household rostering component, as well as a self-reported gender identity item from persons age 16 or older. This presentation will discuss how the survey items were developed, including cognitive testing procedures, how the items were incorporated into the NCVS survey instruments, and translation into languages other than English. We will also discuss how the items are performing in the field, measures of data quality, and preliminary statistical estimates. Data quality measures include item nonresponse, breakoffs, and discordant responses between rostered data and self-reported data.
Dr Amy Alexander (University of Gothenburg)
Professor Catherine Bolzendahl (Oregon State University) - Presenting Author
Dr Patrik Öhberg (University of Gothenburg)
A broadening recognition of the non-binary character of sex and gender has coincided with greater respect for the rights of persons to identify as non-binary and/or transgender. Individual gender identities are formed by these individuals drawing upon their own positions in the social structure and the relevant societal views on issues of gender and inequality. Through gender’s co-constitution along with various social localities, we expect that a number of socio-political factors differentiate individuals’ gender identities through self-assessments of their masculine and feminine traits. Using data from a 2013 Swedish survey, our results show that men and women tend toward traditionally polarized gender identities and that social location is a particularly influential correlate of men’s claims of feminine characteristics and women’s of masculine characteristics. Individuals from younger generations and individuals that are more educated are consistently more likely to ascribe to less-binary feminine and masculine characteristics. This suggests that generational replacement and higher education may increase the tendency of populations to ascribe to less-binary gender identities.