An official website of the United States government
Is there a relationship between mathematics ability beliefs and STEM degrees? Fields such as physics, engineering, mathematics, and computer science (PEMC) are thought to require talent or brilliance. However, the potential effects of difficulty perceptions on students’ participation in STEM have yet to be examined using a gender and race/ethnicity intersectional lens. Using nationally representative U.S. longitudinal data, we measure gender and racial/ethnic variation in secondary students’ orientation towards mathematics difficulty. We observed nuanced relationships between mathematics difficulty orientation, gender, race/ethnicity, and PEMC major and degree outcomes. In secondary school, the gap between boys’ and girls’ mathematics difficulty orientations were wider than gaps between White and non-White students. Mathematics difficulty orientation was positively associated with both declaring majors and earning degrees in PEMC. This relationship varied more strongly based on gender than race/ethnicity. Notably, Black women show higher gains in predicted probability to declare a mathematics-intensive major as compared to all other women, given their mathematics difficulty orientations. This study’s findings show that both gender and racial/ethnic identities may influence the relationship between mathematics difficulty orientation and postsecondary STEM outcomes.
more »
« less
A Rose by Any Other Name would not Smell as Sweet: Social Bias in Names Mistranslation
We ask the question: Are there widespread disparities in machine translations of names across race/ethnicity, and gender? We hypothesize that the translation quality of names and surrounding context will be lower for names associated with US racial and ethnic minorities due to these systems' tendencies to standardize language to predominant language patterns. We develop a dataset of names that are strongly demographically aligned and propose a translation evaluation procedure based on round-trip translation. We analyze the effect of name demographics on translation quality using generalized linear mixed effects models and find that the ability of translation systems to correctly translate female-associated names is significantly lower than male-associated names. This effect is particularly pronounced for female-associated names that are also associated with racial (Black) and ethnic (Hispanic) minorities. This disparity in translation quality between social groups for something as personal as someone's name has significant implications for people's professional, personal, and cultural identities, self-worth and ease of communication. Our findings suggest that more MT research is needed to improve the translation of names and to provide high-quality service for users regardless of gender, race, and ethnicity.
more »
« less
- Award ID(s):
- 2229885
- PAR ID:
- 10522354
- Publisher / Repository:
- Association for Computational Linguistics
- Date Published:
- Page Range / eLocation ID:
- 3933 to 3945
- Format(s):
- Medium: X
- Location:
- Singapore
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
This research paper investigates the relationship between race/ ethnicity, gender, first-generation college student status, and engineering identity using cross-sectional data from early-career engineering majors. Measures of engineering identity are increasingly used in models of engineering education to evaluate how identity contributes to success and persistence of engineering students. Engineering identity is generally assumed to contribute to educational success, with stronger engineering identity leading to persistence. At the same time, data clearly shows that persistence of engineering students varies by race/ethnicity and gender. Given these previous findings, we would expect to find that engineering identity will vary by race/ ethnicity, gender, and first generation status. Yet, relatively little work has quantitatively compared how engineering identity differs across racial/ ethnic groups and gender. While researchers are increasingly trying to gain a better understanding of engineering identity among Latina students, for example, the literature has not yet adequately accounted for how Latina students differ from their non-Hispanic white peers. This works seeks to address that gap in the literature with an exploration of the ways that race/ethnicity, gender, and first generation status work together to impact engineering identity among early-career engineering students at four public Hispanic-Serving Institutions (HSIs) in the Southwestern United States. We conducted surveys as part of a longitudinal study on STEM education. Data discussed here comes from baseline surveys of three cohorts of engineering students (N=475). Approximately two-thirds of the respondents were attending a traditional 4-year university while the remainder (N=159) were attending community college at the time of the survey. Approximately two-thirds of the respondents identified as Latinx, 27% identified as female, and 26.5% reported that they were first-generation college students. While expectations were that engineering identity would vary by race/ethnicity and gender, preliminary analyses of our data unexpectedly reveal no significant differences between Latinx and White students in terms of their engineering identity and no significant differences in engineering identity between male and female students. Interactions between race/ethnicity and gender were also tested and yielded no significant differences between early-career Latinx and White students in terms of their engineering identity. Finally, students who reported that they will be the first in their family to get a college degree had significantly lower engineering identity scores (=-.422; p=.001). These results lead us to conclude that first generation status at HSIs may be more important than gender and race/ ethnicity in the development of engineering identity for early career students.more » « less
-
Abstract In several author name disambiguation studies, some ethnic name groups such as East Asian names are reported to be more difficult to disambiguate than others. This implies that disambiguation approaches might be improved if ethnic name groups are distinguished before disambiguation. We explore the potential of ethnic name partitioning by comparing performance of four machine learning algorithms trained and tested on the entire data or specifically on individual name groups. Results show that ethnicity‐based name partitioning can substantially improve disambiguation performance because the individual models are better suited for their respective name group. The improvements occur across all ethnic name groups with different magnitudes. Performance gains in predicting matched name pairs outweigh losses in predicting nonmatched pairs. Feature (e.g., coauthor name) similarities of name pairs vary across ethnic name groups. Such differences may enable the development of ethnicity‐specific feature weights to improve prediction for specific ethic name categories. These findings are observed for three labeled data with a natural distribution of problem sizes as well as one in which all ethnic name groups are controlled for the same sizes of ambiguous names. This study is expected to motive scholars to group author names based on ethnicity prior to disambiguation.more » « less
-
Abstract Using the Household Pulse Survey and American Community Survey, this study examines employment insecurity experienced across different racial/ethnic groups of the U.S. labor force under the pandemic disruptions. It highlights significant employment security disparities based on race, ethnicity, and income during the pandemic. However, there are no significant gender and racial disparities within the lowest income group when controlling for other conditions. In contrast, gender and racial disparities in EI are much more pronounced among mid-to-high income groups. Non-White individuals were disproportionately affected by job loss due to health and COVID-related employment issues, unlike Whites who faced unemployment more due to other factors. This pattern was more evident among lower-income groups. The trends shifted in later stages, with high-income Black and Hispanic workers becoming more likely to be unemployed due to non-health and non-employment reasons. Middle-income workers across all races were least likely to stop working for reasons other than COVID-related health or employment issues. In addition, regardless race or ethnicity, women more likely to be unemployed due to health reasons and less so due to employment issues compared to men, and the gender disparities increased with higher household incomes. We propose that the apparent immediate effects of the pandemic are actually indicative of deeper, systemic issues within the U.S. labor market, specifically the occupational segregation tied to race/ethnicity, gender, and class. Recovery efforts must take a holistic approach and integrate economic development policies, workforce development strategies, and social policies targeting poverty alleviation, health disparities, and people of color.more » « less
-
null (Ed.)Students belonging to marginalized groups experience positive impacts when taught by a teacher of the same race, ethnicity, and gender. The unique nature of standardized, asynchronous online course taking allows for greater separation of any possible educational benefits of student versus teacher-driven mechanisms contributing to these improved outcomes. Using a student-by-course fixed effect strategy on data from a large urban school district, I examined associations between whether students experienced racial/ethnic or gender congruence with their remote instructor and both engagement and learning outcomes. Students who identified as Black demonstrated higher rates of engagement, although no difference in achievement, within lessons taught by a same-race remote instructor. I find that representation is associated with engagement even when instructors follow closely scripted lessons, representation occurs in only small doses, and instruction occurs in an impersonal setting.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.18653/v1/2023.emnlp-main.239
