Drawing on 12 semi-structured interviews with Black, Latina, and white graduate women who either continued or discontinued their STEM doctoral degrees, the present study examined the psychological impact of navigating marginalizing experiences in white male-dominated STEM environments. Using thematic analysis grounded in a social constructivist paradigm, researchers identified three emergent themes: 1) institutional challenges as contextual barriers, 2) impact on wellbeing and STEM persistence, and 3) contextual supports and coping. These findings indicate that challenging STEM encounters within the higher education environment contributed to increased stress, depression, anxiety, and suicidal ideation among graduate women in STEM from diverse racial/ethnic backgrounds. The compound effect of these STEM stressors and their subsequent psychological toll contributed to decreased STEM persistence among participants. Study implications highlight the need for faculty and university administrators to challenge and address institutional norms that operate as contextual barriers, destigmatize discussions surrounding mental health, and adopt a whole person approach to supporting graduate women in STEM.

Though one might imagine that traditional gender stereotypes about math have lessened over the years, this assumption remains to be tested. We know little about the extent to which parents' gender stereotypes about math abilities and their correlates have changed over time or the extent to which they replicate across research methods and racial/ethnic groups. To address these issues, we used four longitudinal U.S. datasets collected from 1984 to 2009 (n's = 537-14,470, 49-53% girls, 32-95% White, 1-59% Black, 0-22% Latinx) that included similar survey items. Across the datasets, parents believed that boys were better at math than girls. This was particularly consistent among White parents, where the small effects favoring boys replicated across all four datasets covering three decades. Compared to White parents, Black and Latinx parents were significantly less likely to favor boys. After controlling for parent education, income, and their child's math grade, parents' traditional gender stereotypes were significantly and negatively associated with girls' math self-concept, a small effect that replicated across all four datasets. These findings have implications for teachers and parents, as parents (particularly White parents) were significantly more likely to hold traditional math gender stereotypes, which relates to children's math self-concept.

Threshold concepts are transformative elements of domain knowledge that enable those who attain them to engage domain tasks in a more sophisticated way. Existing research tends to focus on the identification of threshold concepts within undergraduate curricula as challenging concepts that prevent attainment of subsequent content until mastered. Recently, threshold concepts have likewise become a research focus at the level of doctoral studies. However, such research faces several limitations. First, the generalizability of findings in past research has been limited due to the relatively small numbers of participants in available studies. Second, it is not clear which specific skills are contingent upon mastery of identified threshold concepts, making it difficult to identify appropriate times for possible intervention. Third, threshold concepts observed across disciplines may or may not mask important nuances that apply within specific disciplinary contexts. The current study therefore employs a novel Bayesian knowledge tracing (BKT) approach to identify possible threshold concepts using a large data set from the biological sciences. Using rubric-scored samples of doctoral students' sole-authored scholarly writing, we apply BKT as a strategy to identify potential threshold concepts by examining the ability of performance scores for specific research skills to predict score gains on other research skills. Findings demonstrate the effectiveness of this strategy, as well as convergence between results of the current study and more conventional, qualitative results identifying threshold concepts at the doctoral level.

Economic stress indices are beginning to be developed as gauges of business cycle conditions for regional economies. Although the popularity of these metrics is increasing, there have been only a small number of studies that analyze the effectiveness of these tools for monitoring regional economic developments. This effort employs data for one such index that is maintained by the University of Texas at El Paso Border Region Modeling Project. The sample period covers December 2002 through March 2019. Specific components of the index include inflation, unemployment, and housing prices. Estimation results indicate that the effects of any changes in economic stress levels may take approximately 16 months to be fully experienced within the El Paso retail sector. Simulation results indicate that a one-time, 1-point increase in economic stress leads to a $2.987 million decrease in monthly retail sales.

[No abstract available]

In prevention science and related fields, large meta-analyses are common, and these analyses often involve dependent effect size estimates. Robust variance estimation (RVE) methods provide a way to include all dependent effect sizes in a single meta-regression model, even when the exact form of the dependence is unknown. RVE uses a working model of the dependence structure, but the two currently available working models are limited to each describing a single type of dependence. Drawing on flexible tools from multilevel and multivariate meta-analysis, this paper describes an expanded range of working models, along with accompanying estimation methods, which offer potential benefits in terms of better capturing the types of data structures that occur in practice and, under some circumstances, improving the efficiency of meta-regression estimates. We describe how the methods can be implemented using existing software (the metafor and clubSandwich packages for R), illustrate the proposed approach in a meta-analysis of randomized trials on the effects of brief alcohol interventions for adolescents and young adults, and report findings from a simulation study evaluating the performance of the new methods.

This theoretical paper sets forth two aspects of predication, which describe how students perceive the relationship between a property and an object. We argue these are consequential for how students make sense of discrete mathematics proofs related to the properties and how they construct a logical structure. These aspects of predication are (1) populating the way students generate sets of examples of the property, and (2) testing membership how one tests whether or not a given object has a specific property. Using data from two teaching experiments in which undergraduate students read proofs of theorems about the discrete concept of multiple relations, we illustrate the nature of these aspects of predication and demonstrate how they help explain student interpretations of the proofs. We argue that these particular properties from number theory likely have correlates in many other discrete mathematics topics because of the role of computation/algorithms for defining and testing properties as well as the role of iteration and recursion in populating examples. We anticipate that these constructs will be useful to teachers and researchers of discrete mathematics to foster and assess student understanding of various mathematical properties. They provide tools for thinking about what it means to understand properties in a rich and coherent way that supports understanding complex lines of inference and generalizations.

Apparel product development is an iterative problem-solving process that is heuristic in nature and involves turning 2D flat patterns into a 3D garment that would fit human anatomy. The digital transformation trend in the apparel industry will accelerate in the post-pandemic era. Therefore, it is crucial to better understand the dynamics as well as the types of information generated during apparel product development to translate this information into the digital realm to better support the apparel industry. Technical apparel designers acquire important knowledge on garment fit during their years in the workforce. Documenting their knowledge could assure that the company's know-how would be saved and used to train the future workforce. Technical designers' problem-solving strategies can be tracked, coded, and made a part of artificial intelligence (AI) technology for product development sessions. This would help strengthen the competitiveness of both existing and new design companies. However, the knowledge possessed by the technical designers to enable much-improved digitalization would require a thorough analysis of tacit or implicit knowledge, which is acquired through years of experience. Due to the nature of how apparel designers learn by directly manipulating materials with their hands, this knowledge type is challenging to document. Nonetheless, if a successful method can be developed to document this knowledge, it would be very rewarding, especially for organizational knowledge management, and allow companies to digitalize using AI.

The ability to identify whether a user is zoning out (mind wandering) from video has many HCI (e.g., distance learning, high-stakes vigilance tasks). However, it remains unknown how well humans can perform this task, how they compare to automatic computerized approaches, and how a fusion of the two might improve accuracy. We analyzed videos of users' faces and upper bodies recorded 10s prior to self-reported mind wandering (i.e., ground truth) while they engaged in a computerized reading task. We found that a state-of-the-art machine learning model had comparable accuracy to aggregated judgments of nine untrained human observers (area under receiver operating characteristic curve [AUC] =.598 versus .589). A fusion of the two (AUC = .644) outperformed each, presumably because each focused on complementary cues. Furthermore, adding more humans beyond 3-4 observers yielded diminishing returns. We discuss implications of human-computer fusion as a means to improve accuracy in complex tasks.