When people are exposed to information that leads them to overestimate the actual amount of genetic difference between racial groups, it can augment their racial biases. However, there is apparently no research that explores if the reverse is possible. Does teaching adolescents scientifically accurate information about genetic variation within and between US census races reduce their racial biases? We randomized 8th and 9th grade students (n = 166) into separate classrooms to learn for an entire week either about the topics of (a) human genetic variation or (b) climate variation. In a cross‐over randomized trial with clustering, we demonstrate that when students learn about genetic variation within and between racial groups it significantly changes their perceptions of human genetic variation, thereby causing a significant decrease in their scores on instruments assessing cognitive forms of prejudice. We then replicate these findings in two computer‐based randomized controlled trials, one with adults (n = 176) and another with biology students (n = 721, 9th–12th graders). These results indicate that teaching about human variation in the domain of genetics has potentially powerful effects on social cognition during adolescence. In turn, we argue that learning about the social and quantitative complexities of human genetic variation research could prepare students to become informed participants in a society where human genetics is invoked as a rationale in sociopolitical debates.

The belief that men and women differ in science ability because of genetics contributes to gender disparities in science, technology, engineering, and math (STEM) in complex ways. In this field experiment, we explored how the content of the genetics curriculum affected beliefs about science ability through its impact on a social-cognitive bias known as neurogenetic essentialism. Students (n = 460, 8th-10th grade) were randomized to read a genetics text that (a) explained plant sex differences, (b) explained human sex differences, or (c) refuted neurogenetic essentialism. After reading, students in the two genetics of sex conditions had significantly greater belief in neurogenetic essentialism and the innate basis of science ability compared with students who read the text that refuted neurogenetic essentialism. Structural equation modeling (SEM) of the experimental data demonstrated that the effect of the readings on the belief that science ability is innate was mediated by neurogenetic essentialism and this indirect effect was significant for girls but not boys. In turn, the belief that science ability is innate predicted lower future interest in STEM for girls, but not for boys. These findings suggest that learning about human genetic difference is not a socially neutral endeavor. Implications for mitigating gender disparities in STEM are discussed.

OpenSciEd is an ambitious effort to implement the vision of the Framework for K-12 Science Education and the Next Generation Science Standards broadly across the United States. The premise of OpenSciEd is that high quality instructional materials can play a critical role in transforming science teaching and learning at a broad scale. To achieve its goal, this collaborative project is developing instructional materials for middle school science that support the shifts in practice required to achieve the outcomes called for by the Framework for K-12 Science Education and the Next Generation Science Standards at a large scale. The OpenSciEd Middle School Program development project is addressing the challenge of making large changes in practice at a large scale through attention to (1) who participates in design and development, and how; (2) providing explicit guidance for developers in a comprehensive design framework; and (3) a design and development process that ensures participation from the desired participants and adherence to the guidelines of the design framework. The resulting instructional materials have shown promise in external reviews and field tests, but their success in achieving the project's goals for transforming science will depend on the circumstances in which the program is implemented.

This article presents the results of exploratory research with community college students from non-dominant linguistic backgrounds (NDLB) in an introductory astronomy class as they collaborated to reconstruct dynamic cosmology visualizations through drawing. Data included student discourse during the drawing activity, post-activity interviews, and the drawings themselves. This work comes from the theoretical perspective that revealing student competence should be an essential part of science education research, and is guided by sociocultural theory. Results indicate that dynamic cosmology visualizations can support the development of cosmological literacy by facilitating heterogeneous sense-making strategies. The activity of drawing the visualizations in groups created fluid, hybrid spaces where students could grapple directly with cosmology content while trying on the language of science. In light of these findings, the author argues that carefully incorporating collaborative activity around the interpretation of visualizations into learning environments can improve access to cosmology content for learners, particularly those who come from non-dominant linguistic backgrounds. # 2016 Wiley Periodicals, Inc. J Res Sci Teach Keywords: science education; language and literacy; language of science and classrooms; equity; science literacy Learners from non-dominant linguistic backgrounds (NDLB) 1 are quickly becoming one of the largest non-mainstream demographics in the educational landscape. (National Center for Education Statistics [NCES], 2011), yet these linguistically diverse learners remain underrepresented in STEM post-secondary education (Stoddart, Pinal, Latzke, & Canaday, 2002). This pattern of inequity can be attributed to a history of compounding educational, economic and social deficits (Ladson-Billings, 2006), including inequitable access to educational opportunities throughout K-15, (Eun, 2016;Mosqueda, 2011) where NDLB students persistently and disproportionally lack access to rigorous content in science (Buxton, 1998;Callahan, 2005;Genesee, Lindholm-Leary, Saunders, & Christian, 2005;Lee, 1999). In response to such inequities, researchers have called for the development of educational tools that are more accessible to a broader range of ways of knowing and doing science (Barton, 1998;Bouillion & Gomez, 2001).Cosmology is the study of the Universe, in particular its structure, organization and dynamics. According to the American Association for the Advancement of Science (AAAS), "finding our place in the cosmic scheme of things and how we got here is a task for the ages À past, present, and future. . . If being educated means having an informed sense of time and place, then it is essential for a person to be familiar with the scientific aspects of the [U]niverse and know something of its origin and structure" (AAAS Project 2061, 1993. Although there is a solid body of research on NDLB students in the science classroom (e.g., Echevarria, Vogt & Short, 2004;Lee & Fradd, 1998;Stoddart, Solis, ...

Argumentation is fundamental to science education, both as a prominent feature of scientific reasoning and as an effective mode of learning—a perspective reflected in contemporary frameworks and standards. The successful implementation of argumentation in school science, however, requires a paradigm shift in science assessment from the measurement of knowledge and understanding to the measurement of performance and knowledge in use. Performance tasks requiring argumentation must capture the many ways students can construct and evaluate arguments in science, yet such tasks are both expensive and resource‐intensive to score. In this study we explore how machine learning text classification techniques can be applied to develop efficient, valid, and accurate constructed‐response measures of students' competency with written scientific argumentation that are aligned with a validated argumentation learning progression. Data come from 933 middle school students in the San Francisco Bay Area and are based on three sets of argumentation items in three different science contexts. The findings demonstrate that we have been able to develop computer scoring models that can achieve substantial to almost perfect agreement between human‐assigned and computer‐predicted scores. Model performance was slightly weaker for harder items targeting higher levels of the learning progression, largely due to the linguistic complexity of these responses and the sparsity of higher‐level responses in the training data set. Comparing the efficacy of different scoring approaches revealed that breaking down students' arguments into multiple components (e.g., the presence of an accurate claim or providing sufficient evidence), developing computer models for each component, and combining scores from these analytic components into a holistic score produced better results than holistic scoring approaches. However, this analytical approach was found to be differentially biased when scoring responses from English learners (EL) students as compared to responses from non‐EL students on some items. Differences in the severity between human and computer scores for EL between these approaches are explored, and potential sources of bias in automated scoring are discussed.