Video-based classroom observation tools can provide constructive feedback to K-6 teachers and assist their training and skill development. However, accurate classroom observation ratings necessitate observer training, requiring the investment of time, money and skilled labor. We propose a new AI-assisted pipeline to automate the classroom observation rating process utilizing a deep learning framework. Specifically, we combine three streams: action recognition, object detection, and age estimation networks to detect classroom instructional activities. We conducted the experiments on a novel labeled K-6 (elementary) classroom observation video dataset to detect human activities in the classroom environment. We labeled the data with the help of experienced annotators trained in classroom observation instruments. We also conducted multiple inter-rater reliability studies to ensure reliable labels. Our preliminary experimental results show promise for detecting multiple classroom instructional activity labels. © 2021 IEEE.

In this letter we address the task of recognizing assembly actions as a structure (e.g. a piece of furniture or a toy block tower) is built up from a set of primitive objects. Recognizing the full range of assembly actions requires perception at a level of spatial detail that has not been attempted in the action recognition literature to date. We extend the fine-grained activity recognition setting to address the task of assembly action recognition in its full generality by unifying assembly actions and kinematic structures within a single framework. We use this framework to develop a general method for recognizing assembly actions from observation sequences, along with observation features that take advantage of a spatial assembly's special structure. Finally, we evaluate our method empirically on two application-driven data sources: 1) An IKEA furniture-assembly dataset, and 2) A block-building dataset. On the first, our system recognizes assembly actions with an average framewise accuracy of 70% and an average normalized edit distance of 10%. On the second, which requires fine-grained geometric reasoning to distinguish between assemblies, our system attains an average normalized edit distance of 23%-a relative improvement of 69% over prior work.

In participatory design research (PDR), all stakeholders actively engage in decision-making and outcomes in ways that are democratic, reflect stakeholder needs and interests, and promote successful design outcomes. During the participatory design process, building and maintaining trust is imperative, especially in youth-based programs where youth have little to no voice or influence over the decisions that affect the programs in which they participate. This is especially true for youth members of minoritized groups. Without trust, youth might not engage, participate, or express their own values and culture within a project [1]. Thus, it is important to create a safe learning environment that allows for the sharing and nurturing of design knowledge, skills, meaning-making, and attitudes. The three dimensions of trust-cognitive, affective, and politicized-examined in this study have been researched within the scope of multiple disciplines, including organizational psychology, biomedical research, communication, and economics. However, little research has been done on the individual and interactional effects of inter-relational trust in the context of participatory design. In this study, we examined the initial development and inter-relationships between cognitive, affective, and politicized trust in a PDR project within an after-school community program that included 2 staff and 11 middle school youth. The goal of this ongoing project is to co-design an educational experience for youth in which they explore how computerized algorithmic processes can reinforce discrimination, racism, and prejudice in socio-technical systems. The research question in this study is: What are the interrelationships among cognitive, affective, and politicized trust development during one initial participatory design activity with youth, staff, and researchers? One discussion activity about Google search algorithms was analyzed using the methodology of quantitative ethnography (QE) and the tools of qualitative critical discourse analysis and Epistemic Network Analysis (ENA). We used critical discourse analysis [2] to examine talk, tone, and grammar through a political lens that focused on status, power, and relationships. In particular, this analysis focused on examining how participants engaged (or did not engage) with each other when discussing issues and inequities around race and gender in the context of internet search algorithms. The coded data was then analyzed using the ENA webtool [3], which counts the co-occurrences of codes and visualizes them in a node-link network model. This allows for the relationships between the codes to be seen in a temporal context. In this study, there were several displays of youth trust construction and deconstruction with staff and the researcher. Cognitive trust tended to be foundational in the building of both affective and politicized trust. Affective trust was established next and began to work and interact with cognitive trust in a bidirectional manner with each dimension influencing the other. Politicized trust seemed to require the presence of both cognitive and affective trust, and breakdowns in either of these dimensions of trust caused breakdowns in politicized trust. All three dimensions of trust were complex, dynamic, and interrelated throughout the study. These findings support a three-dimensional trust framework for educational design involving youth as active participants and co-designers. These findings also suggest that shared knowledge, solidarity, and willingness to show vulnerability were crucial to the initial development process of all three forms of trust. The results further suggest that it was difficult for adults to gain politicized trust within the youth group, and an adult facilitating the direction of politicized discussions involving gender and racial representation decreased levels of vulnerability and engagement displayed by youth. These patterns of interaction require us to further probe how vulnerability, discord, and solidarity impact engagement and opportunities for learning among youth, especially in light of learning that involves more politicized issues.

This paper presents themes that emerge from empirical literature on women of color (WOC) in computer science (CS) graduate education. We ask, According to the literature, what factors affect the experiences, participation, and advancement of WOC in CS graduate degrees? The findings are drawn from a subset of literature on graduate education from our National Science Foundation-funded project, Literature Analysis and Synthesis of Women of Color in Technology and Computing. Findings of on-campus social supports include student support groups and peers who provided community, navigation strategies, and motivation to succeed. Family and friends also provided recognition and encouragement. Students attending Historically Black Colleges and Universities reported that their schools provided them with structural support through recognition and investment in their potential. Findings of barriers include a sense of isolation, as well as professors and male classmates creating a culture of hostility and exclusion for WOC. Despite these challenges, WOC used individual and social strategies to navigate and persist. They drew on their determination, dedication to achieving goals, and past challenges to stay motivated and succeed while also developing soft skills. They were further motivated to use their knowledge of CS as a tool to solve problems and help others. Our synthesis contributes an analysis of the social and structural supports and barriers for WOC in the understudied field of CS graduate education. This research will increase knowledge about success strategies to retain women of color with advanced CS degrees to fill the United States' technological workforce needs.

Student engagement is a key component of learning and teaching, resulting in a plethora of automated methods to measure it. Whereas most of the literature explores student engagement analysis using computer-based learning often in the lab, we focus on using classroom instruction in authentic learning environments. We collected audiovisual recordings of secondary school classes over a one and a half month period, acquired continuous engagement labeling per student (N=15) in repeated sessions, and explored computer vision methods to classify engagement from facial videos. We learned deep embeddings for attentional and affective features by training Attention-Net for head pose estimation and Affect-Net for facial expression recognition using previously-collected large-scale datasets. We used these representations to train engagement classifiers on our data, in individual and multiple channel settings, considering temporal dependencies. The best performing engagement classifiers achieved student-independent AUCs of .620 and .720 for grades 8 and 12, respectively, with attention-based features outperforming affective features. Score-level fusion either improved the engagement classifiers or was on par with the best performing modality. We also investigated the effect of personalization and found that only 60 seconds of person-specific data, selected by margin uncertainty of the base classifier, yielded an average AUC improvement of .084.

While many advanced haptic devices are under development, touchscreens are one of the most readily available platforms. In this paper, we leverage a leap forward in vibration-based haptics, Apple's new CoreHaptics API, and investigate its potential for a multi-finger vibration experience. We present a perceptual user study (N=15) that investigates multi-finger vibration identification and exploration strategies and we conduct a laser doppler vibrometry study, uncovering the challenges of developing consistent, high-quality vibration feedback across hardware platforms that vary in actuation principle, screen size, and external attachments. Repeated-measures ANOVA tests showed no statistically significant results in time, error, or weighted error for vibration identification across all participants based on number of fingers. However, one-way ANOVAs did show significant results within individuals, illustrating benefits of multiple-finger exploration. The most effective strategies for locating vibrations involved multiple fingers grouped together using sweeping motions across the screen. Our LDV study demonstrated that CoreHaptics and the Taptic Engine in the iPhone were capable of more accurately recreating specific frequencies than coin motors in Android devices. This research supports a move to a multi-finger, vibrotactile touchscreen experience, while highlighting the challenges of creating a consistent vibration experience across hardware platforms.

Historic racial disparities in the United States have created an urgent need for evidence-based strategies promoting African American students' academic performance via school-based ethnic-racial socialization and identity development. However, the temporal order among socialization, identity, and academic performance remains unclear in extant literature. This longitudinal study examined whether school cultural socialization predicted 961 African American adolescents' grade point averages through their ethnic-racial identities (49.6% males; M-age = 13.60; 91.9% qualified for free lunch). Results revealed that youth who perceived more school cultural socialization had better grades 1 and 2 years later. In addition, identity commitment (but not exploration) fully mediated these relations. Implications for how educators can help adolescents of color succeed in schools are discussed.

Three hundred and ninety-one children (195 girls; M-age = 9.56 years) attending Grades 1 and 5 completed implicit and explicit measures of math attitudes and math self-concepts. Math grades were obtained. Multilevel analyses showed that first-grade girls held a strong negative implicit attitude about math, despite no gender differences in math grades or self-reported (explicit) positivity about math. The explicit measures significantly predicted math grades, and implicit attitudes accounted for additional variance in boys. The contrast between the implicit (negativity for girls) and explicit (positivity for girls and boys) effects suggest implicit-explicit dissociations in children, which have also been observed in adults. Early-emerging implicit attitudes may be a foundation for the later development of explicit attitudes and beliefs about math.

Maintaining learning engagement throughout adolescence is critical for long-term academic success. This research sought to understand the role of metacognition and motivation in predicting adolescents' engagement in math learning over time. In two longitudinal studies with 2,325 and 207 adolescents (ages 11-15), metacognitive skills, interest, and self-control each uniquely predicted math engagement. Additionally, metacognitive skills worked with interest and self-control interactively to shape engagement. In Study 1, metacognitive skills and interest were found to compensate for one another. This compensatory pattern further interacted with time in Study 2, indicating that the decline in engagement was forestalled among adolescents who had either high metacognitive skills or high interest. Both studies also uncovered an interaction between metacognitive skills and self-control, though with slightly different interaction patterns.

This article used self-regulated learning as a theoretical lens to examine the individual and interactive associations between a growth mindset and metacognition on math engagement for adolescent students from socioeconomically disadvantaged schools. Across three longitudinal studies with 207, 897, and 2,325 11- to 15-year-old adolescents, students' beliefs that intelligence is malleable and capable of growth over time only predicted higher math engagement among students possessing the metacognitive skills to reflect upon and be aware of their learning progress. The results suggest that metacognitive skills may be necessary for students to realize their growth mindset. Thus, growth mindsets and metacognitive skills should be promoted together to capitalize on the mutually reinforcing effects of each, especially among students in socioeconomically disadvantaged schools.