This study examines three dimensions of Text-to-Speech (TTS) use - listening time, frequency, and sequence. Using response and log data from 28,090 students on a Grade 8 mathematics item from the 2017 National Assessment of Educational Progress, we estimated relationships between TTS use and accuracy with logistic regressions and Random Forest partial-dependence plots. TTS use was most common among lower-proficiency students. Among Below Basic students, listening to the problem statement was associated with higher accuracy, though benefits diminished beyond 25 seconds. Among Basic students, frequent toggling was negatively associated with accuracy, whereas among Proficient and Advanced students, no consistent relationship emerged. Overall, TTS appears to support access when used purposefully, while inefficient use may hinder performance. Although limited to one item and observational data, the findings suggest practical design and instructional strategies for promoting equitable access.

Classroom observations are central to research on teaching quality, but the cost and labor of training human coders limits how widely this method can be applied. This paper documents a reproducible method for using a large language model to code instructor and student behaviors from speaker-labeled classroom transcripts. The method was validated against human-coded proportion scores from 25 undergraduate STEM classes across four prompting conditions. Prompt templates, the script that calls the model, and the validation statistics behind the reported results are provided so that researchers can adapt the approach to their own observation instruments and transcription tools. The key elements of the method are summarized below.• The method uses an LLM to estimate how much of a class session is spent on different teaching and learning behaviors from audio transcripts rather than in-person observation.• It tests a progression of prompting strategies, from no guidance to step-by-step reasoning with worked examples, and evaluates agreement with human-coded reference scores.• The method captures broad instructional patterns (e.g., lecturing vs. student activity) but shows limited performance for behaviors requiring nonverbal or auditory cues.

This study investigates the time-use patterns of students with learning disabilities during digital mathematics assessments and explores the role of extended time accommodations (ETA) in shaping these patterns. Using latent profile analysis, four distinct time-use profiles were identified separately for students with and without ETA. “Initial Focusers” spend more time on simpler initial items and less time on later, more difficult items, exhibiting high omission rates and low performance. “Rapid Progressors” complete assessments quickly but exhibit shallow engagement across all items, achieving low performance. “Diligent Time Maximizers” allocate time effortfully across items but often run out of time on the last two items when ETA was not granted, achieving the second-highest scores. “Efficient Prioritizers,” excel in strategic time management, score the highest, and report strong persistence and interest in math. The findings reveal that ETA supports students who adopt meticulous strategies, such as Diligent Time Maximizers, but does not universally address the challenges faced by other profiles. This study underscores the need for tailored interventions and accommodations aligned with individual time-use profiles to foster equitable and effective learning and assessment environments.

This investigation explores the relationship between the use of digital pencil and mathematical problem-solving accuracy among 1,530 students with learning disabilities (LD) and 25,400 general education (GE) peers from the 2017 digital National Assessment of Educational Progress mathematics assessment. The term “digital pencil” in this context refers to NAEP’s “embedded pencil,” a scratchwork tool within the digital assessment interface that allows students to draw or annotate on the screen using either a stylus or their finger. Findings reveal that students with LD utilized digital pencils less frequently than their GE peers, particularly on more complex items. However, digital pencil use was associated with a 20% increase in the likelihood of GE students accurately solving difficult problems and a 26% increase in accuracy for students with LD solving simpler problems. The study highlights the educational implications of incorporating digital tools like the digital pencil in learning and assessment environments, emphasizing the need for tailored instructional strategies to support diverse learners.

This study examines the relationships between three accommodation strategies—Extended Time (ET) only, breaks only, and breaks bundled with ET—and academic performance, test-taking behavior, and attitudes among eighth-grade students with disabilities who participated in the 2017 National Assessment of Educational Progress (NAEP) mathematics assessment. Utilizing propensity score analysis to mitigate selection bias, the study finds that students receiving only breaks tend to exhibit lower performance and engagement compared to those receiving ET alone or those receiving breaks bundled with ET. In addition, results suggest that bundling breaks with ET is associated with lower performance compared to ET alone. These findings highlight the complexities of interactions between accommodations and support the need for reevaluating current accommodation policies to enhance their effectiveness for students with disabilities.

Mental rotation (MR), a key aspect of spatial reasoning, is highly predictive of success in STEM fields. This study analyzed strategies employed by 27,600 eighth-grade students during a digital MR task from the 2017 National Assessment of Educational Progress (NAEP) in mathematics. Utilizing K-means cluster analysis to categorize behavioral and performance patterns, we identified four distinct profiles: Cognitive Offloaders (15% of the sample), Internal Visualizers (55%), External Visualizers (5%), and Non-Triers (25%). Cognitive Offloaders, skilled at minimizing cognitive load by eliminating incorrect options, demonstrated the highest MR accuracy rates at 45%. Internal Visualizers, relying less on digital tools and more on mental strategies, achieved robust performance with an average score of 38%. External Visualizers, despite their extensive use of assistive tools and greater time investment, scored an average of 36%. Non-Triers showed minimal engagement and correspondingly the lowest performance, averaging 29%. These findings not only underscore the diverse strategies students adopt in solving MR tasks but also emphasize the need for educational strategies that are tailored to accommodate different cognitive styles. By integrating MR training into the curriculum and enhancing teacher preparedness to support diverse learning needs, this study advocates for educational reforms to promote equitable outcomes in mathematics and broader STEM fields.

This study investigates the relationship between text-to-speech (TTS) usage and item-by-item performance in the 2017 eighth-grade National Assessment of Educational Progress (NAEP) math assessment, focusing on students with disabilities (SWDs), English language learners (ELLs), and their general education (GE) peers. Results indicate that all students use TTS more for longer and more difficult math items as well as for multiple-choice or short-response formats. Among SWDs and GE students, lower math proficiency and higher perceived time pressure are linked to higher TTS usage. Moreover, among GE students, factors such as male gender, minority status, lower math persistence, and higher math interest and effort during testing contribute to higher TTS usage. TTS usage is positively associated with item performance for SWDs and ELLs who received extended time accommodations but not for those who did not receive such accommodations or for general education students. The study suggests that the time constraints of speeded digital assessments may limit the potential benefits of TTS for SWDs and ELLs in math problem-solving.

This empirical research study investigates the relationship between the utilization of Universal Design (UD) elements and math performance among eighth graders. We analyzed 2017 National Assessment of Educational Progress process data using Poisson Generalized Linear Mixed-Effects Models to examine how the frequency of UD element usage varies across diverse student demographics. Our findings reveal a great divide: lower achievers, students with disabilities, those from lower socioeconomic backgrounds, English Language Learners, and African American students predominantly use text-to-speech, whereas high achievers more frequently employ digital pencils (which allow for manual writing or drawing on the screen) and elimination capacity (which enables students to dismiss incorrect multiple-choice answers). Linear Mixed-Effects Models show that higher levels of utilization of digital pencils and elimination capacity are correlated with improved math performance across various student subgroups. However, frequent changes in color themes, intended to enhance visual differentiation, were associated with lower performance, suggesting that these might serve as distractions rather than aids for some students. We discuss the implications of the findings in terms of the design of UD elements, as well as for equity, accessibility, and math education more broadly.

This study assesses the capabilities of OpenAI’s ChatGPT-4 and ChatGPT-4o in solving mathematics problems from the National Assessment of Educational Progress (NAEP) across grades 4, 8, and 12. Results indicate that ChatGPT-4o slightly outperform ChatGPT-4 and both models generally surpass U.S. students’ performance across all grades, content areas, item type, and difficulty level. However, both models perform worse on geometry and measurement than on algebra and face more difficulties with high-difficulty mathematics items. This investigation highlights the strengths and limitations of AI as a supplementary educational tool, pinpointing areas for improvement in spatial intelligence and complex mathematical problem-solving. These findings suggest that while AI has the potential to support instruction in specific mathematical areas like algebra, there remains a need for careful integration and teacher-mediated strategies in areas where AI is less effective.

For autistic students receiving special education services, little is known about their relative strengths, weaknesses, and enjoyment across different math content areas; their overall math interest and persistence are also not well-studied. Using the 2017 eighth-grade National Assessment of Education Progress data, this study finds, relative to general education peers with the same math proficiency level, autistic students scored higher and exhibited faster speed in solving visuospatial problems (e.g. identifying figures), but scored lower on math word problems with complex language or social context. Autistic students reported a higher level of enjoyment in solving math problems related to finding areas of shapes or figures but a lower level of persistence than their non-autistic, general education peers. Our work points out the need to help autistic students overcome their weaknesses in word problems and develop their mathematical persistence.

This study analyzed performance, process, and survey data of eighth graders with learning disabilities (LDs) who took the 2017 National Assessment of Educational Progress (NAEP) digital math test. Compared with students with LDs who did not receive extended time accommodations (ETAs), students with LDs who received and used ETA scored significantly higher on the test, whereas students with LDs who received but did not use ETA scored significantly lower on the test. In addition, students with LDs in the two ETA groups reported a lower level of perceived time pressure and a higher level of math interest and enjoyment than their peers who did not receive ETA. For students with LDs who received ETA, optimal performance was achieved with 50% additional time, while their peers who did not receive ETA typically performed best when utilizing most of their allotted time. The analysis of process data revealed that students with LDs who used ETA performed more actions, had a higher number of revisits, used universal design digital tools more frequently, and performed better on time-consuming items than their peers who did not receive ETA at the same level of math performance.

ABSTRACT The recent surge of online language learning services in the past decade has benefitted second language learners. However, there is a lack of understanding of whether learners, especially young learners, are engaged in online learning, and how educators can enhance the engagement of the online learning experience. This study examines an artificial intelligence (AI)- powered automated system that uses voice and facial recognition to track both teacher and learner speech, facial expressions, and interactions in real-time in a one-to-one 25-minute online English class. Each learner completed a learner engagement survey within 72 hours of the online class. Results demonstrated that young learners were highly engaged during this one-to-one online learning setting (mean = 4.5, out of 5). Learners’ frontal face exposure (indicating their attentiveness during class) and English proficiency levels are significant and positive predictors of learner engagement. Teachers’ total length of speech and instructional time tended toward significance in predicting learner engagement. Educational implications are discussed.

Background: Diffuse midline gliomas (DMGs) are a universally fatal brain tumor of childhood. While histone mutations are a critical tumor initiating event, they are insufficient to drive gliomagenesis. Histone mutations co-occur with somatic alterations in other pathways including TP53, MAPK, and MYC signaling. However, the mechanisms through which these pathways are activated have not been fully elucidated. Methods: We applied an integrative approach using transcriptomics, epigenetics, proteomics, in vitro cancer models, and in vivo mouse models to systematically evaluate how FOXR2 mediates gliomagenesis. Results: We have recently found that a subset of DMGs aberrantly express FOXR2, a forkhead transcription factor. FOXR2 is both sufficient to enhance tumor formation, and necessary for FOXR2-expressing DMGs. While FOXR2 indeed enhances MYC protein stability, FOXR2 exerts oncogenesis through MYC-independent functions and specifically hijacks E26-transformation specific (ETS) transcriptional circuits and FOXR2 DNA-binding is highly enriched at ETS motifs. We have performed proteomic and phospho-proteomic analysis of FOXR2-expressing human neural stem cells to identify proteins and phospho-sites that are highly enriched in FOXR2-expressing cells. Conclusion: Taken together, this study elucidates how FOXR2 interacts with ETS transcription factors to mediate oncogenesis, and further highlights a role for FOXR2 in activating ETS and MAPK signaling. Citation Format: Jessica W. Tsai, Paloma Cejas, Marissa Coppola, Dayle K. Wang, Smruti Patel, David W. Wu, Phonepasong Arounleut, Xin Wei, Ningxuan Zhou, Sudeepa Syamala, Frank P. Dubois, Kristine Pelton, Jayne Vogelzang, Cecilia Sousa, Audrey Baguette, Xiaolong Chen, Alexandra L. Condurat, Sarah E. Dixon-Clarke, Annarah Charles, Kevin N. Zhou, Sophie D. Lu, Elizabeth M. Gonzalez, Madison S. Chacon, Jeromy J. Digiacomo, Rushil Kumbhani, Dana Novikov, Maria Tsoli, David S. Ziegler, Uta Dirksen, Natalie Jager, Gnana Prakash Balasubramanian, Christof M. Kramm, Michaela Nathrath, Stefan Bielack, Suzanne J. Baker, Jinghui Zhang, James M. McFarland, Gad Getz, Francois Aguet, Nada Jabado, Olaf Witt, Stefan M. Pfister, Keith L. Ligon, Volker Hovestadt, Claudia Kleinman, Henry Long, David T. Jones, Pratiti Bandopadhayay, Timothy N. Phoenix. Dissecting mechanisms underlying FOXR2-mediated gliomagenesis in diffuse midline gliomas. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3562.

Information and Communications technology has been integrated into education for decades, 12 offering accessible ways to reach resources and revolutionizing the teaching and learning 13 paradigm (Huang et al., 2021;Teo, 2009;Wang, 2020). Educational technology (EdTech) 14 provides opportunities to sustain teaching and learning, even in emergency remote learning 15 settings necessitated by the COVID-19 pandemic. However, the effectiveness of EdTech Collectively, these articles underscore that there is no one-size-fits-all approach to EdTech.

Aberrant expression of long non-coding RNAs (lncRNAs) that results in sustained activation of cell growth promoting pathways is an important mechanism in driving prostate cancer progression. In the present study, we explored differentially expressed lncRNAs in two microarray datasets of prostate benign and malignant tissues. We found that MAGI2-AS3 was one of the most downregulated lncRNAs in prostate tumors, which was further confirmed in our collected clinical samples. The function assays showed that MAGI2-AS3 overexpression decreased cell viability and led to obvious cell apoptosis in PC-3 and DU145 prostate cancer cells. Elevation of MAGI2-AS3 decreased the activity of STAT3 in PC-3 and DU145. In addition, microRNA-424-5p (miR-424-5p), a positive regulator of STAT3 pathway, was predicted as a target of MAGI2-AS3, furthermore, the interaction between MAGI2-AS3 and miR-424-5p was confirmed via reverse-transcript polymerase chain reaction (RT-qPCR), dual luciferase reporter assay and RNA immunoprecipitation (RIP). MAGI2-AS3 upregulated miR-424-5p and downregulated COP1 in PC-3 and DU145. More importantly, IL6-induced activation of STAT3 pathway could attenuate the biological effect of MAGI2-AS3 in PC-3 and DU145. In clinical samples, MAGI2-AS3 levels were negatively correlated with miR-424-5p expression, while positively correlated with COP1 mRNA expression. Altogether, the current study revealed MAGI2-AS3 as a novel negative regulator of prostate cancer development.

Background: Understanding how aberrant transcription factors (TFs) hijack normal development to induce oncogenesis is a critical question in oncology. Forkhead box (FOX) proteins are a superfamily of transcriptional regulators characterized by a forkhead DNA-binding domain. Within this family, Forkhead Box R2 (FOXR2) expression has been associated with a subset of cancers including CNS and peripheral neuroblastoma. While FOXR2 has been shown to stabilize MYC isoforms, the mechanistic details through which it enhances tumor formation and the true extent of its role as an oncogene across all cancers have not been systematically evaluated. Methods: We applied an integrative approach using transcriptomics, epigenetics, in vitro cancer models, and in vivo mouse models to systematically evaluate the mechanisms by which FOXR2 is activated across human cancers. Results: We performed a pan-cancer analysis of FOXR2 activation across over 10,000 adult and pediatric cancer samples, and found FOXR2 to be aberrantly upregulated in 70% of all cancer types, and 8% of all individual tumors. We identified genetic and epigenetic mechanisms that induce its expression, including hypomethylation of a novel promoter in the vast majority (78%) of FOXR2-expressing cases. We demonstrate that FOXR2 expression is both sufficient and necessary for transformation across multiple lineages, using both in vitro and in vivo models. Conclusion: Taken together, this study demonstrates the role of FOXR2 as a potent oncogene across human cancers, and highlights a novel mechanism by which its expression is activated. Citation Format: Jessica W. Tsai, Paloma Cejas, Dayle K. Wang, Smruti Patel, David W. Wu, Phonepasong Arounleut, Xin Wei, Ningxuan Zhou, Sudeepa Syamala, Frank P. Dubois, Kristine Pelton, Jayne Vogelzang, Cecilia Sousa, Audrey Baguette, Xiaolong Chen, Alexandra L. Condurat, Sarah E. Dixon-Clarke, Kevin N. Zhou, Sophie D. Lu, Elizabeth M. Gonzalez, Madison S. Chacon, Jeromy J. Digiacomo, Rushil Kumbhani, Dana Novikov, J'Ya Hunter, Maria Tsoli, David S. Ziegler, Uta Dirksen, Natalie Jager, Gnana Prakash Balasubramanian, Christof M. Kramm, Michaela Nathrath, Stefan Bielack, Suzanne J. Baker, Jinghui Zhang, James M. McFarland, Gad Getz, Francois Aguet, Nada Jabado, Olaf Witt, Stefan M. Pfister, Keith L. Ligon, Claudia Kleinman, Henry Long, David T. Jones, Pratiti Bandopadhayay, Timothy N. Phoenix. FOXR2 is an oncogenic driver across adult and pediatric cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5730.

Abstract BACKGROUND: Understanding how aberrant transcription factors (TFs) hijack normal development to induce oncogenesis is a critical question in oncology. Forkhead box (FOX) proteins are a superfamily of transcriptional regulators characterized by a forkhead DNA-binding domain. Within this family, Forkhead Box R2 (FOXR2) has been identified as a candidate structural variant (SV) driver in a subset of pediatric cancers including CNS embryonal tumors and peripheral neuroblastoma. While FOXR2 has been shown to stabilize MYC isoforms, the mechanistic details through which it enhances tumor formation, other non-SV mechanisms of activating aberrant expression, and the true extent of its role as an oncogene across all cancers have not been systematically evaluated. METHODS: We applied an integrative approach using transcriptomics, epigenetics, in vitro cancer models, and in vivo mouse models to systematically evaluate the mechanisms by which FOXR2 is activated across human cancers. RESULTS: We performed a pan-cancer analysis of FOXR2 activation across over 10,000 adult and pediatric cancer samples, and surprisingly found FOXR2 to be aberrantly upregulated in 70% of all cancer types (including diffuse midline gliomas), and 8% of all individual tumors. FOXR2 expression occurred predominantly in the absence of rearrangement/fusions, single nucleotide variants, or copy number aberrations at the DNA level. Transcriptomic and epigenomic analyses show the vast majority of tumors (78%) aberrantly express FOXR2 through a previously undescribed epigenetic mechanism via hypomethylation of a novel promoter. Using both in vitro and in vivo models, we demonstrate that FOXR2 expression is both sufficient and necessary for transformation across multiple lineages, including DMGs. CONCLUSION: Taken together, this study demonstrates that FOXR2 is a novel and potent oncogene across pediatric and adult cancers, and highlights a new epigenetic mechanism by which its expression is activated.

This study examines differential effects of the Cognitive Behavioral Intervention for Trauma in Schools (CBITS) program on behavioral and academic outcomes of middle school students. Researchers administered screenings to grade 6 students to assess traumatic stress and then randomized those with elevated levels to the CBITS treatment (n = 150; 47% female) or comparison group (n = 143; 53% female). Analyses examined the overall impact of CBITS and differential effects among subpopulations of students who reported clinically significant externalizing (n = 75; 67% female) or internalizing behavior (n = 185; 53% female) at baseline. Overall, students who received CBITS reported significantly reduced post-traumatic stress symptoms and marginally significant improvements in internalizing symptoms. Relative to counterparts in the comparison group, students exhibiting externalizing behaviors in the CBITS group reported significantly reduced post-traumatic stress, dissociation, anger, internalizing and total behavior problems, and also significantly improved scores on a standardized literacy assessment at posttest and follow-up. Students with internalizing behavior problems showed differential academic effects at 1-year follow-up; those in CBITS did significantly better on standardized math tests.