Search EdWorkingPapers

Search EdWorkingPapers by author, title, or keywords.

Heather C. Hill

Dorottya Demszky, Jing Liu, Heather C. Hill, Shyamoli Sanghi, Ariel Chung.

While recent studies have demonstrated the potential of automated feedback to enhance teacher instruction in virtual settings, its efficacy in traditional classrooms remains unexplored. In collaboration with TeachFX, we conducted a pre-registered randomized controlled trial involving 523 Utah mathematics and science teachers to assess the impact of automated feedback in K-12 classrooms. This feedback targeted “focusing questions” – questions that probe students’ thinking by pressing for explanations and reflection. Our findings indicate that automated feedback increased teachers’ use of focusing questions by 20%. However, there was no discernible effect on other teaching practices. Qualitative interviews revealed mixed engagement with the automated feedback: some teachers noticed and appreciated the reflective insights from the feedback, while others had no knowledge of it. Teachers also expressed skepticism about the accuracy of feedback, concerns about data security, and/or noted that time constraints prevented their engagement with the feedback. Our findings highlight avenues for future work, including integrating this feedback into existing professional development activities to maximize its effect.

More →


Zid Mancenido, Heather C. Hill, Jeannette Garcia Coppersmith, Hannah Carter, Cynthia Pollard, Chris Monschauer.

Practice-based teacher education has increasingly been adopted as an alternative to more traditional, conceptually-focused pedagogies, yet the field lacks causal evidence regarding the relative efficacy of these approaches. To address this issue, we randomly assigned 185 college students to one of three experimental conditions reflective of common conceptually-focused and practice-based teacher preparation pedagogies. We find significant and large positive effects of practice-based pedagogies on participants’ skills in eliciting and responding to student thinking as demonstrated through a written assessment and a short teaching episode. Our findings contribute to a developing evidence base that can assist policymakers and teacher educators in designing effective teacher preparation at scale.

More →


Zachary Himmelsbach, Heather C. Hill, Jing Liu, Dorottya Demszky.

This study provides the first large-scale quantitative exploration of mathematical language use in U.S. classrooms. Our approach employs natural language processing techniques to describe variation in the use of mathematical language in 1,657 fourth and fifth grade lessons by teachers and students in 317 classrooms in four districts over three years. Students’ exposure to mathematical language varies substantially across lessons and between teachers. Students whose teachers use more mathematical language are more likely to use it themselves, and they perform better on standardized tests. These findings suggest that teachers play a substantial role in students’ mathematical language use.

More →


Dorottya Demszky, Jing Liu, Heather C. Hill, Dan Jurafsky, Chris Piech.

Providing consistent, individualized feedback to teachers is essential for improving instruction but can be prohibitively resource-intensive in most educational contexts. We develop M-Powering Teachers, an automated tool based on natural language processing to give teachers feedback on their uptake of student contributions, a high-leverage dialogic teaching practice that makes students feel heard. We conduct a randomized controlled trial in an online computer science course (n=1,136 instructors), to evaluate the effectiveness of our tool. We find that M-Powering Teachers improves instructors’ uptake of student contributions by 13% and present suggestive evidence that it also improves students’ satisfaction with the course and assignment completion. These results demonstrate the promise of M-Powering Teachers to complement existing efforts in teachers’ professional development.

More →


Dorottya Demszky, Heather C. Hill.

Classroom discourse is a core medium of instruction --- analyzing it can provide a window into teaching and learning as well as driving the development of new tools for improving instruction. We introduce the largest dataset of mathematics classroom transcripts available to researchers, and demonstrate how this data can help improve instruction. The dataset consists of 1,660 45-60 minute long 4th and 5th grade elementary mathematics observations collected by the National Center for Teacher Effectiveness (NCTE) between 2010-2013. The anonymized transcripts represent data from 317 teachers across 4 school districts that serve largely historically marginalized students. The transcripts come with rich metadata, including turn-level annotations for dialogic discourse moves, classroom observation scores, demographic information, survey responses and student test scores. We demonstrate that our natural language processing model, trained on our turn-level annotations, can learn to identify dialogic discourse moves and these moves are correlated with better classroom observation scores and learning outcomes. This dataset opens up several possibilities for researchers, educators and policymakers to learn about and improve K-12 instruction.

The data and its terms of use can be accessed here: https://github.com/ddemszky/classroom-transcript-analysis

More →


Sterling Alic, Dorottya Demszky, Zid Mancenido, Jing Liu, Heather C. Hill, Dan Jurafsky.

Responsive teaching is a highly effective strategy that promotes student learning. In math classrooms, teachers might funnel students towards a normative answer or focus students to reflect on their own thinking, deepening their understanding of math concepts. When teachers focus, they treat students’ contributions as resources for collective sensemaking, and thereby significantly improve students’ achievement and confidence in mathematics. We propose the task of computationally detecting funneling and focusing questions in classroom discourse. We do so by creating and releasing an annotated dataset of 2,348 teacher utterances labeled for funneling and focusing questions, or neither. We introduce supervised and unsupervised approaches to differentiating these questions. Our best model, a supervised RoBERTa model fine-tuned on our dataset, has a strong linear correlation of .76 with human expert labels and with positive educational outcomes, including math instruction quality and student achievement, showing the model’s potential for use in automated teacher feedback tools. Our unsupervised measures show significant but weaker correlations with human labels and outcomes, and they highlight interesting linguistic patterns of funneling and focusing questions. The high performance of the supervised measure indicates its promise for supporting teachers in their instruction.

More →


Heather C. Hill, Virginia S. Lovison.

In recent decades, U.S. education leaders have advocated for more intellectually ambitious mathematics instruction in classrooms. Evidence about whether more ambitious mathematics instruction has filtered into contemporary classrooms, however, is largely anecdotal. To address this issue, we analyzed 93 lessons recorded by a national random sample of middle school mathematics teachers. We find that lesson quality varies, with the typical lesson containing some elements of mathematical reasoning and sense-making, but also teacher-directed instruction with limited student input. Lesson quality correlates with teachers’ use of a textbook and with teachers’ mathematical background. We consider these findings in light of efforts to transform U.S. mathematics instruction.

More →


Kathryn E. Gonzalez, Kathleen Lynch, Heather C. Hill.

Despite growing evidence that classroom interventions in science, technology, engineering, and mathematics (STEM) can increase student achievement, there is little evidence regarding how these interventions affect teachers themselves and whether these changes predict student learning. We present results from a meta-analysis of 37 experimental studies of preK-12 STEM professional learning and curricular interventions, seeking to understand how STEM classroom interventions affect teacher knowledge and classroom instruction, and how these impacts relate to intervention impacts on student achievement. Compared with control group teachers, teachers who participated in STEM classroom interventions experienced improvements in content and pedagogical content knowledge and classroom instruction, with a pooled average impact estimate of +0.56 standard deviations. Programs with larger impacts on teacher practice yielded larger effects on student achievement, on average. Findings highlight the positive effects of STEM instructional interventions on teachers, and shed light on potential teacher-level mechanisms via which these programs influence student learning.

More →


Dorottya Demszky, Jing Liu, Zid Mancenido, Julie Cohen, Heather C. Hill, Dan Jurafsky, Tatsunori Hashimoto.

In conversation, uptake happens when a speaker builds on the contribution of their interlocutor by, for example, acknowledging, repeating or reformulating what they have said. In education, teachers' uptake of student contributions has been linked to higher student achievement. Yet measuring and improving teachers' uptake at scale is challenging, as existing methods require expensive annotation by experts. We propose a framework for computationally measuring uptake, by (1) releasing a dataset of student-teacher exchanges extracted from US math classroom transcripts annotated for uptake by experts; (2) formalizing uptake as pointwise Jensen-Shannon Divergence (pJSD), estimated via next utterance classification; (3) conducting a linguistically-motivated comparison of different unsupervised measures and (4) correlating these measures with educational outcomes. We find that although repetition captures a significant part of uptake, pJSD outperforms repetition-based baselines, as it is capable of identifying a wider range of uptake phenomena like question answering and reformulation. We apply our uptake measure to three different educational datasets with outcome indicators. Unlike baseline measures, pJSD correlates significantly with instruction quality in all three, providing evidence for its generalizability and for its potential to serve as an automated professional development tool for teachers.

More →


Heather C. Hill, Anna Erickson.

Poor program implementation constitutes one explanation for null results in trials of educational interventions. For this reason, researchers often collect data about implementation fidelity when conducting such trials. In this article, we document whether and how researchers report and measure program fidelity in recent cluster-randomized trials. We then create two measures—one describing the level of fidelity reported by authors and another describing whether the study reports null results—and examine the correspondence between the two. We also explore whether fidelity is influenced by study size, type of fidelity measured and reported, and features of the intervention. We find that as expected, fidelity level relates to student outcomes; we also find that the presence of new curriculum materials positively predicts fidelity level.

More →