More Like this

1. Abstract Graph Neural Networks (GNNs) have shown satisfying performance on various graph learning tasks. To achieve better fitting capability, most GNNs are with a large number of parameters, which makes these GNNs computationally expensive. Therefore, it is difficult to deploy them onto edge devices with scarce computational resources, e.g., mobile phones and wearable smart devices. Knowledge Distillation (KD) is a common solution to compress GNNs, where a light-weighted model (i.e., the student model) is encouraged to mimic the behavior of a computationally expensive GNN (i.e., the teacher GNN model). Nevertheless, most existing GNN-based KD methods lack fairness consideration. As a consequence, the student model usually inherits and even exaggerates the bias from the teacher GNN. To handle such a problem, we take initial steps towards fair knowledge distillation for GNNs. Specifically, we first formulate a novel problem of fair knowledge distillation for GNN-based teacher-student frameworks. Then we propose a principled framework named RELIANT to mitigate the bias exhibited by the student model. Notably the design of RELIANT is decoupled from any specific teacher and student model structures, and thus can be easily adapted to various GNN-based KD frameworks. We perform extensive experiments on multiple real-world datasets, which corroborates that RELIANT achieves less biased GNN knowledge distillation while maintaining high prediction utility. Open-source code can be found at https://github.com/yushundong/RELIANT.

   Dong, Yushun ; Zhang, Binichi ; Yuan, Yiling ; Zou, Na ; Wang, Qi ; Li, Jundong ( January 2023 , Proceedings of the 2023 SIAM International Conference on Data Mining (SDM))
2. Examining the effects of undergraduate STEM teacher recruitment and teacher education programs on preservice secondary science and mathematics teacher readiness and teacher performance assessment (edTPA) scores

   https://doi.org/10.1111/ssm.12498  

   Moon, Sungmin ; Carpenter, Stacey L. ; Hansen, Alexandria K. ; Bushong, Leslie ; Bianchini, Julie A. ( November 2021 , School Science and Mathematics)

   We examined the effects of teacher education on preservice secondary science and mathematics teacher readiness. The construct of teacher readiness consisted of three dimensions: (a) preservice teachers’ understanding of how to implement current standards, (b) their understanding of how to teach multilingual learners (MLs), and (c) their beliefs about their abilities and skills as educators. To determine teacher readiness, a subset of preservice teachers enrolled in six teacher education programs (TEPs) completed a pre‐ and post‐survey. We found that the undergraduate STEM teacher recruitment program was associated with developing preservice teachers’ understanding of standards‐based instruction but was not associated with the other two dimensions of teacher readiness. We also found that the post‐baccalaureate TEPs were effective in developing preservice teachers’ understanding of ML instruction, whereas the experimental baccalaureate program was not. Further, we found that teacher readiness did not predict participants’ performance assessment (edTPA) scores. Findings suggest that teacher educators and curriculum developers involved in undergraduate STEM teacher recruitment programs should better address language, literacy, and ML instruction. Both teacher recruitment and TEPs should consider ways to strengthen teacher efficacy. Finally, teacher educators should carefully consider how their coursework and field experiences help preservice teachers prepare for the edTPA.

    

   more » « less
3. Toward Automated Feedback on Teacher Discourse to Enhance Teacher Learning

   https://doi.org/10.1145/3313831.3376418  

   Jensen, Emily ; Dale, Meghan ; Donnelly, Patrick J. ; Stone, Cathlyn ; Kelly, Sean ; Godley, Amanda ; D'Mello, Sidney K. ( April 2020 , Proceedings of the ACM CHI Conference on Human Factors in Computing Systems (CHI 2020))
4. The function of teacher dashboards depends on the amount of time pressure in the classroom situation: Results from teacher interviews and an experimental study

   https://doi.org/10.1007/s42010-022-00156-9  

   van Leeuwen, Anouschka ; Rummel, Nikol ( January 2022 , Unterrichtswissenschaft)

   Teacher dashboards are visual displays that provide information to teachers about their learners. In this article, we address teacher dashboards in the context of computer-supported student collaboration in primary education. We examine the role of different types of dashboards for the specific purpose of aiding teachers in identifying which group of collaborating students is in need of support. This question is addressed using qualitative and quantitative approaches. First, an interview study is reported in which teachers’ views (n = 10) on and perceptions of the acceptability of different types of dashboards were examined. Then, the results of an experimental vignette study are reported, which built upon on the interview study, and in which teachers (n = 35) interacted with mirroring or advising dashboards. Together, the studies revealed that the classroom situation, such as differing levels of time pressure, plays an important role regarding what type of dashboard is beneficial for a teacher to use in the classroom. The theoretical contribution of our study lies in a conceptual and empirical investigation of the relation between teachers’ need for control and their perception of different types of dashboards. Our study also points to several practical implications and directions for future research. 

   more » « less
5. What's in a teacher test? Assessing the relationship between teacher licensure test scores and student STEM achievement and course-taking

   https://doi.org/10.1016/j.econedurev.2017.09.002  

   Goldhaber, Dan ; Gratz, Trevor ; Theobald, Roddy ( December 2017 , Economics of Education Review)