Projects Overview

In short, this efficacy study seeks to test the efficacy of From Here to There (FH2T) compared to two other conditions and a baseline condition.

More details of this project can be found by clicking on the link to the grant above.

My roles in this project include generating the approximately 4,000 IDs used by the students and teachers to access their appropriate experimental condition, coordinating a team of four people to successfully input the IDs into the correct databases, assisting in troubleshooting efforts to ensure that all four conditions function as intended, generation of data visualizations and summary information for weekly fidelity updates to be sent to teachers weekly, and other support tasks.

Teachers invest massive amounts of their time into finding the best ways to help their students learn. As more students shift to using online learning technologies, it becomes increasingly challenging for teachers to keep track of how students are performing while using these resources. While many educational technologies collect rich log data from student interactions, teachers often only have access to very basic information, such as what problems students have completed, whether they got the problem correct, and how long the problem took them to solve. Although more complex data are sometimes available, such as students' problem solving processes and errors, this information is usually not accessible to teachers in a usable way (Mandinach & Jimerson, 2016). Using that data to make informed actionable decisions in their classrooms requires teachers to spend more time processing and understanding that data so that they can act on it to help their students. Unfortunately, despite the dramatic shift toward e-learning within the past year, there is a dearth of basic research on how teachers interact with educational data, dashboards, and interfaces, particularly those that provide teachers with rich process information about students' mathematical understanding.

My research goal is to conduct the basic research of what metrics are the most useful for mathematics teachers and how teachers can best access those metrics so that educational technology developers can optimize their interfaces for teacher use. While there is some research on what information teachers would like to have accessible to them (Holstein et al., 2017), to my knowledge, no research investigates the fundamental question of whether the metrics that teachers request are empirically predictive of learning, or how teachers interact with student data. Until teachers are able to appropriately use metrics provided to them, the effectiveness of these systems for improving classroom instruction will be limited.

This project seeks to determine if the font that mathematics is typed in likely impacts correctness (accuracy) or reaction time of college undergraduate students. We compared Times New Roman, Kalam (the font used in Graspable Math), and a sample of handwriting from one student. The idea here being that if students were particularly bad at one font, we would make the recommendation to not type math in that font. Luckily, the majority of the results came back as non-significant which means that font type likely doesn't impact correctness (accuracy) or reaction time for college aged students.

My roles for this project include literature review, content development, preliminary analysis, and coauthoring the manuscript under the guidance of post-doc Dr. Jenny Yun-Chen Chan.

Prior research has shown negative relations between math anxiety and math performance (Foley et al., 2017). One potential pathway through which math anxiety influences math performance is through help seeking behavior during learning. Here, we examine whether middle school students’ behavior, specifically the frequency of hint requests, within educational technologies mediates the association between math anxiety and math performance. We examine the mediation in two online math learning technologies: From Here to There (FH2T) and problem sets in ASSISTments. FH2T is a dynamic math game, and ASSISTments is an online learning platform that features problem sets with immediate hints and feedback. We examined whether the frequency of hint requests mediates the effects of math anxiety on performance in both conditions. Unlike past work, we found that math anxiety was not a predictor of hint usage or post test math performance in either conditions. Our findings also indicated that students who used more hints in the problem set condition scored lower on the math post-test. This relationship is not significant in the FH2T condition, suggesting a fundamental difference in hint usage between the two technologies.

In short, this bar lab study sought to investigate factors that influence alcohol consumption including childhood trauma, impulsivity, stress, and social factors.

More details of this project can be found by clicking on the link to the grant above.

My roles in this project included managing the majority of the technology used in the study (including laptops, desktops, wearable tech, and coordinating with ASU's tech team), training the eligibility interview team, developing and refining the interview protocol and procedures, administering tasks (digit span, AUDADIS, Shipley, BART/X-BART*, MRBURNS, timeline follow back) as a member of the interview team, website management, and social media management for participant recruitment. Additionally, I assisted in the authoring of several poster presentations and one manuscript (eds.).

The X-BART (eXpected pumps Balloon Analogue Risk Task; modified from Pleskac and Wershbale, 2014) is a modified version of the Balloon Analogue Risk Task (BART; Lejuez, 2002) in which participants estimate the number of pumps they plan to pump the next balloon before they do so on the next screen. The utility of the X-BART is that it appears to be a more strongly correlated to affect-free impulsivity (i.e. lack of forethought) than the original BART. For more information on the specifics of this functionality, please see Can we learn new things about state impulsive choice and drinking by using a modified balloon analogue risk task - BART?

This project was the first real research project that was my original work. On my own I found a suitable BART base program to modify, modified the ePrime code, and, under the guidance of Dr. Julie Patock-Peckham, tested it in her NIH study. With Dr. Patock's support I constructed the literature review, analyzed the results, presented a paper, and am collaborating on a manuscript.