Emily M. Hastings, Ph.D.

Emily M. Hastings, Ph.D.

Assistant Professor of Computer Science at the University of Wisconsin-Eau Claire (she/her)

Research

My overarching research interest lies in using technology to create educational tools, particularly for CS or broader STEM education. My dissertation research focused on the use of algorithmic team formation tools in courses implementing team-based approaches to learning. I also worked on a project for my fellowship with NIST on quantifying human skill level from historical data, which is useful both in the context of team formation and in other disciplines such as manufacturing maintenance.

I have also previously conducted research outside of CS for my undergraduate minor in Renaissance and Medieval Studies, and I remain interested in similar work and in projects combining CS with other academic fields or creative endeavors (e.g., digital humanities, e-textiles, etc.).

Current Projects

I am currently establishing my research agenda at UW-Eau Claire. Check back for updates!

Past Projects

Learner-Centered Algorithmic Team Formation

Fall 2016 – 2024
_Advisors: Brian Bailey, Karrie Karahalios, Darko Marinov _

Instructors are increasingly adopting team-based learning in their courses in a response to demands from industry. One of the first decisions instructors encounter when adopting this approach is how exactly the teams should be formed. A popular approach supported by the literature is a criteria-based one, where students are grouped into teams according to criteria such as demographics, skills, and working styles. As course enrollments grow, instructors are increasingly turning to algorithmic tools like CATME to implement this team formation approach in their courses. However, not much is known about how to most effectively use these tools to support student learning and promote good teamwork experiences.

This line of work examines stakeholder perceptions of algorithmic team formation tools, identifies strengths and weaknesses of current tools, and aims to create a more learner-centered approach to algorithmic team formation, in which students have more control over a previously-inaccessible process. Now funded by the NSF!

Associated Presentations:

Associated Publications:

Quantifying Maintenance Technician Capability

Summer 2018 – 2020
Advisors: Michael Brundage, Rachael Sexton

During my time as a GMSE fellow, I worked with the Knowledge Extraction and Application team in NIST’s Engineering Laboratory toward quantifying human skill level from historical data. The team is using human-in-the-loop machine learning to make analysis of jargon- and misspelling-filled maintenance work order documents (MWOs) feasible. Using the toolkit Nestor, users can classify and annotate the words used in a set of MWOs to produce a clean set of tags that can be used in further analysis.

My early work on the project related to assessing the validity of the approach, with an ultimate goal of using the tags generated by Nestor to estimate the skill level of maintenance technicians with the various tasks and machines mentioned in the MWOs. These estimates can then be used to recommend training, assign technicians to jobs, or form teams. I hope to generalize this work to the educational contexts I study in my dissertation research.

My later work on the project involved the formulation of a data-driven framework for team formation tasks in the maintenance domain, generalized from my LIFT workflow for student teams in academia.

Associated Publications:

Beyond the Black Box: Understanding & Designing for Users’ Expectations of Algorithms

January 2018 – May 2019
Advisors: Karrie Karahalios, Christian Sandvig

During my time on this project, I worked with a team across multiple universities to plan a large-scale survey aimed at better understanding people’s awareness of and literacy about algorithms, especially those of social media platforms like Facebook and Twitter. I specifically contributed to the design of a sub-section of the larger survey which would assess the effects of different visual design cues on algorithmic awareness.

Knoxel: Teaching Introductory Programming with Minecraft

Summer 2015 - March 2016
Advisor: Jaime Spacco

During my ASSET scholarship, I worked with a team to develop an educational plugin for the game Minecraft that helps teach introductory computer science students to program in Java and similar languages. One of the main objectives of Minecraft is to build various structures out of blocks. Our project, which we named Knoxel (formerly Knoxcraft), forces students to do this by writing code in Java (or another language of their instructor’s choice—we support Java, Python, and Blockly, and have a framework for users to extend Knoxel to work with potentially any programming language). This environment allows students to become comfortable with the process of programming in a fun and visual context they may already be familiar with, and may help make computer science as a field more accessible to those groups who are currently underrepresented.

We piloted Knoxel in CS 141 during the Fall 2015 term and received very positive feedback from students. Since my time on the project has ended, Knoxel has grown to include a simple web interface for students who do not have or wish to buy the full Minecraft game.

Associated Presentations:

Responding to Manycore: Teaching Parallel Computing with Higher-Level Languages and Activity-Based Laboratories

Summer 2014
Advisor: David Bunde

This project was part of an NSF-funded collaboration to develop materials to teach parallel programming at undergraduate institutions, especially compelling examples to increase student interest.

My contribution to the project was an adventure game example called “Through the Mines,” in which players had to navigate a level full of enemies and obstacles as quickly as possible. The goal of this example was to demonstrate that in this context, since the game had to run slow enough for humans to play, the increased computing power gained from parallelization was better used to increase the difficulty of the game by making enemies smarter.

My presentation of this work at Knox’s Summer Science Seminar Series won the Best Student Seminar Award for the 2014 season.

Associated Presentations:

Task Mapping for Emerging Network Topologies

Summer 2013
Advisor: David Bunde

This work was part of an NSF-funded project aiming to improve the performance of large-scale scientific simulations on high-performance computing systems through new strategies for task mapping, the assignment of parts of an application to specific parts of a machine. For more information, see here.

I worked with a partner to investigate different cabling methods for the then-new Dragonfly interconnect topology and their effects on task mapping on Dragonfly systems.

Associated Presentations:

Associated Publications:

The History and Construction of Elizabethan English Costume

Spring 2013, Spring 2016
Advisor: Margo Shively

In an independent research project bookending my time at Knox, I researched the dress of English women of the 1560s-70s, from both the working- and upper-classes. This research took the form of examining both primary sources (mostly art) and the work of costume scholars such as Janet Arnold, Margo Anderson, and Drea Leed. I then used the knowledge I had gained to design and construct an outfit representative of each class, consisting of garments from the skin out. The techniques and materials used were as historically accurate as I could afford. This project offered an extraordinary opportunity to step into the shoes (in a nearly literal sense) of the subjects of Elizabethan portraits, and taught me skills I still use in my creative projects.

Associated Presentations:

Associated Projects: