The University of Michigan Institute for Firearm Injury Prevention seeks a motivated and skilled master's level student to assist Dr. Heinze and Dr. Thulin’s research of technology-facilitated reporting systems (e.g., anonymous reporting systems, confidential reporting systems).

The purpose of the current study is to analyze approximately 40,000 adolescent-submitted tips about themselves or their peers to the state-wide Say Something Anonymous Reporting System (SS-ARS) in North Carolina. The goal of this role is to leverage large language models in a secure environment to characterize tips across a variety of features.

The student should excel at professional communication and collaboration, pay attention to details, be able to work independently and meet deadlines, and create consistently high-quality products. They will gain experience working with faculty, postdoctoral fellows, and staff, with mentorship and career development opportunities.

This is a summer position with flexible working hours and the opportunity to make significant contributions to a crucial area of research. Interested candidates should submit their resume and a cover letter detailing their relevant experience and interest in the project.
 

Role and Responsibilities:

- Utilize the UMGPT Toolkit to perform classification tasks on the tip line data.

- Engage in prompt engineering to refine and optimize input prompts for the UMGPT model.

- Interpret model fit statistics to evaluate the performance and accuracy of classification tasks.

- Establish an efficient workflow for managing and processing a large dataset.

- Collaborate with the research team to develop insights and contribute to the overall aims of the project.

- Participate in regular project meetings to discuss progress and troubleshoot challenges.

- Document methodologies and maintain accurate records of analyses and findings.

- Contribute to draft reports or manuscripts, with opportunities for authorship based on contributions to the project.
 

Work Schedule:
This position is for a hybrid or remote work schedule.

Time-Frame: ASAP? August 31, 2025 with potential for extension in the fall depending on grant funding. 

Qualifications and Education Requirements:

- Currently enrolled in a graduate program in a relevant field (e.g., data science, computer science, information, psychology, or education).

- Experience with the UMGPT Toolkit or similar natural language processing tools.

- Proficiency in prompt engineering and optimizing input data for machine learning models.

- Strong understanding of statistical concepts and experience interpreting model fit statistics.

- Excellent problem-solving with great attention to detail.

- Strong written and verbal communication skills in English, with the ability to communicate findings and collaborate with team members and external partners.

- Strong organizational skills and the ability to manage multiple tasks within specified timelines.

- Clear articulation of ideas, rationalization of choices, and deduction based on information.

- Excellent interpersonal skills and demonstrated organizational and project management abilities.

- Capacity to work independently, meet deadlines, identify resources and solutions for code issues, and manage multiple tasks within specified timelines.

- Familiarity with online/virtual work environments and strong work ethic and professionalism.