Research Associate MIT: Physics Education Research
RELATE.MIT.edu group, Prof. Dave Pritchard
The REsearch on
Learning Assessing and Tutoring Effectively group is studying
and improving assessment and learning in novel and impactful ways. Using
psychometrics and machine learning we’ve analyzed 180k administrations of the
Force Concept Inventory (FCI) to discover student misconceptions and then
find the degree to which each student or class harbors each misconception. Our misconceptions are weighted clusters of
wrong answers exhibiting an identifiable misconception that are preferentially
selected by the subset students who hold this misconception.
The postdoc will help set the research
agenda for our powerful new analysis tools, carry out the research with the
group, and write papers. Research directions include
characterizing the nature of misconceptions, studying the effects of various
pedagogies on their remediation, and placing them in a cognitive context. An
important short-time goal is a web site to analyze administrations of the FCI
to inform students and teachers on how deeply they hold each of the ~ 25
misconceptions that we have discovered and (given pre- and post-tests) as well
as the extent to which their instruction has reduced these levels. We also plan to probe the universality
of misconceptions on the FCI by
similarly analyzing the Force and Motion Conceptual Evaluation.
RELATE has expertise in educational data
mining, psychometrics including multi-dimensional item response theory, and
experience in designing online experiments and learning environments (including
MasteringPhysics.com).
Our postdocs have collaborated in the
physics department teaching and online education development group. All former RELATE alumni have obtained jobs
in academia (~ 7 at MIT) or education companies.
Additionally, the PI is restarting an
informal month-long “Puzzles and Paradoxes of Physics” and the postdoc may
serve as a co-teacher and co-keeper of the associated puzzles and paradoxes
plus discussions.
Requisite
Knowledge and Skills:
Thoughtful
analysis skills for understanding data
Understand
statistical packages and programming (Python helpful)
Writing
impactful and clear research articles
Knowledge
of education research literature and teaching experience
A Ph.D
or Ed.D in physics involving data analysis, PER, AI, machine learning,
statistics and data mining, or cognitive science is necessary. Applicants
must be comfortable working both alone and with others.
Inquiries
and applications (cover letter, CV, first author paper, and list of at least
three references) should be sent to relateMIT@gmail.com (please include
“postdoc application” in subject line). Salary $73k. Review of
applications will begin immediately; starting date Summer 2026.
MIT is
an Affirmative Action/Equal Opportunity employer, and this job has the standard
MIT benefits.
Prof. David E. Pritchard dpritch@mit.edu
Room 26-241 Dept. of Physics, MIT
77 Massachusetts Ave. , Cambridge, MA 02139.
