Project Activities
The Partnership will work together to identify predictors of success observable during student-teaching with information collected by cooperating teachers and human resources evaluators through survey and in-class observation data that are used for early job offers. The team will also examine teacher applicant data from MPS's centralized structured-interview process, the psychometric properties of these measures, and their relation to future teacher outcomes (i.e., evaluation scores, absences, and retention). Partners will analyze how performance monitoring and incentives affect teacher effort and sorting into schools of varying risks (e.g., percentage of students from low-income backgrounds). The project will result in database software changes at MPS to facilitate sustained data-driven personnel decision-making.
Structured Abstract
Setting
This project will take place in the Minneapolis Public Schools (MPS), which is the third largest school district in the state, and the metropolitan area of Minneapolis and St. Paul. Over 2,800 teachers are responsible for 34,400 students (68% racial or ethnic minority, 21% English language learners, and 65% of students eligible for free or reduced price lunch).
Sample
The district's personnel database has records on over 6,000 teachers who have worked in the district at any time since 2007, while the district's current applicant database has over 10,000 applicants (since 2007) and prior applicant database holds records for approximately 4000 applicants.
Data analytic strategy
The team will model the expected value of a potential hire using dimension-reduction and data-mining techniques such as Latent Dirichlet Allocation (LDA) and related Bayesian approaches. They will use generalized difference-in-difference models to examine whether: 1) low-absence, high-effort, more-effective teachers sort towards schools with performance pay or not; 2) teachers increase effort when monitoring and feedback are in place, and 3) teachers increase effort when performance-pay is in place; 4) teachers with desirable observables sort towards more affluent schools when monitoring is not in place; and 5) the introduction of monitoring drives highly effective teachers towards more affluent schools.
Key outcomes
The Partnership will produce more efficient measures and procedures for screening applicants for MPS. The results will inform the design of incentives for recruiting/retaining teachers who are prepared to teach at high-needs schools.
People and institutions involved
IES program contact(s)
Project contributors
Supplemental information
Co-Principal Investigators: Maggie Sullivan (Minneapolis Public Schools), Christopher Moore (Minneapolis Public Schools), Nathan Kuncel (University of Minnesota), Aaron Sojourner (University of Minnesota), Kristine West (St. Catherine's University)
Partner Institutions: Minneapolis Public Schools (MPS), the University of Minnesota (UMN), and St. Catherine's University
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
