Skip to main content
Grantee Research Improving High School Transcript Records via Item Response Theory to Predict Postsecondary College and Labor Market Outcomes
Last updated on
Display Date

Add Urban on Google

This project examines course grades and course‑taking patterns to develop a scalable measure of transcript strength that can identify which course choices most strongly correlate with postsecondary and labor market success, offering a more accurate way to account for course‑selection behavior and selection effects when comparing student trajectories. Building on prior work using Delaware data, the team will expand the transcript‑strength measure to Texas, refine it from an end‑of‑high‑school indicator to grade‑specific measures, and extend it into domain‑specific strength indexes, such as STEM (science, technology, engineering, and mathematics) and humanities, rather than a single unidimensional score. In so doing, the team will make transcripts a more useful tool for measuring skills. By linking domain‑specific measures to college major, occupational field, and outcome data (including college‑level earnings, employment, and early labor market trajectories), the study will generate actionable insights into how transcript‑embedded skills translate into later opportunities and how schools can better guide students toward courses that build the competencies most predictive of long‑term success. This project also complements SUMI’s niche of using existing, familiar data for new skill and mobility driver measurement and analysis.

Data Sources

  • Data: Texas transcript and demographic data, National Student Clearinghouse postsecondary enrollment and completion records, and degree field and institution-level earnings from the College Mobility Report Card
  • Mobility Outcomes: Texas Unemployment Insurance employment and earnings through Texas State Longitudinal Data System (SLDS)
Body

Research Team

Kenneth Shores

University of Delaware

Sanford Student

University of Delaware


Cohort 2

Academic Achievement | Driver Validation