Data Quality Dashboard

Per-university scorecard for the 106 universities tracked on this site. Generated from scripts/qa_report.py.

Records by University, colored by Degree Labeling Quality

Each bar is one university. Width = total records collected from the university's repository (PhD theses + master's theses). Color shows what fraction of those records have a degree level (PhD or MS) parsed out vs left blank — green ≥ 80% labeled, yellow 50–80%, orange 25–50%, red < 25%. Universities with very few records (< 100) are flagged with a red border. Hover for details.
≥ 80% labeled 50–80% 25–50% < 25% 0% (no labels at all)

Quality Heatmap — All Dimensions

Each cell is colored by quality. Green = good, yellow = ok, orange = concerning, red = problem, gray = N/A. Click a column header to sort.

Top Issues

Frequency of distinct quality issues across all 106 universities.
For each university, every department in our records is checked against four link types: Department homepage, Faculty directory, Students directory, and Repository search. Each bar shows how many departments fall into each coverage tier — green = all 4 fields populated, light green = 3, yellow = 2, orange = 1, red = no link entries at all. Sorted by total department count.
All 4 link types 3 of 4 2 of 4 1 of 4 0
Sitewide totals for each of the 4 link types across all dept-occurrences (106 universities x ~54 average departments). Repository search URLs are auto-generated from the OAI-PMH base URL so they're near-complete; the other three are research-driven and reflect curation effort.
How many departments at each university have each link type. Cells colored by % coverage relative to that university's total department count.