English COLLEGEVERSE RESEARCH · 49 ARTICLES
Data & College Choices
Understand the population, year and definition behind a number before using it to compare colleges.
Newest publications first · Page 3 of 5
An Average Growth Rate Can Be Positive While Enrolment Ends Where It Started
Compare arithmetic average changes with a compounded endpoint rate using two invented annual paths.
Compounding experimentNot Awarded Here Does Not Mean Dropped Out Everywhere
Reconcile an invented entry cohort across awards, continuing study, other institutions and unknown destinations.
Outcome-accounting tableA Student-to-Staff Ratio Does Not Tell You Your Seminar Size
Compare invented FTE ratios with teaching-group arrangements and identify the course-level evidence still needed.
Ratio interpretationA Three-Year Average Can Rise While the Latest Result Falls
Reproduce a fictional rolling average and show why overlapping windows are not independent evidence of improvement.
Window decompositionFall Headcount and Annual Headcount Answer Different Questions
Reconcile fictional term snapshots with a distinct annual student count instead of adding registrations.
Stock-flow reconciliationA Test-Score Range Describes Its Submitters, Not Every Entrant
Map an invented admissions score table to the population actually represented before comparing ranges.
Coverage mapOne Decimal Place Should Not Decide Your Entire Shortlist
Test an invented cutoff against rounding and measurement uncertainty before excluding colleges.
Boundary auditWould Your College Ranking Survive Different Weights?
Find the exact preference-switch point in an invented two-criterion comparison.
Weight sensitivityHigher Spending and Higher Completion: What the Plot Cannot Prove
Use a fictional three-college comparison to distinguish association, selection and a causal claim.
Scatter-plot auditWhy Student Categories Cannot Always Be Added Together
Use an invented set-overlap table to reconcile international, part-time and total student counts.
Set reconciliationOne High Salary Can Move the Mean While the Median Stays Still
Recalculate a small invented salary list and show what two measures of a typical outcome reveal and conceal.
Outlier experimentTwo Survey Estimates Differ: Is the Gap Larger Than the Uncertainty?
Use invented confidence intervals to separate descriptive gaps from claims about underlying populations.
Uncertainty comparison