How to compare published education and employment outcomes

Replace ranking language with a method for comparing definitions, cohorts, dates, denominators, and missing data.

Questions to ask

• Are the populations and denominators comparable?
• What date and definition does each outcome use?
• Which missing data changes the interpretation?

Methodology

Prefer a reproducible official dataset, preserve the source definition, and show missingness. Do not create a ranking when the underlying populations cannot be compared.

Trade-offs

• A comparable dataset may be narrower than marketing tables.
• A transparent range can be more useful than a precise but incompatible rank.

Evidence and sources

• NCES Digest of Education Statistics organizes education-outcome tables around educational attainment and workforce characteristics such as labor-force status, income, occupation, and unemployment, so comparisons must preserve each table's definition and population. Digest of Education Statistics — Chapter 5: Outcomes of Education