Return to Article Details
Towards a more robust teaching evaluation by articulating M-estimators, the Friedman test, additivity tests, and reliability metrics
Download
Download PDF