Concepts that stop short produce empirical score distributions spiked at the top

Evidence

Depending on my mood, I call this the “Lake Wobegon” or “Harvard grading” problem: in Lake Wobegon everyone is above average and at Harvard almost everyone gets an A. The problem is that the concept does not capture all the interesting variation and hence when applied empirically you can get distributions with spikes at the pole. — Gary Goertz, Social Science Concepts and Measurement, 423

Categorization

• SHARE: A punchy diagnostic metaphor — ‘Harvard grading’ — worth sending to anyone who builds indices or rubrics.
• CANDY: Lake Wobegon meets Harvard grade inflation as a name for a measurement pathology; tweetable on its own.
• SEED: Angle: ‘The Lake Wobegon Problem in Metrics’ — why every country, school, and platform ends up with everyone above average.