Measurement error
The difference between a measured value and the true value. Every measurement has some, and a single noisy number shouldn't decide anything important.
In plain English
Some error is random: the scale reads a little high one day and low the next. Some is systematic: a strict manager rates everyone lower. Random error shrinks when you measure several times and average; systematic error doesn't, and needs a different measurer or a better scale. Ratings, surveys and judgements of people usually carry a lot of both.
Uncle Gary explains
In a study of 4,492 managers each rated by six people, about 62% of the differences in ratings came from who did the rating and about 21% from actual performance. In a team of ten, the top-rated person was truly the best only 27% of the time in our simulation.
Why it matters for your money
Fund ratings, analyst targets and star reviews are measurements too. Several independent ones, or a long record, beat a single confident number.
Where we tested it
- Are Performance Reviews Accurate? What a Rating Really Measures
- Segal's Law: Why Two Reports Never Show the Same Number
Related terms
All glossary terms. Educational content, not financial advice.