3 ms·
If you had a multiple choice quiz with the actual questions hidden you would get a normal distribution in results. Anything where you sum lots of small random e
by salty_biscuits 7y ago
If you had a multiple choice quiz with the actual questions hidden you would get a normal distribution in results. Anything where you sum lots of small random effects.
- im3w1l 7y agoIf the test was not measuring anything at all, then you would expect repeat takers to have completely unrelated scores on first and last test.
- salty_biscuits 7y agoThat's true, I'm just pointing out that the normality of the distribution of the results alone proves nothing.