I have been in a bit of a nostalgic mood tonight, and so I went back to look at the journal I kept in college. In it, I found an entry where I feel statistics was used improperly. Not by me, but by my Algebra professor.
here’s the data on the numbertheory exam i took on tuesday:
number of students taking exam: 10
mean (average): 23.8 (out of 40)
standard deviation: 6.596211034
median: 26.25
kurtosis: 1.333906232
geometric mean: 22.82177487
harmonic mean: 21.82918952
maximum score: 40 (by the lecturer), 32.5 (by one student)
first decile (i.e., 10% of scores are below): 15
second decile: 15.5
third decile: 16.5
fourth decile: 17.5
fifth decile: 24.5
sixth decile: 28
seventh decile: 28
eighth decile: 29.5
ninth decile: 31
i dont know what i got. i guess ill see tomorrow at 10ish.
Um. 10 students. Professor decides to publish deciles? Seriously? That’s an improper, terrible use of statistics if I ever saw one. Wow.
You probably clicked here because you wanted to know what I scored on it. NOSY. Apparently, from what my journal says the next day, I got 29.5. I also was apparently not pleased with that score. There was a four letter word next to that score in my journal.
Continuous Everywhere but Differentiable Nowhere 
Oddly enough, I know who this professor was. I seem to recall that he was just more amused by the fact that the program he used could automatically generate this stuff than anything else.
I think the most worrying is the Kurtosis! In turbulent flows you need at least 10^4-10^5 to get statistically stationarity.
I’m not sure what a harmonic and geometric mean means with marks as well ;-)
I think a Histogram and a percentage of passing would give a lot more data than all those numbers ;-)