A pet peeve: reporting a mean as a range
“Las Vegas averages 5 to 6 inches of rain a year.”
“Average classroom sizes are 15 to 20 students per teacher.”
Et cetera.
A mean (or arithmetic mean, or average, whichever you want to call it) is a specific number. It is computed by adding the samples in your survey, and dividing it by the number of samples.
For example, if the annual rainfalls in Las Vegas were for the past five years were 5.4 (inches), 6.2, 3.7, 4.9, and 5.9, the mean would be found by adding those together: 5.4+6.2+3.7+4.9+5.9=26.1; and then dividing by the number of samples, which in our situation is 5 (one for each of the five years): 26.1 / 5 = 5.22. Thus, our average yearly rainfall is not “5 to 6 inches” but 5.22 inches.
I think the reason people often erroneously report means is to try to give some idea at the magnitude for which rainfalls vary from year to year: “5 to 6 inches” is a lot less spread out than “3 to 8 inches” — even though the actual mean is likely the same. In essence, the reporter is trying to relate both the mean AND the standard deviation* in one fell swoop.
Unfortunately, it is still technically incorrect to report a mean as a range of values.
* The standard deviation measures the “spread” of the samples — if you’re interested in an explanation of how the standard deviation is computed or its statistical implications, let me know. If enough people are curious, I might write something about it.