Stand. Dev. in Analysis of Changes in Crime Stats
Message posted by on February 1, 2000 at 12:00 AM (ET)

Can standard deviation be used to evaluate the statistical significance of an apparantly "sharp" rise/fall in (for example) reported crime?

I've had it explained to me that standard deviation is the average variation from the mean of a group of stats. If so it would seem that it could be used to evaluate whether the change from year 4 to year 5 is within this "average variablity".

Example:

Anywhere, Fla had the following Homicides:
1994 17
1995 14
1996 5
1997 11
1998 4

Excel tells me that the S.D.of these 5 numbers is 5.6. The Mean is 10.2.

Am I wrong to assume that for this 5 yr. period a change of 5.6 plus or minus should be considered an "average change"?

If so, what appears to be a sharp rise from 96 to 97, should only be considered within the average variabilty for this period.