Char, you need to learn what a correlation really is, and what it isn't.

If we found a correlation of .85 existed between happiness and personal wealth we could say they are (or appear to be) somehow related--but we couldn't say we have any idea of how they might be related. That is, does one cause the other or are both generated by some undefined third (or nth) factor.

The situation offered earlier regarding doctors and blacksmiths is something of a standard used in statistics classes to illustrate this point.

Correlations may be either positive or negative in their indication of relationship between two statistics or parameters (whether the two sets of numbers move together or move in opposite directions). The nearer the correlation is to 1 or -1 the greater the likelihood a change in one statistic will correspond to a change in the other statistic. But consider the doctors and blacksmiths.

Another oft misused statistical maeasure concerns averages. Mean, median, and mode might all be listed as averages, but the meanings are totaly different.

Median is a number such that one-half of all the numbers in the population in question are above it and one-half below it.
The m

edian of 1, 2, 3, and 1,000 is the same as the median of 1, 2, 3, and 4.

Mean is simply an arithmetic average of a group of numbers comprising a statistic (range).

The mode is the most commonly occuring number in a range. Of 1, 2, 3, 4, 4, and 1,000, four is the modal value.

In any large range of values showing normal ditribution the mean, median, and mode will be very close to identical.

I list this as it also comes up once in a while when attempts are made to lie with statistics.
President Reagan summed it up nicely, "Some people go through their entire lives wondering if they made a difference, Marines don't have that problem!!!"