Microsoft Office Tutorials and References
In Depth Information
Deviating from the Middle
A negative result means there is an inverse correlation. As one set of data
goes up, the other goes down. The actual negative value tells to what degree
the inverse correlation is. A value of –1 means the two sets of data move
perfectly in opposite directions. A value of –0.5, for example, means the two sets
move in somewhat opposite directions.
The CORREL function returns a value between –1 and 1. A positive value
means the two data sets move in the same direction. A negative value means
the two sets of data move in opposite directions. A value of 0 means there is
no relation between the sets of data.
Figure 9-13 shows three correlation results. The correlations display how
customers reacted (as a percentage increase in sales) with regard to three
types of advertising. All three advertising campaigns show a positive
correlation. As more money is spent on advertising, customer responsiveness
increases (or at least doesn’t reverse its direction).
All three returned correlation values fall within the range of 0 to 1 and,
therefore, are easy to compare with each other. The evidence is clear — direct
mail is not as efficient as magazine or radio advertising. Both the magazine
and radio advertising score high; the returned values are close to 1. However,
direct mail returns a correlation of 0.4472. A positive correlation does exist —
that is, direct-mail expenditures create an increase in customer
responsiveness. But the correlation is not as strong as magazine or radio advertising.
The money spent on direct mail would be better spent elsewhere.