Can you relate? Correlation and simple linear regression - Improving the User Experience through Practical Data Analytics - page 234

FIGURE 9.27
Linear regression output for Behemoth.com data; Excel.
(i.e., high) for just one X variable, there is still 80% of the variability (1−r 2 = 0.8)
unaccounted for.
The p -value of the slope is 0.000; thus, the slope is highly signiicant; remem-
ber that the value “2.47E-10” actually means 0.000000000247 as the notation
indicates that we should move the decimal point 10 places to the left. It is, of
course, the same p -value we found much earlier when we analyzed the correla-
tion. This very low p -value indicates that there is virtually no doubt that there is a
positive linear relationship between the usefulness of the Ability to do a Boolean
search, and the Likelihood of Adoption of the search engine. Furthermore, the
r-square value of 0.202 means we estimate that the former, by itself , explains
more than 20% of the responder's choice for the Likelihood of Adoption of the
search engine query.
The results are the same, of course, in SPSS. See Figure 9.28 .