Can you relate? Correlation and simple linear regression - Improving the User Experience through Practical Data Analytics - page 233

FIGURE 9.26
Augmented SPSS output; illustrative data.
So, for each unit increase in the perceived usefulness of the ability to do a Bool-
ean search, there is a corresponding increase of 0.46 (on the 5-point Likert scale) of
“Likelihood of Adoption” of the search engine. A person who gives a 5 for useful-
ness of the ability to do a Boolean search has a predicted value of about 4.76 on the
5-point scale of likelihood to adopt the search engine.
A conidence interval for this value, using the methodology described in the pre-
vious section, is 3.85 to 5.00. A person who gives a 4 for the usefulness of doing the
Boolean search has a predicted value of 4.30 for likelihood of adoption of the search
engine. A conidence interval for this value is 3.39 to 5.00. We would, of course, pre-
fer narrower intervals, but we must remember that we are using only one X variable
in the prediction process, and while the r 2 is about 20%, which is often very good