Can you relate? Correlation and simple linear regression - Improving the User Experience through Practical Data Analytics - page 228

If we determine the actual value for the conidence interval, using the relatively
complex formula, we would get 1.339 to 5, a bit wider, but not that different, even
though, after all, n is only 5. In a real application, in which n is not so small (such as
in the Behemoth.com data set, in which n = 180), the difference from the theoreti-
cally true conidence interval will be very much smaller, and virtually for sure, the
difference will be immaterial. Indeed, the difference is not that big even with our
sample size of only 5 !
This conidence interval for what will actually occur in an individual case when
X = 3 is wider that you might like, but that is because, as we have noted, it is based on
only ive data values. If we had pretty much the same results for n = 30, the interval
would be much more precise, around 2.45 to 4.75.
9.4.2 SPSS
Figure 9.20 shows the same sample data in SPSS. We have already gone into “Vari-
able View” to label the columns Y and X.
FIGURE 9.20
SPSS template for illustrative data regression analysis.