Comparing two designs (or anything else!) using paired sample T-tests - Improving the User Experience through Practical Data Analytics - page 79

We click on the command and enter the location of each column of data (see
arrows in Figure 3.2 ).
We continue to choose to ask for the analysis to appear on a new page (here,
arbitrarily named “jared.” See next to the last row in Figure 3.2 : New Worksheet Ply).
After clicking “OK” in Figure 3.2 , we obtain our output in Figure 3.3 .
In examining the Figure 3.3 output, we see that the two-sided p -value is 0.0262,
certainly less than the traditional α = 0.05 (see arrow in Figure 3.3 ).
Thus, we reject H0, and, beyond a reasonable doubt, we conclude that there is a
difference in mean time to post a job between the two designs; design 1 (“The Long
Scroller”) requires less time than design 2 (“The Wizard”).
In other words, if the means for the two designs were the same after we magically
tested with everyone on Earth, a difference in means that large (14.26 − 12.88 = 1.38)
FIGURE 3.2
Entering the location of the two columns of data; Excel.