Database Reference
In-Depth Information
evaluating each design, or performing each task, etc. The hypotheses were expanded
to (for C columns, for C > 2):
H0: μ1=μ2=⋯=μ C
H1: not all C µ's are equal
Here, in this chapter, we generalize the problem addressed in Chapter 3 also to
the case of having more than two designs, or tasks, or anything else, but with the
condition of having what is analogous to “paired data,” as introduced in Chapter 3.
When there are more than two columns whose means are being compared, clearly
the word, “paired,” doesn't it the description (grammatically!!), but the idea is
the same. Each person evaluates each design, or performs each task. The person
“repeats” his/her “measure” for all columns. This is referred to as a “within-sub-
jects” design/experiment, or, is sometimes referred to as a “repeated measures
design/experiment.” 1 We have the same hypotheses as in Chapter 6:
H0: μ1=μ2=⋯=μ C
H1: not all C µ's are equal
But, just as there was the difference in the exact set of commands that you give
the software to indicate that you have independent samples versus paired data, the
same idea exists here, where, as you shall see, you will use a very similar process
to what we used in Chapter 6, but with a modest difference in the software process
to indicate that we now have a within-subjects design, instead of Chapter 6's inde-
pendent samples (or what is sometimes referred to as a “between-subjects design.”)
7.2 CASE STUDY: COMPARING MULTIPLE EASE-OF-USE
RATINGS AT MADEMOISELLE LA LA
Let's return to our favorite fashion site, Mademoiselle La La. To refresh your memory, you
were hired as the UX researcher at Mademoiselle La La, a high-end online store aimed
at urbane women from 18-55 years of age with well-above-average disposable income.
You'll remember that you launched an online survey to determine which new home
page was more sophisticated. Using a t-test for independent samples, you proved that
Design 1—a scene of a young demoiselle sipping a coffee at an outdoor French café,
ignoring the adoring eyes of a nearby young man—was considered more “sophisticated”
1 We consider the case of independent samples (previous chapter), and the case in which a person sees/
tries all “columns.” There are more complex cases of within-subject designs that are beyond the scope
of this topic. For example, if there are six tasks to be performed, there might be a within-subject design
in which one person performs three of the six tasks and another person sees the other three of the six
tasks. Clearly, we do not have independent data, but the “within-subjects” aspect of the design is more
complex than having one person performing all six designs.
 
Search WWH ::




Custom Search