Testing Multiple User Croups - I

Now let's look at a slightly more complex, yet realistic scenario. Suppose your user profile consists of two different user groups, managers and clerks, who will be using your product- One of your test objectives is to see if there are differences in ability to use the product between or among user groups. In addition, you also want to see if there are differences in novice and experienced users within each group. You will therefore need to vary experience and job type, each of which will have two levels. Once again, you will use a matrix
design, as shown below.

Each one of the four conditions or cells shown in the table above will be populated with a different set of participants. If you want to acquire at least four participants per cell, as shown, you will need a total of 16 participants.
If this is too many participants for your budget and time, (four participants is about the bare minimum per group required to evaluate group differences), then you cannot simply apply a within-subjects design. Instead, you will either have to limit each cell to fewer participants or simplify the study. Remember, limiting a cell to less than four participants severely limits the conclusions you can draw about each group. You will probably need to simplify the research to exclude a study of group differences (see Figure 5-4).