Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

Share this Page URL
Help

7 What Sample Sizes Do We Need?: Part 2:... > Additional Applications of the Model - Pg. 149

Additional Applications of the Model 149 · · · participates in a study, typically relying on employment agencies to draw from their pools to obtain participants who are consistent with the target population. Observations among participants are independent because the events of interest experienced by one participant cannot have an effect on those experienced by another participant. Note that the model does not require independence among the different types of events that can occur in the study. The two mutually exclusive and exhaustive event categories are (1) the event occurred during a session with a participant or (2) the event did not occur during the session. Finally, the sampled observations in a usability study do not deplete the source. Another assumption of the model is that the value of p is constant from trial to trial (Ennis and Bi, 1998). It seems likely that this assumption does not strictly hold in user research due to differences in users' capabilities and experiences (Caulton, 2001; Woolrych and Cockton, 2001; Schmettow, 2008). The extent to which this can affect the use of the binomial formula in modeling problem discovery is an ongoing topic of research (Briand et al., 2000; Kanis, 2011; Lewis, 2001; Schmettow, 2008, 2009)--a topic to which we will return later in this chapter. But note that the procedures provided earlier in this chapter are not affected by this assumption because they take as given (not as estimated) a specific value of p.