Information Technology Reference
In-Depth Information
The Web artifacts selected to be evaluated were all the platform-independent mod-
els and the final UIs (i.e., 1 class model, 4 NADs, 4 APDs and 4 final UIs). PSMs
cannot be evaluated since they are embedded in the model compiler. Figure 2(a)
shows an excerpt of a NAD and Figure 2(b) shows an excerpt of an APD.
A set of 20 usability attributes were selected from the Web Usability Model
through the consensus of the evaluator designers and the Web usability experts. The
attributes were selected by considering which of them would be more relevant to the
type of Web application and the context in which it is going to be used (e.g., Inter-
connectivity, Clickability, Permanence of links, etc.). Only 20 attributes were selected
in order to avoid extending the duration of the execution stage.
With regard to the
specification of the evaluation
stage of WUEP, metrics associ-
ated with the selected attributes were obtained from the Web Usability Model, and
then associated with the artifact in which they could be applied. Since metrics can be
applied at different abstraction levels, the highest level of application was selected.
For example, the metrics presented in Appendix A indicate that: the Compactness
metric could be applied to models that specify the navigation (i.e., NADs); and the
Discernible links metric could be applied to models that specify the UI (i.e., APDs).
Once the metrics had been associated with the artifacts, metrics were operational-
ized in order to provide a calculation formula for artifacts from the OO-H method (in
this case), and to establish rating levels for them. Appendix B shows an example of
operationalization and rating levels for the metrics from Appendix A.
With regard to the
design of the evaluation
stage of WUEP, a template for usability
reports was defined by considering the same fields proposed in the previous section.
The evaluation plan that was elaborated takes into consideration the following evalua-
tion order of the artifacts: class model; NADs, APDs, and final UIs. Five evaluators
were selected to perform the execution stage of WUEP, as mentioned above. This
number was selected by following the recommendations of studies which claim that
five or more evaluators are needed to perform usability inspections [21]. The evalua-
tors attended a training session of 2 hours at which they were informed of the artifacts
from the OO-H method and the tasks to be performed.
Fig. 2.
Excerpts from artifacts (PIMs) of the Web application developed in OO-H
