The Usability Problem Classification (UPC) tool is for analyzing, diagnosing, and classifying usability problems by problem type and by causes in the interaction design, to locate the problem within the usability design space, as contained in the UAF.
During usability evaluation (e.g., lab-based usability testing, usability inspection, or remote usability evaluation), critical incidents (events occurring during a task that reveal a negative effect on the user's performance or satisfaction) are noted as indicators of possible usability problems.
The context of a critical incident is a usability situation, which can involve more than one usability problem. For example, a critical incident centered on an error message can indicate something about the poor design of the message content. Or it can indicate something about how difficult the message is to read. Additionally, the fact that an error message was displayed is an indication that the user encountered an error, a problem with determining how to carry out a certain intention correctly (e.g., difficulty in knowing which action to perform on which object).
Before analysis and diagnosis, each usability problem must be extracted and isolated from the other possible usability problems in a usability situation. Otherwise, each practitioner or developer, unable to agree on a diagnosis, may see a different problem in the usability situation but think they are talking about the same thing. The UPC tool encourages and supports problem extraction.
See our page on motivation and justification.
Practitioners use the UPC tool to analyze and classify each usability problem extracted from a usability situation indicated or revealed by a critical incident occurring in usability evaluation. Analysis and diagnosis of a usability problem has two goals:
Thus, to determine how a problem interferes with use, analysis begins with identifying the effects of the problem on the user, including these kinds of effects:
Next, to determine the why a problem interferes with use (the causse), classification identifies the type and subtypes of a problem and its causes within the interaction design, which means that classification is diagnosis. Through a sequence of classification decisions at UAF nodes, practitioners identify a sequence of usability attributes (taken from the UAF) that define and characterize a problem. Each decision adds one descriptive usability attribute and takes the process one level deeper into the detail, while pruning off all but one sub-trees at each decision node. The resulting set of descriptive attributes is an "encoding" of the usability problem in the "standard" usability vocabulary of the UAF, unique precisely locating the usability problem in the usability design concept space of the UAF.
{give example here, from NSF or Basic}
As we said above, usability situations can involve more than one problem and thus the need for problem extraction. Similarly, each usability problem can have more than one cause. [give example of font size and color contrast} each different cause usually requires a different solution. e.g., fixing either font size or color contrast alone doesn't fix the problem. As a matter of cost importance, we do try to find single solutions that fix multiple problems and multiple causes. We have to be aware of all the problems and causes in order to do this effectively.
The UPC tool is used closely with the Usability Data Base (UDB) tool. As a new problem is to be analyzed, a database record is created. As each part of the analysis is completed in the UPC tool, the results are put into the database record for the problem.
describe and explain. a set of pre-analyzed and classified ups taken from real-world usability situations. link to the book (coming soon)
useful for teaching usability, too
Future: plan to attach each of these to the corresponding node in the UAF, so that a practitioner, arriving at a node duing classification and unsure of the choice, can see a representative sample usability problem of the type represented by that UAF node, along with solutions that have worked for other practitioners.