Cooperative evaluation is a method presented by Peter Wright and Andrew Monk. This method combines elements from thinking aloud and question-asking protocol, since the users are asked to think aloud and also to ask questions from the tutor if they are uncertain or puzzled with something. The designers of the systems act as moderators of the tests, and they need not to be experts in usability or human factors but to follow brief instructions on what sort of questions to ask from the test users.
Wright and Monk categorize cooperative evaluation as a discount method and suitable for formative evaluation. It has three essential characteristics:
1. The user performs pre-defined tasks
2. The user thinks aloud while doing the tasks
3. Designer works as the evaluator and the moderator of the test.
Wright and Monk recommend cooperative evaluation to be used in such an early phase of iterative design that the system requirements are already gathered but the design decisions are still under consideration and need to get feedback from the users. In their experiments, Wright and Monk found out that the designers of the systems detected more usability problems than evaluators who had not been involved in the design, although neither these groups had previous experience in usability evaluation. In these experiments, the designers were quite poor in predicting the usability problems, and so were the non-designers, so user testing is recommended.