Seminar 2

Chapter 13 is about evaluation, why it should be done, what needs to be evaluated, where it should take place, and when. Simply put, evaluation is very important for the development of a product to improve its design and to inform the designer that the product is user-friendly and enjoyable for the user. As the book mentions, well-designed products sell. Depending on the product or prototype, different types of feed back will be of interest by the developer.

There are three categories of evaluation depending on the setting, if users are involved or not and the level of control. Two of these involve users, with one being in a controlled setting, this one is good for when more control is needed and when outside influences and distractions have to be excluded. The other is in a natural setting and is optimal when the evaluator doesn’t want to affect what people do during evaluation. This is also called field studies. The third one is without users and the setting can be anywhere, often imagined or modeld. These methods are often combined to obtain a clearer understanding, but knowing when to use which method is important. 

Chapter 15 talks about this third category where users are not present. These methods were developed to use when real usability testing wasn’t possible because of users not being available or just to save on costs. The inspection methods that were developed are heuristic evaluation and walkthroughs, were there are cognitive walkthroughs, were a users problem-solving process to accomplish tasks with a system is simulated and analyzed, and pluralistic walkthroughs, mostly involving experts discussing usability issues in different given scenarios. These are in turn conducted with so called experts that are, as the name implies, experts in both interaction design and user behavior, that often role-play typical users and evaluate whether user-interfaces follow previous tried and tested principals. The original heuristic evaluation methods for HCI were developed by Jakob Nielsen, the book summarizes the ten usability heuristics that were derived from an analysis of 249 usability problems. Predictive models are also used without the users being present and simliar to analytics data is collected. It provides estimates of efficiency for various task in different systems. 

What is the best evaluation method for our product?