Seminar 2 - Individual Post - Per

The evaluation process is one of the most important parts of designing a system according to me. Even if you think that you have conducted the data gathering, data analysis and  establishment of the requirements according to what the user needs, you can’t know if you are doing things right until you have gone through the the first evaluation. One evaluation framework is the DECIDE Framework. From my own experience, concepts or lists that aims to make their abbreviations become an actual related word are usually shitty.

Determining the goals is the first stage of DECIDE. For our project, a goal might be to find out whether less tourists would give up on trying to buy the ticket electronically with this system. Restating the general goal statement is of course also relevant for us, “Does the design sketch address the description of requirements”?  For us,  exploring the questions would be related to the bigger changes we have made and whether it was the right way to go. Choosing the evaluation methods is a very important part of the evaluation process, choosing the wrong one might generate corrupt data. For us, the ost obvious way to go about it would maybe to bring a tablet with the prototype to the subway and ask tourists if they want to participate. Several practical issues would probably come up with this way of conducting the evaluation, but since time and resources are scarce, it might be our only valid option. I cannot think of any ethical issues that would come up when doing the evaluation this way, buying tickets is also an activity that doesn’t infringe very much on one’s personal life.

Chapter 15 talks a lot about Heuristics. For us engineer student this is more comfortable ground since our first years is a lot about mathematical modelling that has less connection with reality. The Heuristic Evaluation breaks down UI design into a couple of principles.  Walkthroughs is at least a little bit more grounded in reality since you put yourself in the clothes of a user and try to analyze upcoming problematic usability features.  Whereas cognitive walkthrough aims at breaking down however the correct action is the most obvious one, pluraistic walkthrough will let each evaluator assume the roles of typical users and then give them different scenarios to walkthrough. I think these more scientifical ways of evaluating UI:s is definitely useful, but will never be superior to having an actual user participate in evaluation.