Reading Seminar 2 - Notes

When thinking about your product and what you want to develop, it’s very difficult to just ask people what they want without pointing them in the direction of your own ideas, however if they use something, they fairly quickly know what they DON’T want. This is one of the main reason why prototyping is so very important.

   But before we start prototyping we need to have a sound understanding of our users, their needs and what goals our product should fulfil for them. The best way to do this is to observe the users in question and perform an evaluation, to ensure that our product is right for the users.

   First we need to break down and determine the overall goals with our product, which our evaluation should address. Secondly we need to break down the “Why do people want this?” question into several more pinpointed sub-questions. “Is it because of this/that?” This will give us the opportunity to test or confirm ideas, making the evaluation more detailed.

   In order to answer these questions we need to study the users in question.  Choosing evaluation paradigm and techniques is not easy. What may seem most appropriate may take too much time or cost too much. Sometimes the best solution is to combine several techniques.

   Usability testing is one way, but it should be pointed out that this environment is strictly controlled by the evaluator. We also need to take special care when forming the prepared task the users should perform, and how long it should take. The number of users are typically too small for a statistical analysis and the results are mainly quantitative.

   If we want more qualitative data, field studies may be a better option. But the problem here is how to observe without disturbing. Environment may or may not influence data which in turn may not be 100% reliable due to it sometimes being verbal. But compared to usability testing you will gain invaluable insight into how users “naturally” use our product.

   Both of these techniques raise several questions about practical issues. Such as what users are “typical” for our product? What type of equipment should we use? What type of environment should we study etc. Also don’t forget to analyse all these questions with time and budget constraints in mind.

   Of course if you do interact with users, ethical issues may arise as well. Common practice is to keep the users well informed and assert anonymity (no personal info will be shared unless permission is given). It is of vital importance that no person is identifiable through the research notes!

   Last but not least. All your data need to be thoroughly analysed. Is it reliable? Will it produce same (or very similar) result if the test was performed at a diff time but in the same environment? Does it measure what it was supposed to? Does the environment have any influence on it?

   All this is needed BEFORE we even start to think about our conceptual design or prototype AND it is something that probably will be used several times during the iteration cycle for the prototyping.

At early stages of development, it’s important to hold discussions and review meetings. It enables you to get diff perspectives and in case of low-fidelity prototypes, rapid feedback from colleagues. Also, informal feedback from users early on may prove invaluable when developing an interface metaphor or interaction paradigm.

   If you reach high-fidelity prototyping again it’s important to note the DECIDE framework. So that the prototype still is in line with the users’ needs.

Depending on what type of feedback you want on your product, sometimes it might be better to have a high-fidelity prototype with a horizontal prototyping rather than low-fidelity prototype. For ex if you want feedback on a design for an interface.

Question:

What would be more efficient for usability testing of an app, evolutionary- or throwaway-prototyping?