Seminar 2 - Notes

The chapters required for this seminar were chapter 13 and 15 with the first one introducing us to the general concept of evaluation, its most common forms and the second zooming in on the ones performed without user involvement. The chapter started of discussing the main reasons for why different forms of evaluation are used in modern development, It’s stated that the practice of iterative evaluation from an early stage can significaly diminish the time required for the product to achieve the same level of result as if developed without those measurements in place. Being able to notice obscure flaws or bugs before the official launch and winnings of time and money are also brought up. Some main forms of evaluation are stated. Field studies, more controlled forms of studies involving custom lab environments built to eliminate extern variables and the use of evaluation methods fully missing of user involvement mainly consisting of experts performing some kind of predetermined test like heuristic evaluation, walkthroughs or purely mathematical models. These methods all differ greatly in ease of use, cost and time requirements and also the data they have the possibility of generating. Field studies have as in most cases the largest room for external interference which can skew results but also highlight problems not previously known about the product. The main cons of this method are the cost and the relative difficulty of execution. More controlled forms of evaluation can include participants being located in some form of lab or controlled environment and introduced to a task which they are expected to perform. It’s stated that these are a good way for scientist and researchers to test their hypothesis and receive feedback on specific scenarios. Evaluation without the use of user input can be both cheaper and faster to conduct than other choices and as stated before often involves the use of standardized tests and predetermined models. My question for the seminar is for the necessity of these tests conducted by experts. If the product is designed for the average consumer why should it be tested by an expert who might overlock or run in to completely different problems?