Assessing health informatics and computing experiments

The controlled experiment is highly regarded because its properties permit conclusions with the most scientific rigor. Controlled experimentation is important for the foundation of disciplines that claim to be scientific. It is also important to conduct them properly: they come at a high cost in time, effort and participation; there is an esteem that confers credibility; there is an ethical responsibility to human subjects. However, the quality of controlled experiments performed in health informatics and computer science is often poor; evidence shows that informatics and computing researchers and evaluators need further training in experimental methods. One way to address quality issues is to measure quality. This follows the example of the creation of questionnaire instruments to measure quality of controlled medical trials, which have also had their problems. This book describes the development and use of a questionnaire (the MICE index) to quantify the quality of controlled experiments in informatics that involve human participants. Such a tool will be helpful to those planning or evaluating informatics experiments.