I was having a discussion with fellow researchers in an academic writing module. We were discussing the difficultly of evaluating our research against some objective criteria. Three of people??(TM)s PhD projects are about improving the ease of performing a certain task (e.g. building an ontology). However, the measure ??oeease?? a series of usability tests are required. HCI however, is something this computer science department does not teach (at all). It is not ??oehard-code engineering?? enough.
Ultimately, these students may end up changing what they do so that they come up with a research hypothesis that is easier to prove. I think this is a major flaw in the way research is conducted. Everything is far too focused on evaluation, evaluation, evaluation. Usability is difficult to objectively evaluate, so most research ends up avoiding usability altogether. The result: completely unusable software that bewilders the average human being.
If only we could relax the so-called objectivity of modern science and introduce some subjectivity. Scientists would be more inclined to the process of improving their subjective state of consciousness and computers might actually become easy to use.