Assessment is a thorny issue, one that causes the most problems and tension between staff and students. We may think the reason for assessment is so we can ultimately give students a degree classification that represents their achievements to the outside world. But assessment serves a much wider function.

The phrase “Assessment drives Learning” is often quoted – students will focus their learning on what is required for their assessment. Typically this means spending a significant amount of study time on coursework assessments (often much more time than is necessary) because they are ‘worth marks’. Any learning activities that are not ‘worth marks’ (including attendance at lectures and tutorials) tend to be seen as low priority. Examination assessment tends to foster a short-term ‘cram-and-forget’ attitude.

I don’t think this is a good situation – we either need to find ways in which we can encourage students to engage in learning activities that are not associated with marks, or we must ensure that the assessments are carefully designed so as to encompass all the intended learning outcomes:

“If assessment drives learning, then we have to be clear about the kind of learning that we want from students before we choose assessment methods” (J.Moon (2002), The Module and Programme Development Handbook, Routledge, pg115)

If the former solution is hard, the latter is probably harder!

Below is the HCI assignment given to first year students in week three of their first semester (when they are still new to university) – it is closely based on the assignment given by the previous lecturer for many years, and is worth 6% of a course which is one-twenty-fourth of their total first year studies. This was the first time I have taught this course.

This assignment embodies some of the principles I think should underlie HCI assessments:

1. Design is a process, not a product. There is no single ‘right’ answer to a design problem. I think it is useful to introduce this concept to these students early on in their HCI studies. Computing Science students (who may be used to getting full marks for more quantitative assessments) sometimes find this frustrating. In particular, this assessment does not give any reward for the quality of the design itself, but only gives marks for the design and evaluation process.
2. Peer assessment encourages critical analysis.The evaluation component of this exercise was performed with other members of the class – this means that students saw a range of other solutions to the same problem. Students thus see design solutions that they would consider better or worse than their own, and they are encouraged to think about criteria for a ‘good solution.’ At the extreme, a ‘studio’ format is best, where students work together in a shared location, and can continually seek feedback and advice from their peers.
3. HCI needs to be seen in the wide context of system development.‘Beginning-to-end’ projects are the most worthwhile, that is, assessments that cover the whole iterative process of design, implementation (even if only low-fidelity implementation) and evaluation. Students can then see how the different phases of development inform each other. This is not always practical – in this case, while the students touched on all aspects of the development cycle, it was not possible to take this idea to the extent I would have liked because of the low weighting of the assignment (and the correspondingly low number of hours the students would be expected to work on it).
4. Group work allows for combining of skills. HCI is an interdisciplinary subject, and the greater the variety of skills in a group, the richer, the more extensive, and the more ‘rounded’ the outcome. At least, this is the ideal! In this case, some preliminary design work was done in groups in a tutorial, but the final submission was individual. Group work is not always practical or appropriate: my Masters HCI course has included a substantial group assessment, but in my department group submissions are not considered appropriate at first or second year level. There is also the infamous problem of deltas (“I did the most work and he did nothing!”) Developing a system in a group is, however, a more realistic scenario than individual development when compared with development activities in industry.

While this assessment can be considered successful (in that it embodies four key principles of assessment, and satifies my aim of students following through the whole design-implementation-evaluation process), as this is the first non-programming assessment that these students undertake in their first year, some students fail to see the point of doing this exercise. At this stage in their Computing Science degree (first semester of first year), many of them still believe that computing science is solely about computers, and find any material that is not clearly related to computers a distraction.

A final word about examinations. They are a necessary evil, and most institutions (including my own) insist on them (or, at least, it is very difficult get permission to remove them entirely from the assessment for a course). Given that this is the case, my HCI examination papers at both first year and second year level assess both theory and practice in equal measure.