Helsinki

In Helsinki this week for NordiCHI where I participated in a Doctoral Consortium run by Kari Kuuttii and Gilbert Cockton two big names in European HCI research going back 30 years or so. Kuuttii’s main contribution has been in an integration of the sociological with the computational, right in my area. Cockton has been a strong advocate for the role of design, and practice based research in HCI, also relevant to me. They had two main bits of advice, both of which were well observed. Kuutti said I had too much in my theoretical toolkit and that a common problem with PhD students is they often attempt over-ambitious theoretical synthesis. My own proposal is a triangulation of multimodality, mediation in social constructionism, and an anthropological reading of interactive artefacts. All of this is framed by design. See conceptual model below.

Kuutti’s main advice was to start with the artefacts I have so far such as the network tiles and browser history comics, use only the theory necessary to justify, position, and explain how and why they came about and ditch the rest. This has the effect of sounding refreshingly simple while disguising the kind of focus and refinement needed to deliver a salient theoretical background to my research.

The advice from Gilbert was more straightforward and stems from the question of how to analyse the large amount of video data I have from various research workshops. Analysing video is notoriously difficult, time-consuming, and complex in qualitative research. There are some good video annotation tools around such as Transcribe!, Anvil and VideoANT but they still involve hours of looking at videos, writing text annotations, and re-watching in order to identify common themes. Gilbert suggested a method used by Malcolm Jones at Northumbria University which involves sketching over video footage and focusing analysis on those sketches. This method is potentially powerful and efficient because it provides a first pass filtration on the data. Sketches can be done repeatedly to highlight different categories of data and they then directly constitute a multimodal transcript.

Another highlight of NordiCHI was the opening keynote by design research legend Don Norman. He spoke about pleasure and enjoyment in HCI and how this is provided by designers, often against the awareness or comprehension of engineers. Two nice examples illustrated the idea; bounce scroll (when getting to the bottom of a mobile screen the list will bounce rather than just stop) and viscous friction (scrolling down and letting go will cause the list to continue to scroll for a short time before stopping), both have zero function beyond delight and pleasure. He made the point that a lot of important functions have disappeared from mobile screens such as navigation, discoverability, and reversability – all these are compensated for by the richness of emotional design qualities. Designers, in Norman’s view, were originally called into HCI ‘to make the invisible mechanisms of the microprocessors visible, understandable, and usable’ the result was (and remains) files, folders, trash cans, and menus. Visual and interactional metaphors that make computational processes tractable. The extension to this argument I would make is; most computer users now understand these metaphors and symbols. A much deeper level of invisiblity is happening at an algorithmic level quite independently of microprocessors. The renewed duty of the desgner is to reveal the effects of algorithmic intelligence. Partly what my research is about.

In response to a question about mental models, a topic on which Norman made an important contribution, he defined the duty of designers in providing good conceptual models. If we have a good conceptual model of an interactive system we can work out what’s going on. The complexity for me comes in the fact conceptual models are subject to transformation over time, and in response to changing social and technical circumstances. There are also few models able to take account of hybrid activities such as texting while driving or typing while talking. The idea of conceptual models is still relevant but how those models are used in design is open to question and creative interrogation.