The Not So Private Personal Informatics

The language used around HCI research into digital health and wellbeing monitors, trackers and coaches reinforces the idea of information ownership by the data subjects themselves. We read about personal informatics, quantified self, life logging, self-tracking, and personal enhancement.

These labels suggest an intimacy and a sense of possession. Are we to believe personal informatics are exemplars of Suchman’s complementary machines and humans, where the outcome is relational, situational and can change over time? People’s capacity to act is reconfigured as they interact, but is their agency being thwarted by lack of awareness?

Neither an individual view, nor a peer-to-peer view, seem to capture the richness of relationships, or the variety of motivations, or the range of use. As individuals interact with their technology and social clusters, and as they experience, curate and share their lived data, we need a broader perspective (Kuutti, 1996) which examines how other parties in the data universe interact. The situation boundaries are wider than the person and their apps.

Personal informatics extends beyond the individuals themselves and their own social clusters. The data have value as assets to the organisation providing the services; it also has value to other organisations who might want to use the data for legitimate or improper purposes, and finally it has societal value as interpreted by local and national government, regulators and other bodies (Watson and Leach, 2010).

Individuals have little knowledge of the ways their data might be used, and how or when this might affect them negatively, as it flows through these different parties. It is not so much an ambiguity in explanation, but a complete hopelessness in understanding and control, despite changing legislation and regulation or the existence of privacy notices. The situation they believe they are in is far removed from reality – the plans exist and the individuals are not in control.

If citizens cannot achieve an understanding and insight into what is happening in personal informatics and the trade-offs being made (Pirolli and Russell, 2011), what hope is there that they can take intelligent action? Personal informatics is a long way from individual sensemaking, and currently more akin to data collection sensors for exploitation by other parties – individuals as instrumentation.

Fortunately, the individual-centric research view is changing, with research into the social motivations of these technologies such as understanding the social contexts and practices. Our models, methods, techniques need to perceive the wider picture to understand what is happening and what the side-effects are, at both individual and societal levels. In the meantime, change the language from Personal Informatics – it is Exposure Informatics.

References

Chris Elsden, David S. Kirk, Abigail C. Durrant. 2016. A Quantified Past: Toward Design for Remembering With Personal Informatics. Human-Computer Interaction.
https://doi.org/10.1080/07370024.2015.1093422

Kari Kuutti. 1996. Activity Theory as a Potential Framework for Human-Computer Interaction Research. Context and Consciousness: Activity Theory and Human Computer Interaction, MIT, Massachusetts, USA.

Peter Pirolli, Daniel M Russell. 2011. Introduction to this Special Issue on Sensemaking. Human–Computer Interaction 26, 1–2: 1–8.
https://doi.org/10.1080/07370024.2011.556557

Colin Watson, John Leach. 2010. The Privacy Dividend : the business case for investing in proactive privacy protection. UK Information Commissioner’s Office.
https://ico.org.uk/media/1042345/privacy-dividend.pdf

Photo

Author’s own. Cyclists participating in Sky Ride London 2010.

Learning Styles as Planning, and Learning Styles as Situated Action

Continuing my thoughts about how learning styles could affect people’s encounters with machines, I wanted to examine Suchman’s Planning and Situated Action (1987 and 2007) in an educational context. Curriculum as experienced by humans might have similarities that can help inform about interaction as experienced between individual and groups of humans and machines.

Firstly, it appears learning style models have weaknesses. A review of 13 of the most influential learning style models (Coffield et al, 2004) highlights a lack of theoretical rigour, conceptual confusion and poor quality in learning style models, and an over-reliance on categorisation schemes. Attempts to categorise and then design pedagogy around these feels much more like planning, than planning with situated action. It underplays the idea that “lessons are always co-constructed by teacher and students together, through the unfolding actions and interactions” (Lemke, 1985). Coffield et al (2004) do not rule out the existence of learning styles; their primary concerns are with the research field, and use of learning styles to dictate interventions.

Wells (2003) provides an early discussion of “situated enactment of learning and teaching” highlighting the non-deterministic nature of plans. Whilst undertaking recent team-based activities in our MRes Digital Civics modules, each person does not use a single consistent learning style. Instead it is more fluid – an improvisation based on the materials, objectives and most importantly the other participants – just like Suchman’s analysis (1987) of photocopier users when they try to help each other. These interactions vary session-to-session, and group-to-group. Fortunately, our combined group agency is not entirely pre-scripted and predictable. It is situated action influenced by our experiences, conventions, traditions, knowledge, collaborations, and of course the instructor and lesson plan.

There are equivalences in the dynamics of (human-human) teacher-learner interaction with machine-human interaction – from an initial imbalance of understanding/knowledge, agency emerges during interpretation. The photocopier (Suchman, 1987) was trying to teach “users” its plan, and various interfaces, guides and handbooks were simply alternative methods of broadcasting a fixed plan to address different imagined learning styles.

Furthermore, teachers and learners are not the same, and Suchman (1987) proposes that machines and humans are complementary rather than equivalent. The result (knowledge) again is not a fixed pre-determined outcome, but like Suchman’s flexible and moving “boundaries”, instead is something which is relational, situational and changes over time.

The outcome of an interaction cannot simply be pre-planned, but needs to consider the context, and the interactions between all the participating parties at the time. Human learning styles should inform research in Human Computer Interaction (HCI).

By considering the ways people approach and make sense of unfamiliar problems with other humans and technology, we reveal alternative approaches to how interpretation is encountered, working our way towards better solutions. In turn we can embrace some degree of ambiguity so digital technology is permitted to understand and facilitate people’s actions and circumstances, rather than pre-define these encounters. Learning styles are considerations in both planning and situated action.

References

Jay L. Lemke. 1985. Using Language in the Classroom (Specialised curriculum: language & learning). Deakin University Press, Australia. ISBN 0730003086.

Frank C. Coffield, David V. M. Moseley, Elaine Hall, Kathryn Ecclestone. 2004. Learning Styles and Pedagogy in Post‐16 Learning: Findings of a Systematic and Critical Review. Learning and Skills Research Centre, London.
http://www.leerbeleving.nl/wp-content/uploads/2011/09/learning-styles.pdf

Lucy Suchman. 1987. Plans and Situated Actions: The Problem of Human–Machine Communication. Cambridge University Press. ISBN 0521337399.

Lucy Suchman. 2007. Human-Machine Reconfigurations – Plans and Situated Actions. 2nd Edition. Cambridge University Press. ISBN: 052167588X.

Gordon Wells. 2003. Lesson Plans and Situated Learning-and-Teaching. Journal of the Learning Sciences, 12:2, 265-272.
https://doi.org/10.1207/S15327809JLS1202_6

Photograph

Author’s own. School reports.

The Interpretation of HCI

HCI as an encounter

Human Computer Interaction (HCI) turns out to be a Pandora’s Box of cross-disciplinary (Reeves, 2015) viewpoints, ideas, approaches, methods and theories, with a very wide range of potential research topics. The features of ambiguity (Gaver et al, 2003) and flexibility in how people interpret their computer technology encounters (Sengers and Gaver, 2006) have caught my imagination. The suggestion that there should not necessarily be “a single correct way to interpret a computer system” (Gaver et al, 2003) is fascinating to me since it is in contrast with my initial assumptions that there could always be a single or best way.

Interpretation in the display of information

In making available weather data summaries for a local community newsletter a few years ago, I was restricted to textual presentation at that time. I was surprised by how different people reacted to the presentation of numbers versus written prose, even when the contained facts were identical. I had not realised the effect of different learning styles, thinking there was instead a best way. Miller (2001) suggests individual “learning preferences and styles” have a “significant impact on how students learn”.

Interpretation in the use of computer technology

It appears to make sense that individual preferences and styles also have a role in encounters between people and computer technology. How computer technology that includes this idea of flexibility of interpretation is encountered (found, chosen, seen, used and changed by people), might not suit everyone, and this will depend upon the context, their cultural expectations, their experiences, their knowledge and their preferences. Designing for multiple interpretations will require greater effort, and thinking about learning styles might be a way to consider these opportunities in some king of existing framework. The consideration of multiple interpretations will also be of help in designing systems that want to avoid ambiguity, and that could be another area to investigate.

The issue of interpretation has become a more noticeable issue as the machines have moved from the computer room, to the workplace desk, to our homes, to our mobile devices, to our apparel, and onto objects with behaviours (Levillain and Zibetti, 2017) and in the future components of our bodies and minds.

There is a large volume of prior work around learning styles in the educational field such as the use of online and other e-learning systems (and of course about what learning styles computer science students exhibit), but much less about how learning styles of people in the wider world affect their relationship with machines. Some notable exceptions are using cognitive styles as a way of modelling user preferences (Brown et al, 2006) and considering learning style when evaluating web pages (Papaeconomou et al, 2008).

It seems the learning styles of individuals ought to influence their interaction with computer technology, and this is an area I would like to consider further.

References

Brown, Elizabeth; Brailsford, Tim; Fisher, Tony; Moore, Adam; Ashman, Helen. (2006) Reappraising cognitive styles in adaptive web applications. WWW ’06 Proceedings of the 15th international conference on World Wide Web. Pages 327-335.
https://doi.org/10.1145/1135777.1135827

Gaver, William W.; Beaver, Jacob; Benford, Steve. (2003) Ambiguity as a resource for design. CHI ’03 Proceedings of the SIGCHI Conference on Human Factors in Computing Systems
ACM, New York, pages 233-240.
https://doi.org/10.1145/642611.642653

Levillain, Florent; Zibetti, Elisabetta. (2017) Behavioral objects: the rise of the evocative machines. Journal of Human-Robot Interaction archive. Volume 6 Issue 1, May 2017. Pages 4-24.
https://doi.org/10.5898/JHRI.6.1.Levillain

Miller, Pamela. (2001) Learning Styles: The Multimedia of the Mind. Research Report.
https://eric.ed.gov/?id=ED451140

Papaeconomou, Chariste; Zijlema, Annemarie F.; Ingwersen, Peter. (2008) Searchers’ relevance judgments and criteria in evaluating web pages in a learning style perspective. IIiX ’08 Proceedings of the second international symposium on Information interaction in context. ACM, New York, pages 123-132.
https://doi.org/10.1145/1414694.1414722

Reeves, Stuart. Human-computer interaction as science. (2015) In Proceedings of The Fifth Decennial Aarhus Conference on Critical Alternatives, AA ’15, pages 73-84. Aarhus University Press, August 2015
https://doi.org/10.7146/aahcc.v1i1.21296

Sengers, Phoebe; Gaver, Bill. (2006) Staying open to interpretation: engaging multiple meanings in design and evaluation. DIS ’06 Proceedings of the 6th conference on Designing Interactive systems. ACM, New York, pages 99-108.
https://doi.org/10.1145/1142405.1142422

Photograph

Author’s own. Exhibit in Farmiloe Building. (2014) Clerkenwell Design Festival. London. Designer unknown.

Edited 23 and 24 Oct 2018 to add photo credit.

Edited 30 Oct to include publication year in text references.