Aarhus University Seal / Aarhus Universitets segl

SummerPIT 2019

Join the crowd! Fuel and charge! Dig deep!

Participatory IT research is the study of how people experience, understand, design, and shape information technology as part of their lived experience. Aarhus University’s interdisciplinary research centre for Participatory IT establishes a new foundation for this area of research.

The centre extends the Scandinavian participatory design tradition, which has historically focused on involving people in the introduction of technology to their workplaces. However, during the recent decades, information technology has become an integrated element of almost all parts of people’s everyday lives, including leisure, civic activity, art, and culture, thereby establishing new forms of participation and social practices. The pervasiveness of information technology in human life poses new challenges for the way participation occurs, is supported, and understood.

Accordingly, the centre poses the fundamental question of what participation currently means, and how it may be supported by IT, today and in the future.

PIT will host SummerPIT in August to bring together international researchers from across PIT-related research areas, local researchers, and PhD students.

SummerPIT has, in previous years, been a lively place where we had a lot of fruitful discussions and where people had a chance to meet and discuss. We hope to continue this spirit this year, and encourage you to stay as long as you possibly can of the period.


Understandable AI | Wednesday, August 14

See the full programme of the day

Artificial Intelligence (AI) and Machine Learning (ML) techniques are increasingly being incorporated into digital technology and used to assist in decision-making in various domains such as credit scoring, insurance risk or disease diagnosis, which impact society at large. However, these algorithms may also be biased. To ensure that algorithmic decision-making is interpretable, fair, accountable and transparent, recent research has investigated how these algorithms can explain how they arrive at their decisions. Challenges not only include technical difficulties of reducing complex and non-transparent algorithms (e.g. deep learning) to a simpler representation that is understandable for non-ML-experts, but also psychological and social factors of dealing with explanations such as biases in interpreting uncertainty, and that explanations may provide a false perception of fairness. Finally, with AI and ML increasingly moving into the workplace, it is unclear what challenges people encounter in the active application of interpretable machine learning in their work.

Evaluation | Thursday, August 15

See the full programme of the day

The evaluation of HCI research has provoked many discussions over the past decades, such as arguments about a “damaged merchandise” following the introduction of discount usability (1995) or “usability evaluation considered harmful” (2008). Yet, many open questions remain, including fundamental ones: What constitutes an evaluation of HCI research? Why, when, and how do we evaluate? There can hardly be a one-size-fit-all answer to either of those questions that will be satisfying to all researchers in our field, which has caused communities even to abandon evaluation as a necessary requirement for publication altogether, or accept evaluations that are likely to carry little value (“the users liked our system”, “participants found the design to be intuitive”). What would it take to enrich our toolbox with evaluation strategies that are more malleable to changing contexts of our research? How can we ensure that we not only evaluate the short-term impressions, but long-term effects - and what does “long-term” even mean? Those and other questions remain unanswered despite years of debate, but we invite the PIT community to share their experiences - including stories of “what didn’t work”, which are often not published despite being potentially more useful to derive lessons learned - and discuss novel ways to think about evaluation of HCI research.

Practice in Research: The Role(s) of Artistic Exploration in Digital Design | Friday, August 16

See the full programme of the day

Artistic practice carries a long tradition in which practice can contribute to art and design research. Artists create both art and new knowledge via actively leading research activities fuelled by their particular knowledge of materials, scientific methods and critical reflection on wider social and cultural contexts. According to Sullivan (2006), “art practice can be seen as a form of intellectual and imaginative inquiry, and as a place where research can be carried out that is robust enough to yield reliable insights that are well grounded and culturally relevant.” Human-computer interaction (HCI) research with roots in participatory design and design practice has a similar interest in creating interventionist, activist, critical, and alternative work. The HCI community continues to grow and involve more researchers and practitioners from art and design, and as such, these researchers and practitioners contribute diverse backgrounds and methodologies.

We are interested in the role(s) of artistic practice or artistic exploration in digital design, the relationship and similarities between an artistic approach and design approach to HCI research, critical technological aesthetics in design and artistic exploration, and curation as method(s). How can art practice be a method for HCI research, design and inquiry? How can creative practice and research practice come together? How can we teach artistic practice as inquiry to young researchers? How is HCI research currently incorporating creative practice, and what can art and design methods and practice further contribute?

In this day focused on Practice in Research, we aim to explore and develop common understandings for collaboration and future work by reflecting on the histories, traditions and emerging parallel practices in artistic and design research and HCI, and where they intersect and converge.