Logging and Visualizing Affect to Facilitate Communication in a Therapeutic Context
Abstract
In recent years there has been an increasing focus on monitoring health and wellbeing with the intent to prevent and manage different illnesses. Advancements in mobile and wearable technology are making data logging easier and more reliable than ever before. An area that has benefited from this development is the area of mental health informatics. One emerging trend aims to provide users with enough data to self-manage their illness, without the involvement of a health care professional. A wide array of different systems meant to prevent or manage mental health issues is now available to the general public. Advanced systems are being researched and developed with the intent to automatically register emotions and moods. The area of affective interaction disagrees with this way of trying to formalize emotional states. They view emotions, or affect, as something more complex than physiological responses. This thesis discusses the challenges of logging and visualizing data about affect. Findings from a literature review show that most applications target specific user groups or conditions. Furthermore, the way affect is viewed influences the way systems are designed with regards to data collecting and data registration. The area of affective interaction views affect as a product of emotions, moods, behaviors and cognition, which in turn is influenced by social and cultural interactions. Guidelines for designing systems within the affective interaction domain argue that the design should be support flexible interpretation and expression of affect, and support a range of communications The Clasp platform, under development by Bryggen Research, allow for data registration of any affective state. The Clutch platform is defined as an affective interaction system that allows registration of any affect as defined by the patient and the therapist. The platform expands the method of active data registration in two ways: firstly, it offers a lower threshold for data registration than many other reviewed systems, and secondly, it captures varying degrees of intensity over a duration determined by the patient. To design visualizations that allow for all types of affect a wide range of prototypes were developed. The prototypes were a result of different types of evaluations with stakeholders and other users, proto-personas, guidelines from Bryggen Research and Boehner et al. (2005), and a design-oriented approach. This finally resulted in high-fidelity prototypes that underwent usability testing. The tests explored user responses on the graphical presentations. The overall feedback and score of the system was good, suggesting that this tool is worth exploring further, however feedback indicate that the system would benefit from a redesign. Examining other types of visualizations or modifications on the ones presented in this thesis should also be a priority. Lastly, the visualizations are only as good as the data, and testing the entire platform with real users and real data is a required. This thesis contributes to the scientific discourse on affective interaction by providing two things: a literature review on how visualizations are being used to represent affect, and a design-proposal for data visualizations generated through design- oriented research.