Developing systems to respond, recognize or record affect in a meaningful manner requires copious amounts of initial data and participants willingness to share their most intimate details. But how as researchers and designers do we determine when we have enough data, who to include and exclude, and what to include.

When developing systems to aid users’ in emotional reflection or learning, user input and feedback on those systems is critical to ensuring the systems are designed in a meaningful manner. In developing AffectAura, user feedback played a critical role in researchers ability to understand how to improve the system[2]. By listening to participants, researchers were able to understand that some design features were to limiting and would require adjustment, and that the interface was not as user friendly as it should be [2]. A system that enables users to reflect on their emotions and/or memories, users need to be able to decided how much data is to be collected and how that data should be represented. Balancing users who wish to have data abstracted verses those who wish to have a more analytical approach, is a balancing game and continues to be a theme within Affect research.

D’Mello et al designed a system that would engage students in a meaningful manner [1]. While the system was designed to engage students when bored, when it came to how someone learns and what pedagogy to implement, the system fell flat. To develop a system that responds to a specific learner, professional educators should have been more involved in order to address learning and teaching styles and how to keep students engaged. While D’Mello et al. gathered useful data on detection, involving professional educators would have been more valuable in order to increase user engagement and how to address what to do when a student seems to become bored, frustrated or confused.

The type of system being developed will ultimately determine what type of stakeholders need to be involved. Developing an emotional/ memory aid researchers and developers are able to rely on their participants to provide feedback and data. While this may be okay for a creation of an app or software, users ability to modify and provide feedback is vital for the life of the app/software. Further, if that app starts to suggest modifications in behavior, mental health professionals should be consulted and involved in the development. There is a difference between developing a system that allows the user to ‘journal’ verses a system that engages with the user. For systems that are designed to provide feedback to users based on their emotional or physical affect, additional stakeholder involvement will need to be considered as well as various types of participants in initial studies.

references

[1] Sidney D’Mello, Rosalind W. Picard, and Arthur Graesser. 2007. Toward an Affect-Sensitive AutoTutor. IEEE Intelligent Systems 22, 4 (July 2007), 53–61. DOI:https://doi.org/10.1109/mis.2007.79

[2] Daniel McDuff, Amy Karlson, Ashish Kapoor, Asta Roseway, and Mary Czerwinski. 2012. AffectAura: An Intelligent System for Emotional Memory. http://www.microsoft.com, 849–858. Retrieved April 1, 2022 from https://www.microsoft.com/en-us/research/publication/affectaura-an-intelligent-system-for-emotional-memory/