Introduction

It is well established that affect influences human cognition and behavior in such aspects as creativity and problem solving [12, 16], motivation [5], attention [6], memory [14], and social behavior [1]. Affect in the context of interactions between computers and users has also been of deep interest to HCI researchers and practitioners [17, 18].

One of the key challenges of evaluating affective interfaces and interactions lies with measuring affect. Different terms have been used to describe the measurement of affect such as sensing, detection and recognition. Here we use ‘measurement’ as an umbrella term to signify all these. Traditionally, two categories of affect measurement techniques have been used: physiological measures and self-report. Physiological measures involve looking into such signals as facial expressions, vocal tone, skin conductance, heart rate, blood pressure, respiration, pupillary dilation, electroencephalography (EEG) or muscle action, to determine the intensity and quality of an individual’s internal affective states. Concerns with physiological measures involve the difficulties relating to (1) setup, invasiveness, and analysis and (2) the association of specific physical responses with a particular type of emotion because of individual variability [4].

Self-report measures involve a plethora of affect inventories: verbal descriptions of an emotion or emotional state, rating scales, standardized checklists, questionnaires, semantic and graphical differentials and projective methods. Criticisms of self-report methods include the possibility that they draw attention to what the experimenter is trying to measure, that they fail to measure mild (low intensity) emotions, and that they are not construct valid [13].

New challenges with affect measurement have emerged with the evolution of computing from single-user-single -computer to multiple users interacting through various technologies. This shift is reflected in the emergence of the field of Computer Supported Cooperative Work (CSCW) and Ubiquitous Computing. Conventional models of emotion have assumed that affect is an internal, strictly individual phenomenon [3]. Yet, when dyads or groups of individuals interact through technology, a number of interesting questions arise. For instance, does a ‘group affect’ [7] emerge from the affective experiences of the group’s members? When members of a group experience different affect, how do these divergent experiences combine into a group affect? Is group affect more than the sum of individual group member’s affect? How does the group perceive and track this emergent quality? Is colocation necessary for the emergence of group affect? What cues are effective in face-to-face engagements, and what cues are effective in remote or computer mediated engagement? How do cultural differences in affect influence group experiences, and the emergence of group affect?

The conversation around affect measurement has already started within the CHI community. The CHI ‘05 workshop on evaluating affective interfaces addressed evaluation strategies for affective interfaces [10]. One of the identified critical issues of the highly successful SIG on emotion research in HCI at CHI ‘07 was affect measurement [2]. This workshop thus aims to continue this conversation through addressing the issues and challenges of affect measurement when moving from the individual to the dyadic and group levels of analysis. As computing moves to an increasingly collaborative and ubiquitous model, it is timely to address affect measurement beyond the individual. Balancing the art of designing affective interfaces and the science of measurement also fits in well with the theme of this year’s conference.

Goals

HCI researchers have recently started to develop various techniques for measuring affect [e.g 11, 15]. However, a systematic discussion of their effectiveness and applicability in different contexts remains missing. Furthermore, techniques for measuring the affect of dyads or groups interacting through technology have received little research attention. The goals of this workshop are to act as a forum where designers, practitioners and researchers can introduce novel methods of affect measurement that go beyond physiological and self-report measures, to examine ways that existing measurement methods can be expanded, and to critically evaluate issues around affect measurement in shared environments. While the specific issues the workshop will address will be determined by paper submissions rather than a priori by workshop organizers, examples of some issues include:

Yours, mine or ours: Is 'group affect' merely a summary of individual group member affect or do we need measurement methods beyond the individual? How is affect transferred from one group member to another through emotional contagion, behavioral entrainment and interaction synchrony [9]? How do we measure such transfer processes?

Implicit measures: A method that overcomes many of the criticisms of self-report and physiological measures of affect measurement are implicit measures such as analysis of linguistic cues [8]. What other implicit measures can the HCI community utilize?

Objective and subjective measures: In what contexts are objective measures of emotion [e.g. 15] and more subjective measures of emotion [e.g. 11] useful? Can there be a common ground between the two? A structured conversation between researchers using these different measures holds great promise for the community.

Mild emotions, mixed emotions and single emotions: Majority of measurement techniques focus on single emotions or an umbrella of emotions generally referred to as positive or negative affect. Is there measurement techniques that can measure subtle (low intensity) emotions or different emotions experienced simultaneously?

Cross cultural applicability: What are the cross-cultural issues associated with affect measurement? Are there measurement techniques that can be reasonably applied across cultures?

Workshop Plan

Our one day workshop will be structured as follows:

9:00-9:15 Welcome and introduction of organizers
9:15-10:30 Poster 'madness' where participants introduce themselves and briefly advertise their work
10:30-11:00 Coffee break
11:00-11:10 Participants fill out post-it notes on specific issues related to affect measurement they are interested in
11:10-12:30 Poster session with mingling to encourage small group discussion AND collaborative card sorting exercise of post-it notes
12:30-14:00 Lunch (organizers map card-sorting themes to break-out groups)
14:00-16:00 Break-out groups on selected topics
16:00-16:30 Coffee break
16:30-17:30 Report from break-out groups
17:30-18:00 Wrap up
18:00-18:30 Break
18:30-21:00 Dinner (individuals responsible for costs)

Workshop proceedings will be published with an ISBN. We anticipate publishing a special journal issue or edited book with revised and longer versions of workshop submissions.

References

1. Berkowitz, L. Aggression: Its causes, consequences, and control. McGraw-Hill, New York, USA, 1993.
2. Crane, E., Shami, N.S., and Peter, C. Let's Get Emotional: Emotion Research in Human Computer Interaction. Ext. Abstracts CHI 2007, ACM Press (2007), 2101-2104.
3. DePaula, R., and Dourish, P. Cognitive and Cultural Views of Emotions. In Proc. HCIC, (2005).
4. Eid, M., and Diener, E. Intraindividual variability in affect, reliability, validity, and personality correlates. Journal of Personality and Social Psychology, 76, (1999), 662-676.
5. Erez, A., and Isen, A.M. The Influence of Positive Affect on Components of Expectancy Motivation. Journal of Applied Psychology 87, 6 (2002), 1055-1067.
6. Forgas, J.P., and Bower, G.H. Mood effects on Person-Perception Judgments. Journal of Personality and Social Psychology 53, 1 (1987), 53-60.
7. George, J. M. Group affective tone. In M. A. West (ed.), Handbook of work group psychology (pp. 77-93). Wiley, Chicester, UK, 1996.
8. Hancock, J. T., Landrigan, C., and Silver, C. Expressing emotion in text-based communication. In Proc. CHI 2007, ACM Press (2007), 929-932.
9. Hatfield, E., Cacioppo, J., & Rapson, R. L. Emotional contagion. Cambridge University Press, New York, USA, 1994.
10. Isbister, K., and Höök, K. Evaluating Affective Interfaces: Innovative Approaches. Ext. Abstracts CHI 2005, ACM Press (2005), 2119.
11. Isbister, K., Höök, K., Sharp, M., and Laaksolahti, J. The sensual evaluation instrument: developing an affective evaluation tool. In Proc. CHI 2006, ACM Press (2006), 1163-1172.
12. Isen, A.M. Positive Affect and Decision Making, in M.Lewis & J.M. Haviland-Jones (eds.), Handbook of Emotions, second edition, The Guildford Press, 417- 435, 2000.
13. Isen, A. M., and Erez, A. Some measurement issues in the study of affect. In A. Ong, & M. vanDulman (eds.), Oxford Handbook of Methods in Positive Psychology. Oxford University Press, New York, USA, 2006.
14. Lee, A.Y., and Sternthal, B. The Effects of Positive Mood on Memory. The Journal of Consumer Research 26, 2 (1999), 115-127.
15. Mandryk, R. L., Atkins, M. S., and Inkpen, K. M. 2006. A continuous and objective evaluation of emotional experience with interactive play environments. In Proc. CHI 2006, ACM Press (2006), 1027-1036.
16. Norman, D. Emotion and Design: Attractive Things Work Better, Interactions, 9, 4, (2002), 36-42.
17. Norman, D. Emotional Design: Why We Love (or Hate) Everyday Things. Basic Books, 2003.
18. Picard, R.W. Affective Computing. MIT Press, Cambridge, USA, 1997.