Dana Kirsch
Department of Brain and Cognitive Sciences
Bachelor of Science in Brain and Cognitive Sciences
May 1997
Rosalind W. Picard
Associate Professor
Whitman A. Richards
Professor Media Arts & Sciences
Prof. Alan V. Hein
Chairman, Department Committee
This paper provides a discussion of the results from three experimental studies on emotion, and presents a new experiment inspired by them, that is a first step toward designing a tool to measure a subject's valenced response. Peter J. Lang and others showed subjects a series of pictures and asked them to self-rate their emotional response. Ward Winton, Lois Putnam, and Robert Krauss measured heart rate and skin conductance while subjects viewed emotionally evocative stimuli. Dr. Manfred Clynes conducted a series of sentic experiments, gathering data from the vertical and horizontal components of finger pressure. Each of these experiments attempted to quantify emotions and map them into a predictive model of emotion theory. Under the auspices of affective computing, these three models are applied to the interaction between humans and computers. Using a computer to provide the affective stimulus to the human subject, an experiment is conducted which combines all three emotion studies. An ordinary computer mouse was augmented with a sensor to collect sentic data as in Dr. Clynes experiments. Subjects were hooked up to various other bio sensors as in the Winton, Putnam, and Krauss tests and viewed the affective picture database from Lang's work. The three measured results: sentic data, heart rate, and self-assessment, are then readily compared against each other as well as the theory predicted results and the valence for each slide. The results show that a strong correlation between the self-reported valence assessment of our subjects and the results from Lang's numerous experiments exists. The data collected from the sentic mouse also significantly correlated to the self-reported information. Valence information can be captured by the sentic mouse.