In LISA, sensor data like EDA and HRV are provided by a learner’s sensor wristband. To better interpret sensor data with respect to the learner’s emotional state, we’ll do an experiment showing emotional pictures and having students rate, while measuring EDA und ECG. The resulting data sets can be analyzed by different methods, using a statistical approach or methods of machine learning, to get insight into the correlation of sensor data and emotional states.
IAPS Emotional Pictures
We are using the International Affective Picture System (IAPS) as provided by the pictures provided by Center for the Study of Emotion and Attention (CSEA), as these pictures have a set of normative emotional stimuli for experimental investigations of emotion. The library of IAPS, previous ratings for the pictures and related research papers were attained.
To select pictures for this emotional pictures experiment of around 40-45 minutes, we have taken several steps to choose 96 pictures. First, we have distributed the equal number of pictures for four intended emotional stimuli (Happy, Angry, Satisfied and Bored) by using the two-dimensional space of valence and arousal. Second, we have selected pictures that has no statistically significant difference between male and female. To do this the mean arousal rating difference between female and male was set less than 0,8 and the mean valence difference was set less than 1. Independent t-test was conducted to check the significance of difference between female and male.
The pictures containing explicit violence and sexually explicit were excluded and pictures with rating value higher than 6 and less than 4 were selected respectively.
In this experiment, a participant will wear EDA and ECG sensors on his or her non-dominant hand and on chest respectively. The experiment will include a baseline task, IAPS picture rating task and video. The duration of the experiment will be 45 minutes.
The experiment is taking in place at H214 and please see the following time slot for November and December 2017 for participation. If you are interested in participating in the experiment, please email here with preferred date&time.
If the following timeslot does not suit you but interested, please contact here.
|01-05 January 2018|
|08-12 January 2018|
|15-19 January 2018|
|22-26 January 2018|