Collecting measurements
One of the things I will need to do is collect data … from people.
However, this is a little more complex than it looks at first glance.
Identifying emotion from physiological measures
I think this really needs a post on it’s own.
People are complex. A lot of that complexity comes from emotions. Our emotions are our own. So no two people will react the same way to a situation, event, circumstances, stimulus … whatever. What research is in this area is trying to establish is that specific emotions will elicit specific and identifiable physiological responses that are consistent across people. There are lots of things we can measure about our physiological processes. The aim is to identify, for different emotional response, the right combination of these physiological responses.
Not easy.
So I have been searching literature, to see what researchers have found about this. Here’s my whiteboard summary that I have developed so far, updated from yesterday.
A little hard to read. The papers have looked at:
- enjoyment
- fear, sadness, neutral
- stress
The real issue is that not everyone uses completes the same analysis to determine significance of the changes in the measures and/or correlation to the targeted emotion.
At this point, I am just looking for an indicator. My initial thoughts are Galvanic Skin Response is a good indicator. It is a broad brush indicator, just a measure of arousal. But it’s easy to measure.
The one that I am really interested in is surface EMG. Mainly because it is quick (see below) and measures somethings that is also visible. I wonder if this along with a few other simple measures might be a good combination.
Time
The real issue is timing.
Whilst I am yet to do a literature search on this, timing for the response is critical. Responses in the range of 100–200ms are important. So far, only EMG is going to allow this.
