The user behaviour while in interaction with an interface designed for measuring the perception of emotion can give us an idea about the cognitive load of the particular interface in different conditions. Parameters like gaze fixation, saccade movements, gaze counts and so on are there, but what are the most used and reliable parameters.
Hi Junmoni
Not sure I understand the purpose of your test - What will you be asked your participants to look at, or is that yet to be decided? I would have thought that your outcomes would be task dependent? You might want to message two colleagues of mine Dr Sam Stuart (conducted a lot of work in eye tracking) and Dr Joanna Wincenciak (examines emotion recognition, visual system processing). Both on ResearchGate
Regards
Alan
Dear Alan,
Thank you for your reference. We have designed interfaces and expect that they will be perceived differently after they listen to the input-music.
Depending on how people will use the interface, I recommend using eye-tracking based estimators of "workload" or "cognitive load" as well as other non-eye-tracking estimators that have a long-history of application in HCI and Human Factors research. Some metrics for cognitive load from eye tracking that I have observed were the "nearest neighbor index" (refer to the work of F. Di Nocera from around 2005-2007) and pupil diameter (easy to find many examples on Google scholar). As an external validation, of sorts anyway, you could refer to the standard human factors assessment of workload, the NASA Task Load Index (or NASA-TLX for short; Hart & Staveland, 1988, Advances in Psychology). The downside of the NASA TLX is that it is not dynamic and cannot be examined DURING task performance without also interrupting the task of interest. A final option, which I know much less about, is the use of behavioral indicators of workload - some have suggested, for instance, that one could use reaction time (or response time) as an indication.
We've successfully used pupil diameter as an index of cognitive load (with an Eyelink II) and the same tasks that indicated greater load also raised the score afterwards on the NASA-TLX. But if you do use pupil, be sure to compare its diameter under equal illumination and at similar gaze angles, as the eye tracker gives different pupil size outputs in oblique gaze, just because the camera image becomes elliptical.
Percent change in Pupil Dilation is a good measure for cognitive load.
Thanks Idris. I have read some papers, they have have used pupil dilation as a measure.
- Badges
- Science topic