Personas, Goals, and Emotional Design
IRB application writing tips | IRB guidelines
BIOPAC video & screencasts | my notes
BIOPAC Training Seminar
We followed the data analysis technique as suggested in
Wise, K., Bolls, P.D., Schaefer, S.R. (2008). Choosing and reading online news: How available choice affects cognitive processing. Journal of Broadcasting & Electronic Media, 52, 69-85.
To achieve this, fir the waveforms have to be transformed into an excel spreadsheet so that we can see numbers rather than waves. The next step would be to reduce those waves to what we need. For the purpose of our study, we divided the time taken to view a stimulus into 3 equal sections and compared them to learn about the pattern of emotional and/or cognitive change. We also calculated the baseline (which is important and essential as that changes for every person) meaning that a person might have a high wave not because of the stimulus but because his heart normally beats faster. We then subtracted the baseline from each of the 3 recordings that we received per stimulus to get to the fluctuation that can be attributed to the stimulus and this is what can be termed as the final dataset.
This screencast was prepared by a representative from Biopac: http://www.screencast.com/t/8bSTKjDCtYse
Once we got the waveform between 2 event markers, we divided it into 3 equal sections (as done by Dr Bolls and Dr Wise in their paper). In case of a time that is not divisible by 3, we took it to the nearest approximation (keeping in mind that their sum does not exceed the total time between the two event markers. For example, if the time was 30.1 seconds, we divided it into three 10.3 second intervals (Dr Bolls and Dr Wise did it as 10 secs, 10 secs and 11 secs which you have to do manually.)
The software doesn’t do very well with uneven intervals.
So, with that in mind, select the area between the two event markers, by clicking on the first marker, and then using Ctrl+ left mouse click on the second marker you’ll get a perfect event to event area.
Then call up the cycle detector and set the dialog box to the following:
Then choose the output you want and then click on FIND IN SELECTED AREA. In the above example the Delta T from first flag to second is 816 seconds, divided by 3 gave me 272 seconds…and that’s the interval that is highlighted.
If the total interval time between the two markers is less than the sum of the intervals you want reported, the software will only give you a partial interval count. In other words, and going by the example, the interval is 31 seconds. Well, if you set the divided interval to be 11 seconds (as you only have this choice or 10 seconds), the software will capture only 22 seconds as it cannot go beyond the remaining 9 seconds.
To achieve the baseline reading (reading just before stimulus so that we understand what a person's heartrate pattern is like), we can take the mean heartrate over the first few minutes before the stimulus is shown. The approach I did was to take the heartrate just the second before the stimulus is shown.
To do this, and using the cycle detector, you would do the following:
Set up your cycle detector to look at the events as single points, not intervals
Then click on SELECTION tab and set as shown: