The Data Export tool in Pro Lab enables you to export the raw and filtered GSR data along with gaze data and stimulus display information. The export produces a tab-separated values file (.tsv) that can be imported into other software for further processing and analysis, e.g., Excel or MATLAB scripts. GSR data must be exported from Pro Lab when you need to apply advanced GSR analyses that are not supported in Pro Lab. Examples of these analyses include: separation of tonic and phasic components of the signal, calculation of SCR by means of deconvolution analysis in software tools such as Ledalab (MATLAB) or normalization of your GSR data.
Once we generated the recordings with a Shimmer3 GSR+ device , to export the GSR data from Pro lab:
The data is saved in a column in a tab-separated values file (.tsv) together with the selected gaze data, fixations, stimulus and manually logged events. This data can easily be imported and processed further by Excel, Matlab or Python scripts.
Each column of the *.tsv output file contains the type of data specified by the name in the top row of the column. All data types are described in the Tobii Pro Lab User Manual, in section 8.5.
The GSR data will get its own column in the data export. All rows in a data export file have a recording timestamp value (except for the first row which contains column data type name). The timestamp is shown in milliseconds and starts at 0 at the beginning of each recording. Since all recorded GSR data samples have been recorded in a sequence, all data points in a recording will have different timestamps. Note: some GSR data points may have the same timestamp as eye gaze data points, while others may have timestamps between two eye gaze data point timestamps. Since, the gaze data points and the GSR data have their own rows in the export file, the relationship between the number of rows and time is not always linear - e.g. when the GSR data point has the same timestamp as the gaze data point it will still originate 2 rows.
In the image above we can also observe that for everyone eye tracker timestamp you have two GSR timestamps, This is due to the fact that the eye tracker and the GSR device have different sampling rates (60 Hz and 120Hz respectively).
All participant, custom and stimuli events will be reported in the "Event" and "Event value" columns. Additionally, all skin conductance responses (SCRs) detected in the GSR signal will also be reported in the “Event” and “Event value” columns. For each SCR, two GSR events will be created: “SCROnset” at the SCR onset timestamp and “SCRPeak” at the SCR peak timestamp.
Benedek, M., & Kaernbach, C. (2010). Decomposition of skin conductance data by means of
nonnegative deconvolution. Psychophysiology. https://doi.org/10.1111/j.1469-
Boucsein, W. (2012). Electrodermal activity. Springer Science & Business Media.
Ledalab (Matlab): http://www.ledalab.de/