Public Dataset: ERPs for Processing Canonical and Non-canonical Finger Numeral Representations
Fırat Soylu · Harvard Dataverse · 2017
Public Dataset: ERPs for Processing Canonical and Non-canonical Finger Numeral Representations Related Publication: Soylu, F. (in review). ERP differences in processing canonical and non-canonical finger numeral representations. Manuscript submitted for publication. Keywords: Numerical cognition, Finger counting, Montring, Gestures, EEG, ERP Access to dataset (Harvard Dataverse): http://dx.doi.org/10.7910/DVN/BLNR8S Created by: Firat Soylu ([email protected]), University of Alabama, on 11/27/2017 Location: The data was collected in the ELDEN Lab (http://elden.ua.edu) at The University of Alabama, Tuscaloosa. DESCRIPTION OF DATA The stimuli for the EEG session included 24 pictures of hand number gestures; 4 cardinal finger montring, 4 ordinal finger counting, and 4 unconventional finger gestures, separately for left and right hands, all showing the palm and matching with numerosities from one to four. The unconventional gestures were based on a previous study comparing montring and unconventional gestures (Di Luca et al., 2010). The gesture images were shot with a digital camera, and were edited to replace the background with a uniform black and to balance color and luminance. The experiment included a total of 960 trials in10 blocks, each block including 96 trials, generated by combining four sets of the 24 gestures, each of them randomized separately, which allowed an even distribution of different stimuli across each block while avoiding predictability. In each trial a number gesture was presented for 500 ms, followed by a validation step, where a single-digit Arabic numeral was presented. Participants pressed one of the two buttons on the controller using either their left or right index finger to indicate whether the Arabic numeral shown matches the number presented in the preceding gesture. To counterbalance use of response buttons, participants used one of the two (right: match, left:no-match, or, left:match, right:no-match) response button configurations in the first five blocks, and the other one in the remaining five blocks, the order randomly chosen for each subject. The EEG portion of the experiment took place in a sound attenuated experiment room. Neurobs Presentation (www.neurobs.com) was used for stimulus presentation and data collection. EEG Data was collected using a BrainVision 32 Channel ActiChamp system (www.brainvision.com), with Easy Cap recording caps using Ag/AgCl electrodes. The 32 electrodes were attached according to the international 10-20 system at the locations Fp1/2, F7/8, F3/4, Fz, FT9/10, FC1/2, FC5/6, T7/8, C3/4, Cz, TP9/10, CP1/2, CP5/6, P7/8, P3/4, Pz, O1/2, Oz and referenced to Cz. BrianVision Recorder was used to record data (electrode impedance<10 kΩ, 0.5-70 Hz, 500 samples/sec). A custom MATLAB script using ERPLAB (http:// erpinfo.org/erplab/) and EEGLAB (http://sccn.ucsd.edu/eeglab) functions were used to analyze data. Inferential statistics was conducted with SPSS (http://www.spss.com). A Logitech F310 game controller was used as the input device. Participants used their left and right index fingers to provide input. HOW TO USE After you download the data, uncompress it using the command "tar -zxvf Soylu(2017)_ERP_Data.tar.gz" in MAC OS X or Linux/Unix. For windows you might need to download additional software. The uncompressed folder will have the raw data and the necessary empty folders for the script to work. You can use the "AnalysisScript.m" script under the "Scripts" folder to produce the results reported in the related publication (see above). You will need to change "home_path" in the script with the location of the main folder for the script to work. You will also need to have EEGLAB & ERPLAB installed in your MATLAB.