Identifying conceptual neural responses to symbolic numerals
Data files
Apr 23, 2024 version files 2.93 GB
-
DATA_BDFs_SHARE.zip
-
README.md
-
Stimuli.zip
Abstract
The goal of measuring conceptual processing in numerical cognition is distanced by the possibility that neural responses to symbolic numerals are influenced by physical stimulus confounds. Here, we targeted conceptual responses to parity (even vs. odd), using electroencephalographic (EEG) frequency-tagging with a symmetry/asymmetry design. Arabic numerals (2–9) were presented at 7.5 Hz in 50-s sequences; odd and even numbers were alternated to target differential, “asymmetry” responses to parity at 3.75 Hz (7.5 Hz/2). Parity responses were probed with four different stimulus sets, increasing in intra-numeral stimulus variability, and with two control conditions comprised of non-conceptual numeral alternations. Significant asymmetry responses were found over the occipitotemporal cortex to all conditions, even for the arbitrary controls. The large physical-differences control condition elicited the largest response in the stimulus set with the lowest level of variability (1 font). Only in the stimulus set with the highest level of variability (20 drawn, colored exemplars/numeral) did the response to parity surpass both control conditions. These findings show that physical differences across small sets of Arabic numerals can strongly influence, and even account for, automatic brain responses. However, carefully designed control conditions and highly variable stimulus sets may be used towards identifying truly conceptual neural responses.
README: Identifying conceptual neural responses to symbolic numerals
https://doi.org/10.5061/dryad.1zcrjdg13
Talia L. Retter, Lucas Eraßmy & Christine Schiltz
Proceedings of the Royal Society B: Biological Sciences (2024)
Description of the data and file structure
BioSemi BDF files (EEG data)
15 participants (NumberShape_S01-S15); none excluded
CONDITION TRIGGERS:
1 font: parity: 11; small phys.-diff. control: 21; large phys.-diff. control: 31
10 fonts: parity: 12; small phys.-diff. control: 22; large phys.-diff. control: 32
10 mixed: parity: 13; small phys.-diff. control: 23; large phys.-diff. control: 33
20 fonts: parity: 14; small phys.-diff. control: 24; large phys.-diff. control: 34
EEG notes and analysis steps
Data proceeded with LetsWave 6 (https://www.letswave.org/, running over Matlab R2019b (MathWorks, USA))
IMPORT/BASIC PREPROCESSING
1. import bdfs
1b. assign electrode labels+coordinates
*S05: cut first block, recorded without glasses and unused (from 263 to 3017 s, w/cropper)
**S10: one ep 21 trigger missing, can recover at end of fade-in (1229.0879 s)
***some trigger fixes: S06, 38-->34; S10, 28-->24; S13: 37-->34; S15: 36-->32
2. bc
3. Butterworth bandpass filter, hpf 0.1, lpf 80 order 4 *bbpf
4. Segment -1 for 56 s (1 pre,2 fade-in;2 fade-out,1 post) *epm
PREPROCESSING CONTINUED: CLEANING
5. electrode check - replacements? +delete status *ch68
6. ICA for blinks >.2/s (3 subj only: S05:IC2,S06:IC1,S07:IC1)
7. interpolation of noisy channels
S01 none
S02 P1,F6
S03 FC1,P2
S04 P2,P4
S05 none
S06 FP2,PO9
S07 CP1,P9
S08 P1,PO3
S09 CP1
S10 P1,T8
S11 none
S12 F7,AF3,AF8
S13 none
S14 none
S15 FPz,Afz,P2
8. rereference
9. Crop epochs by bin *epb (don't merge) 2 s to 25532 bins
FREQUENCY-DOMAIN PROCESSING
10. Average epochs *avg
11. FFT *fft
12. Chunk for harmonics
-latency: 3.439198774
-duration: 0.621602452 (31 bins)
-interval: 3.75
13. baseline sub 2-11 ex. 1 *sbl
14. sum epochs (1:3.75 Hz; 3:11.25 Hz; 5:18.75 Hz for asymmetry responses)
Demographics
Age | Sex | Handedness | |
---|---|---|---|
S01 | 24-26 | F | R |
S02 | 21-23 | M | L |
S03 | 30-32 | F | R |
S04 | 27-29 | F | R |
S05 | 18-20 | F | R |
S06 | 24-26 | F | R |
S07 | 24-26 | F | R |
S08 | 21-23 | F | R |
S09 | 21-23 | F | R |
S10 | 18-20 | F | R |
S11 | 18-20 | F | R |
S12 | 21-23 | M | R |
S13 | 24-26 | F | R |
S14 | 18-20 | F | R |
S15 | 21-23 | F | L |
*age given as an approximate range to preserve individual confidentiality
Stimuli
.jpg image files
Methods
1) EEG data for 15 participants.
-acquired with a BioSemi, ActiveTwo system (.bdf)
-standard 64 channel headcap active recording electrodes + PO9, I1, I2, and PO10 https://www.biosemi.com/headcap.htm
-We processed this data with LetsWave6, an open-source toolbox running over MATLAB (https://www.letswave.org/). Bdf (BioSemi Data Format) files (https://www.biosemi.com/faq/file_format.htm) can also be accessed directly and with many other common EEG analysis toolboxes for Matlab and Python.
Please see the README document for processing details.
2) Stimuli. 4 sets: 1 font; 10 fonts; 10 mixed; 20 drawn.