My code for the SRT Paradigm introduced by Crouzet et al., 2010.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.

This is my code for the Saccadic Reaction Time paradigm, as first introduced by Sébastien M. Crouzet et al. in his paper Crouzet, S. M., Kirchner, H., & Thorpe, S. J. (2010). Fast saccades toward faces: face detection in just 100 ms. Journal of Vision, 10(4), 16.1–17. I used this code to test our setup for our work on familiar faces, Visconti di Oleggio Castello, M., & Gobbini, M. I. (2015). Familiar Face Detection in 180ms. PLoS ONE, 10(8), e0136548.

If you use this code, please acknowledge the original paper that introduced the paradigm, and share your improvements with pull requests.

General requirements

The code requires

  • Psychtoolbox 3 (the code was run with version 3.0.11)
  • MATLAB (tested with version 2014b. It might work with octave, but I haven't tried it)
  • Eyelink Plus 1000 system—the code needs to be modified if you plan to use it with other eye-tracking systems.

Also, you won't find stimuli in this repository as I do not have the rights to share them. Please contact the author of the original paper (Sébastien M. Crouzet) to obtain them.

How it works

The code could be made more user-friendly (I welcome pull requests), however this is how it works as of now.

The structure of the folder is as follows: the main code resides in the code directory; subject-specific trial-order files are under the csv directory, with one folder for each subject id, for example csv/mv00. In the subject-specific folder you need two txt files containing the filenames of the stimuli corresponding to faces and objects. These stimuli must exist in the stim directory. See csv/test-before as an example of the directory before running make_order_trials (see below).

My approach was to create a modular code in which the main experiment code reads in CSV files containing the trial order for each block and subject, and another code creates the trial orders. The main files of the experiment are

  • setupExp.m for general setup,
  • make_order_trials.m to generate the CSV files of the trial order, and
  • srt_paradigm.m, the code that reads in the CSV files and shows the experiment.

You must be in the code directory to make the code run. Do not add all the directories to the matlab path. Alternatively, you need to fix setupExp.m.


Here you should only change what is between lines 23–44, in particular the distance of the subject from the screen DIST_CM, the width of the screen SCREEN_W_CM, and the resolution and refresh rate of your system RESOLUTION.

For debugging, you can also set DEBUG to 1 on line 24. In this way there's no need to have the eye tracker connected and it will use mouse position instead of eye position.


This function generate blocks of pseudo-randomized trials, making sure that targets occur in each hemifield equally often within each block. You are welcome to write your own code to generate these CSV files. See below for the type of input that srt_paradigm.m accepts.

To generate the CSV files with this function, you need the following:

>> help make_order_trials
  Saccadic Reaction Time paradigm
  make_order_trials(subid, cfg) makes csv files for subject subid

        subid           subject id

        cfg             a structure with the following fields

           stimulitxt   a cell with the filenames of the txt files containing
                        the filenames of the stimuli. It assumes that these txt
                        files are stored under ./csv/subid/. Stimuli need to be
                        stored in ./stim/
           stimulitype  a cell containing the labels for the stimuli type, 
                        in the same order as stimulitxt 
           imagerep     (optional) number of repetitions for each image
           blocklength  (optional) specify length of the blocks. The program
                        will try to split each task into smaller blocks of
                        length blocklength, with targets balanced into
                        left/right hemifields. If it cannot do it (because
                        blocklength does not evenly divide the total number of
                        trials), it will complain and abort.

        csv files under ./csv/subid/
        a txt file under ./csv/subid/ containing the codes for each condition

Note that it checks that the stimuli exist under ./stim/—better to fail early that during the experiment. As an example of the output, see the files contained in csv/test-after/. Those where generated running the function in the following way (again, if you do not have the stimuli in the stim directory it will fail):

subid = 'test-after';
cfg = [];
cfg.stimulitxt = {'faces.txt', 'objects.txt'};
cfg.stimulitype = {'faces', 'objects'};
cfg.blocklength = 50;

make_order_trials(subid, cfg);

this generated the files in csv/test-after:

├── faces.txt
├── objects.txt
├── test-after_blocks_orig.txt
├── test-after_code_description.txt
├── test-after_tar-faces_dis-objects_1.csv
├── test-after_tar-faces_dis-objects_2.csv
├── test-after_tar-faces_dis-objects_3.csv
├── test-after_tar-faces_dis-objects_4.csv
├── test-after_tar-objects_dis-faces_1.csv
├── test-after_tar-objects_dis-faces_2.csv
├── test-after_tar-objects_dis-faces_3.csv
└── test-after_tar-objects_dis-faces_4.csv

The last important step is to counterbalance the order of the blocks. You need to create a file named subid_blocks.txt where each line is the filename of a block. You can look at csv/test-after/test-after_blocks_orig.txt as an example of an unbalanced block order.

The generated CSV have this structure, which can be read-in by srt_paradigm.m.



This is the code that shows the stimuli to the subject. It is run as srt_paradigm(subid, subnr, blocknr). It starts with a 9-point calibration of the eye-tracker, followed by the experimental paradigm. The trial starts only after the subject has maintained fixation within the allowed range (set to 2 visual degrees from the center of the screen; it can be modified in setupExp.m with the var MAX_DIST_FIXATION_DEG.

When the block ends, the EDF file generated by the eye tracker is moved to res/edf/subid/.

After data collection: analysis

The output EDF can then be converted to ASCII using the EDF2ASCII converter available on the SR research website , and analyzed offline for saccade initiation.

I included an example parser in the directory ana(lysis). It should work with the output generated by this code. If you make any changes to the code of the experiment, especially to the triggers such as the following, you might have to adapt the parser since it uses these triggers as sentinels:

Eyelink('Message', sprintf('Trial %d Code %s: STIM ON', itrl, ...
        output{1+itrl, 7}));

The parser will generate a structure whose design is inspired by the Fieldtrip toolbox:

s = 

         dimord: 'trl_sample_x_y_na'
            fix: {1x162 cell}
          blank: {1x162 cell}
           stim: {1x162 cell}
      trialinfo: [1x162 double]
        targets: {162x1 cell}
    distractors: {162x1 cell}

each of the fields fix, blank and stim contain the x, y position of the eye (and the pupil area, I believe) over time in the respective conditions: initial fixation, blank period, and stimulus presentation. Each element represents a single trial, whose code is stored in trialinfo, with targets and distractors in the respective fields. To sum up, for trial i, the code of that trial is in s.trialinfo(i), the target was s.targets{i} and the distractor s.distractors{i}; the position of the eye is stored in s.stim{i}.