Classification task template

---

This is a template for a jsPsych-based online categorisation task. It is made to be a lightweight, easy to amend script that you can easily use for your own experiment. The task implemented here is an animacy task, where participants classify images as inanimate or animate.

How to use this?

The repository is centered around the main experiment script, index.html. This is what browsers read to show the task to participants. To run the experiment, you will need a browser to read this file and have access to the files around it. This is mainly done in two ways:

1. Locally: turn off the exp_online flag, and you can run index.html locally. You can simply open it with a browser like Chrome, but it might be the easiest to open the whole repository in a friendly code environment that will handle it along with all the other files (a good choice is VSCode). From there, your script can be ran using your local browser.
2. Online: with the exp_online flag on, this script will try to connect to Pavlovia. If correctly placed in a GitLab repository behind the scenes, your experiment will run and be accessible at the click of a Pavlovia link.

Directory structure

.
├── data
├── demographics
├── informed_consent
├── instructions
│   └── instruction_images
├── readme_files
└── stimuli

About the different folders

• data is where pavlovia automatically sends the collected data.
• demographics contains an external .js file with the content of the quick demographics survey to appear at the start of the experiment.
• informed consent contains an external .js file with the content of the informed consent participants will accept at the start of the experiment, alongside a .css file to determine some visual features of the consent form.
• instructions contains an external .js file containing instructions to be displayed at different moments of the experiment, and serves to de-clutter the main script. It also contains a .css file to style some visual elements of the instructions, and some images in instruction_images to illustrate things.
• stimuli contains your images, including task images (starting with task) and training images (starting with train).
• readme_files contains illustrations for this readme document.

Note.The extra file jspsych-7-pavlovia-2021.12.js comes from this gitlab repository and is the only way around using the readily available jspsych-7-pavlovia-2022.1.1.js which faces consistent issues with terminating the pavlovia connection. It's a temporary solution.

Parameters

Here are the parameters to set before running your script:

• exp_online determines whether your script tries to connect to Pavlovia.

• exp_fullscreen determines whether to run the experiment in full screen.

• ask_informed_consent should participants are presented with an informed consent to sign.

• ask_demographics should participants are asked for some demographic information.

• exp_instructions show instructions at the start of the experiment or not

• exp_training play training trials at the start of the experiment.

• n_runs number of times all the trial are shown. Each run is divided in n_bocks.

• n_blocks how many blocks a full run is divided into, with breaks in between each block.

• img_width commented out by default, uncomment to constrain the width of your images.

• img_height commented out by default, uncomment to constrain the height of your images.

• jitter_duration how long the jitter can be, from 0 to x.

• fixation_duration pre-stimulus fixation duration (ms).

• image_duration target stimulus presentation duration (ms).

• training_threshold successive correct trials needed to pass training.

• response_labels what are the participants responding?

• response_buttons declare the keys to use to respond.

• debugging_mode turn this on to make the experiment much shorter (useful for debugging.)

• debugging_length how many trials of training and main task are played in debug mode.

• save_data_locally whether or not to download a copy of the created data locall at the end of the experiment.

Experiment timeline

When all events are turned on, here is what the timeline of a full experiment looks like:

1. The connection is initialised with jsPsych and Pavlovia.
2. All file names contained in the stimulus list are preloaded.
3. The experiment starts in full screen.
4. The informed consent form is presented and signed.
5. The demographics questionaire is presented and filled.
6. The main instructions are displayed.
7. Training starts:
   • Training instructions come on screen.
   • Training trials are played
   • Training finishes when a threshold is reached or all the training stimuli have been presented.
   • Post-training instructions come on screen.
8. Main task starts:
   • Pre-task instructions are shown.
   • Runs follow each other. Each run is divided into blocks, with breaks in between.
   • A quick post-task debriefing is shown.
9. The connection is closed with Pavlovia.

Stimulus list

To have jsPsych read and use your files, you need to refer to them in your main script. There are two ways of doing so.

1. If you don't have a lot of images, it's probably the easiest to list them inside your index.html, like this: let stimuli = ['./stimuli/image1.png, './stimuli/image2.png', …]
2. In most cases, if you have a lot of images and/or if you want a cleaner script, your can make use of an external .js file. This is in use by default in this template. Such a file should contain a list of stimuli like this:

var stimuli = [
  {
    "filename": "./stimuli/image1.png"
  },
  {
    "filename": "./stimuli/image2.png"
  },
  ...
 ]

You can create this external .js file as you please. One way to go about it is illustrated in make_stim_list.py, a python script that will take the stimuli you uploaded in the stimuli folders and automatically create an external .js list for you (see stimuli.js).