Embedding the jsPsych Flanker Task into Qualtrics
Please NOTE!! This tutorial series is based on version 6.x
of jsPsych. Due to extensive changes in jsPsych 7.x and above, many things in this tutorial may not work with jsPsych v7.x and above. So, I instead use the 6.1.0 version
below, which I forked previously. Or, you can get jsPsych v6.3.1 (the last jsPsych 6.x release) from here.
Here, I used v6.1.0
, which is not much different from v6.3.1
**
Let's start from the demo-flanker.html
in the jsPsych repository and embed it to qualtrics. The current demo file was revised to incorporate jsPsych v7, so it's different from what I used to create this tutorial. For details of demo-flanker.html
in jsPsych v6.x, please read the separate tutorial in this GitHub repository.
You can first try the task by clicking HERE.
The start: demo-flanker-task.html
The experiment code, demo-flanker.html
is included in the flanker
folder in this tutorial repository.
Let's make the code work in a separate folder like the hello-world example.
To do so, you need to change the url of jsPsych scripts, stylesheet, and stimuli of the experiment code. In this tutorial, we will change those twice to make the flanker task (1) run on a new folder and (2) use the GitHub-hosted resources.
First, let's make it run locally. To do so, The jsPsych library, plugins, and stylesheet come from the local folder. As you can see, the demo-flanker.html
in this repository looks at the local jspsych-6.1.0
folder for these files.
<script src="jspsych-6.1.0/jspsych.js"></script>
<script src="jspsych-6.1.0/plugins/jspsych-html-keyboard-response.js"></script>
<script src="jspsych-6.1.0/plugins/jspsych-image-keyboard-response.js"></script>
<link rel="stylesheet" href="jspsych-6.1.0/css/jspsych.css"></link>
The experiment also needs stimuli images. So, you need to create a folder called img
in the experiment folder and copy
con1.png,
con2.png,
inc1.png,
and inc2.png into the img
folder.
Your folder should look like this (and imagine the html file has the name demo-flanker.html
).
Using files in public GitHub repositories
To use jsPsych in Qualtrics, the jsPsych javascript and CSS files need to be online and loadable from Qualtrics. You can make your files online using GitHub Pages, as explained in the Hosting jsPsych or the Hosting RT Task.
But you can do so WITHOUT setting up your own GitHub Pages.
If you see a GitHub file that you want to embed in your experiment, you can use jSDelivr to fetch the file like https://cdn.jsdelivr.net/gh/<github-username>/<repository-name>/<file-name>
.
WARNING: If you are developing your experiment and making frequent changes, jsDelivr is NOT your best option because it takes up to 24 hours to serve the modified files. Read How to use jsDelivr to learn more about jsDelivr and Content Delivery Network (CDN) in general.
Using this tutorial's GitHub files for the flanker task
This jsPsych-in-Qualtrics repository has the files necessary to run the flanker task.
These files can be embedded via jSDelivr (github-username: kywch
, repository-name: jsPsych-in-Qualtrics
, experiment-folder: flanker
).
You need eight files to run the flanker experiment.
jspsych.css
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/css/jspsych.cssjspsych.js
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/jspsych.jsjspsych-html-keyboard-response.js
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/plugins/jspsych-html-keyboard-response.jsjspsych-image-keyboard-response.js
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/plugins/jspsych-image-keyboard-response.jscon1.png
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/img/con1.pngcon2.png
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/img/con2.pnginc1.png
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/img/inc1.pnginc2.png
: https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/img/inc2.png
First transformation: demo-flanker-transformed.html
and flanker_main.js
I prefer to separate out the experiment-related code from the operation-related code, following the Experiment Factory style. If you haven't, please visit the Experiment Factory library, which contain a ton of experiments with running demos and their code.
In the demo-flanker.html
, constructing the timeline
can be considered to be experiment-related,
and the remaining parts that also appear in the Hello-World example, such as loading jspsych.js
and jsPsych.init()
, can be considered to be operation-related.
Here, I am separating the experiment-related code and the operation-related code and putting the code into flanker_main.js
and demo-flanker-transformed.html
, respectively, by following steps:
Step 1. Copy demo-flanker.html
to demo-flanker-transformed.html
Step 2. Create a new file flanker_main.js
And open flanker_main.js
in a new editor window.
Step 3. Cut and paste the experiment-related code
Open demo-flanker-transformed.html
. You should have both flanker_main.js
and demo-flanker-transformed.html
opened.
Select the code from the /* experiment parameters */
line to the timeline.push(debrief_block);
line, cut the selection, and paste it to flanker_main.js
. REPEAT. The code is being cut-and-pasted (NOT copy-pasted).
/* experiment parameters */
var reps_per_trial_type = 4;
/*set up welcome block*/
var welcome = {
type: "html-keyboard-response",
stimulus: "Welcome to the experiment. Press any key to begin."
};
/*set up instructions block*/
var instructions = {
type: "html-keyboard-response",
stimulus: "<p>In this task, you will see five arrows on the screen, like the example below.</p>" +
"<img src='img/inc1.png'></img>" +
"<p>Press the left arrow key if the middle arrow is pointing left. (<)</p>" +
"<p>Press the right arrow key if the middle arrow is pointing right. (>)</p>" +
"<p>Press any key to begin.</p>",
post_trial_gap: 1000
};
/*defining stimuli*/
var test_stimuli = [
{
stimulus: "img/con1.png",
data: { stim_type: 'congruent', direction: 'left'}
},
{
stimulus: "img/con2.png",
data: { stim_type: 'congruent', direction: 'right'}
},
{
stimulus: "img/inc1.png",
data: { stim_type: 'incongruent', direction: 'right'}
},
{
stimulus: "img/inc2.png",
data: { stim_type: 'incongruent', direction: 'left'}
}
];
/* defining test timeline */
var test = {
timeline: [{
type: 'image-keyboard-response',
choices: [37, 39],
trial_duration: 1500,
stimulus: jsPsych.timelineVariable('stimulus'),
data: jsPsych.timelineVariable('data'),
on_finish: function (data) {
var correct = false;
if (data.direction == 'left' && data.key_press == 37 && data.rt > -1) {
correct = true;
} else if (data.direction == 'right' && data.key_press == 39 && data.rt > -1) {
correct = true;
}
data.correct = correct;
},
post_trial_gap: function () {
return Math.floor(Math.random() * 1500) + 500;
}
}],
timeline_variables: test_stimuli,
sample: {
type: 'fixed-repetitions',
size: reps_per_trial_type
}
};
/*defining debriefing block*/
var debrief = {
type: "html-keyboard-response",
stimulus: function () {
var total_trials = jsPsych.data.get().filter({
trial_type: 'image-keyboard-response'
}).count();
var accuracy = Math.round(jsPsych.data.get().filter({
correct: true
}).count() / total_trials * 100);
var congruent_rt = Math.round(jsPsych.data.get().filter({
correct: true,
stim_type: 'congruent'
}).select('rt').mean());
var incongruent_rt = Math.round(jsPsych.data.get().filter({
correct: true,
stim_type: 'incongruent'
}).select('rt').mean());
return "<p>You responded correctly on <strong>" + accuracy + "%</strong> of the trials.</p> " +
"<p>Your average response time for congruent trials was <strong>" + congruent_rt + "ms</strong>.</p>" +
"<p>Your average response time for incongruent trials was <strong>" + incongruent_rt + "ms</strong>.</p>" +
"<p>Press any key to complete the experiment. Thank you!</p>";
}
};
/*set up experiment structure*/
var timeline = [];
timeline.push(welcome);
timeline.push(instructions);
timeline.push(test);
timeline.push(debrief);
Step 4. flanker_main.js
: Use the GitHub-hosted stimuli images
Let's use the above GitHub-hosted images.
The flanker_main.js
in this tutorial has a few changes to do so. Let's see these changes.
Change 1: Adding the image hosting site
The repo_site
variable was added to locate the JSDELIVR site that hosts stimuli images. You can easily use your repository by changing this line.
// Use JSDELIVR to get the files from a GitHub repository
// https://cdn.jsdelivr.net/gh/<github-username>/<repository-name>/<specific folder or file>
var repo_site = "https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/";
Change 2: Adding repo_site
in instructions
The repo_site
variable should be added to wherever img/con1.png
, img/con2.png
, img/inc1.png
, and img/inc2.png
are used like the below.
The find
function really helps in doing so.
var instructions = {
type: "html-keyboard-response",
stimulus: "<p>In this task, you will see five arrows on the screen, like the example below.</p>" +
"<img src='" + repo_site + "img/inc1.png'></img>" +
"<p>Press the left arrow key if the middle arrow is pointing left. (<)</p>" +
"<p>Press the right arrow key if the middle arrow is pointing right. (>)</p>" +
"<p>Press any key to begin.</p>",
post_trial_gap: 1000
};
Change 3: Adding repo_site
in test_stimuli
var test_stimuli = [
{
stimulus: repo_site + "img/con1.png",
data: { stim_type: 'congruent', direction: 'left'}
},
{
stimulus: repo_site + "img/con2.png",
data: { stim_type: 'congruent', direction: 'right'}
},
{
stimulus: repo_site + "img/inc1.png",
data: { stim_type: 'incongruent', direction: 'right'}
},
{
stimulus: repo_site + "img/inc2.png",
data: { stim_type: 'incongruent', direction: 'left'}
}
];
Step 5. demo-flanker-transformed.html
: Import flanker_main.js
Cutting the experiment-related code makes the demo-flanker-transformed.html
file much shorter.
Let's also use the GitHub-hosted JS and CSS files (above).
In addition, you also need to load flanker_main.js
by using a <script>
tag like the other jspsych plugin files (see the original jsPsych v6.3 tutorial).
You can use your own repository by creating the jsDelivr link by following this guide.
CHANGE 1: Using the GitHub-hosted JS and CSS files
<!-- CHANGE 1: Using the GitHub-hosted files -->
<script src="https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/jspsych.js"></script>
<script src="https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/plugins/jspsych-html-keyboard-response.js"></script>
<script src="https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/plugins/jspsych-image-keyboard-response.js"></script>
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/css/jspsych.css">
CHANGE 2: Loading the main experiment script in the same local folder
<!-- CHANGE 2: Loading the main experiment script in the same folder-->
<script src="flanker_main.js"></script>
Does this work?
Let's find out by opening demo-flanker-transformed.html
in a browser.
Step 6: Upload demo-flanker-transformed.html
and flanker_main.js
to GitHub
If this works, you may want to start a new GitHub repository for this task, as I did with my simple Reaction Time Task.
Or, you may just want to try this flanker task in the Qualtrics. In that case, you can use the flanker_main.js
in this tutorial GitHub repository,
which is at https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/flanker_main.js and available for Qualtrics to load.
Second transformation: demo-flanker-transformed-with-display-element.html
Adding the display_element
parameter in the jsPsych.init()
and having additional control over the display element is important for embedding jsPsych into Qualtrics.
The demo-flanker-transformed-with-display-element.html
file in this GitHub repository contains three changes from the demo-flanker-transformed.html
.
When you open demo-flanker-transformed-with-display-element.html
in a browser, you should see the same flanker task.
Let's look at each change.
Change 1: Using display_element
By adding the display_element
parameter in the jsPsych.init()
like below, jsPsych will render the experiment in the <div>
called display_stage
, which the current script does not have yet. So, let's add the display_stage
.
/* start the experiment */
jsPsych.init({
timeline: timeline,
/* Change 1: Using `display_element` */
display_element: 'display_stage',
on_finish: function () {
jsPsych.data.displayData();
}
});
Change 2: Adding display_stage
CSS and Div
The display_stage
needs below CSS and HTML code after the <body>
tag, which also include the display_stage_background
. Both are necessary to function well.
<style>
#display_stage_background {
width: 100vw;
background-color: white;
z-index: -1;
}
#display_stage {
position: fixed;
left: 1vw;
top: 1vh;
height: 98vh;
width: 98vw;
background-color: white;
box-shadow: 1px 1px 1px #999;
border-radius: 15px;
z-index: 0;
overflow-y: hidden;
overflow-x: hidden;
}
</style>
<div id='display_stage_background'></div>
<div id='display_stage'></div>
Change 3: Adding extra scripts for Qualtrics
The below scripts are not necessary to run this reaction time experiment, but you need to copy-paste these in Qualtrics later.
One critical point is that now we use the jsPsych stylesheet hosted in the GitHub Pages (explained above in the Using the jsPsych files section) instead of the CSS file in your computer. This is important because Qualtrics can also access the GitHub-hosted files.
<!-- Change 3: Adding extra scripts for Qualtrics -->
<!-- Use JSDELIVR to get the files from a GitHub repository
https://cdn.jsdelivr.net/gh/<github-username>/<repository-name>/<specific folder or file> -->
<link href="https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/css/jspsych.css" rel="stylesheet" type="text/css"></link>
<div>
<span style="font-size: 24px;">
<br><br>
If you are seeing this message for <span style="color: rgb(255, 0, 0);"><b>more than 5 minutes</b></span>,<br>
please screen-capture this screen and send the image to us.
<br><br>
<span style="font-size: 28px;">We are very sorry for the inconvenience.</span>
</span>
</div>
Third transformation: demo-flanker-qualtrics.js
To make your jsPsych experiment work in Qualtrics, you need to add your code to the below skeleton provided by Qualtrics Question JavaScript Editor. For details, see the Add JavaScript help page.
Qualtrics.SurveyEngine.addOnload(function()
{
/*Place your JavaScript here to run when the page loads*/
});
Qualtrics.SurveyEngine.addOnReady(function()
{
/*Place your JavaScript here to run when the page is fully displayed*/
});
Qualtrics.SurveyEngine.addOnUnload(function()
{
/*Place your JavaScript here to run when the page is unloaded*/
});
The demo-flanker-qualtrics.js
file in this GitHub repository contains several changes from demo-flanker-transformed-with-display-element.html
and can be direclty copy-pasted into the Qualtrics Question JavaScript Editor.
By separating the experiment-related code from the operation-related code, converting the html file to Qualtrics JavaScript function is pretty much similar to the Hello-World example.
Let's look at each change.
Change 1: Hiding the Next button
The below javascript code hides the Next button and puts the javascript code in the driving seat.
// Retrieve Qualtrics object and save in qthis
var qthis = this;
// Hide buttons
qthis.hideNextButton();
Change 2: Defining and load required resources
The below javascript defines where the necessary files are so that Qualtrics can load these. If you enabled GitHub Pages of your repository, Qualtrics can load the necessary JS, CSS, and image files right away. Or, you can provide the embeddable links via jsDelivr. Here, let's continue to use jsDelivr.
One way to check whether these files are accessilbe is to view the source of demo-flanker-transformed.html
and click the imported scripts. For example of jspsych.js
, https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/jspsych.js.
// https://cdn.jsdelivr.net/gh/<github-username>/<repository-name>/<experiment-folder>
var task_github = "https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/";
// requiredResources must include all the JS files that demo-simple-rt-task-transformed.html uses.
var requiredResources = [
task_github + "jspsych-6.1.0/jspsych.js",
task_github + "jspsych-6.1.0/plugins/jspsych-html-keyboard-response.js",
task_github + "jspsych-6.1.0/plugins/jspsych-image-keyboard-response.js",
task_github + "flanker_main.js"
];
function loadScript(idx) {
console.log("Loading ", requiredResources[idx]);
jQuery.getScript(requiredResources[idx], function () {
if ((idx + 1) < requiredResources.length) {
loadScript(idx + 1);
} else {
initExp();
}
});
}
if (window.Qualtrics && (!window.frameElement || window.frameElement.id !== "mobile-preview-view")) {
loadScript(0);
}
Change 3: Appending the display_stage Div using jQuery
In Qualtrics, jQuery (loaded by default) is used to append the display_stage_background
and display_stage
Divs. The CSS for these elements will be added directly to Question HTML later in this tutorial.
// jQuery is loaded in Qualtrics by default
jQuery("<div id = 'display_stage_background'></div>").appendTo('body');
jQuery("<div id = 'display_stage'></div>").appendTo('body');
Change 4: Wrapping jsPsych.init() in a function
The main experiment codes are wrapped in the initExp
function to make sure it runs after all the necessary library and plugin files are loaded (as defined in the loadScript
function above).
Change 5: Summarizing and save the results to Qualtrics
The jsPsych v6.3 tutorial shows you how you can process and summarize the experiment results online. Wouldn't it be great if your experiment results are directly included in your Qualtrics data?
Well, you can actually summarize the results online and save those results in Qualtrics' Embedded Data using the setEmbeddedData function.
demo-flanker-qualtrics.js
does so when the jsPsych experiment finishes by running the below code from debrief_block
.
jsPsych.init({
timeline: timeline,
display_element: 'display_stage',
on_finish: function (data) {
/* Change 5: Summarizing and save the results to Qualtrics */
// summarize the results
var total_trials = jsPsych.data.get().filter({
trial_type: 'image-keyboard-response'
}).count();
var accuracy = Math.round(jsPsych.data.get().filter({
correct: true
}).count() / total_trials * 100);
var congruent_rt = Math.round(jsPsych.data.get().filter({
correct: true,
stim_type: 'congruent'
}).select('rt').mean());
var incongruent_rt = Math.round(jsPsych.data.get().filter({
correct: true,
stim_type: 'incongruent'
}).select('rt').mean());
// save to qualtrics embedded data
Qualtrics.SurveyEngine.setEmbeddedData("accuracy", accuracy);
Qualtrics.SurveyEngine.setEmbeddedData("congruent_rt", congruent_rt);
Qualtrics.SurveyEngine.setEmbeddedData("incongruent_rt", incongruent_rt);
/* Change 6: Adding the clean up and continue functions.*/
// clear the stage
jQuery('#display_stage').remove();
jQuery('#display_stage_background').remove();
// simulate click on Qualtrics "next" button, making use of the Qualtrics JS API
qthis.clickNextButton();
}
});
Change 6: Adding the clean up and continue functions
When the jsPsych ends, display_stage
and display_stage_background
should be removed. Then, execulte the clickNextButton
to simulate clicking the Next button and proceed to the next question.
Finally, embedding jsPsych in Qualtrics
Let's log in to Qualtrics and take a look at each step.
Step 1. Create a new Qualtrics project and then a new question
Create a new Qualtrics project. Then, create a new question and (1) change its format to Text/Graphic.
Step 2. Open the Question JavaScript editor and copy paste the whole demo-simple-rt-task-qualtrics.js
Keep going in the above picture. (2) Click the gear to open the dropdown menu, and then (3) find and click Add JavaScript.
Step 3. Open the Question HTML editor and copy paste the portion of demo-flanker-transformed-with-display-element.html
The Change 3 section of the demo-flanker-transformed-with-display-element.html
file contains the link to jsPsych CSS file, the inline styles for display_stage
Div, and the error message to be displayed when things go wrong.
To open the HTML editor, click the HTML View
button.
Then, copy paste the portion of demo-flanker-transformed-with-display-element.html
to the HTML editor, starting from <!-- COPY PASTE TO QUALTRICS FROM HERE -->
to <!-- COPY PASTE TO QUALTRICS UP TO HERE -->
.
<!-- Change 3: Adding extra scripts for Qualtrics -->
<!-- COPY PASTE TO QUALTRICS FROM HERE -->
<link href="https://cdn.jsdelivr.net/gh/kywch/jsPsych-in-Qualtrics/flanker/jspsych-6.1.0/css/jspsych.css" rel="stylesheet" type="text/css"></link>
<div>
<span style="font-size: 24px;">
<br><br>
If you are seeing this message for <span style="color: rgb(255, 0, 0);"><b>more than 5
minutes</b></span>,<br>
please screen-capture this screen and send the image to us.
<br><br>
<span style="font-size: 28px;">We are very sorry for the inconvenience.</span>
</span>
</div>
<!-- Change 2: Adding `display_stage` CSS and Div -->
<style>
#display_stage_background {
width: 100vw;
background-color: white;
z-index: -1;
}
#display_stage {
position: fixed;
left: 1vw;
top: 1vh;
height: 98vh;
width: 98vw;
background-color: white;
box-shadow: 1px 1px 1px #999;
border-radius: 15px;
z-index: 0;
overflow-y: hidden;
overflow-x: hidden;
}
</style>
<!-- COPY PASTE TO QUALTRICS UP TO HERE -->
After copy pasting, you should see something like below.
Step 4. Creating Embedded Data elements
To store the experiment results -- accuracy, congruent_rt, and incongruent_rt -- in Qualtrics, you need to create the Embedded Data elements named accuracy, congruent_rt, and incongruent_rt in your survey by following this Qualtrics tutorial.
- Click Survey Flow from the Survey tab
- Click Add a New Element Here
- Choose Embedded Data
- Click Create New Field or Choose From Dropdown and type accuracy
- Click Set a Value Now and type -1
- Click Add a New Field and type congruent_rt
- Click Set a Value Now and type -1
- Click Add a New Field and type incogruent_rt
- Click Set a Value Now and type -1
Step 5. Move Set Embedded Data
block to the top of Survey Flow
The accuracy
, congruent_rt
, and incongruent_rt
fields should be defined before the jsPsych reaction time task,
so you need to move the Embedded Data block to the top of the Survey Flow.
Click and hold down Move to drag the Embedded Data to the top of the Survey Flow. After this, you should see the Survey Flow like below.
Step 6. Add a check for the Embedded Data
You should make sure that the experiment results are saved correctly.
You can do so by showing the accuracy
, congruent_rt
, and incongruent_rt
Embedded Data after the reaction task.
- Click Add Page Break
- Click Create a New Question
- Change Question Type to Text/Graphic
- Type
${e://Field/accuracy}
,${e://Field/congruent_rt}
, and${e://Field/incongruent_rt}
to access these Embedded Data
Step 7. Publish and test!
Publish the survey by following this Qualtrics tutorial. Then, an anonymous Qualtrics link is generated. If you click this link, you should be able to see the same reaction time task running in Qualtrics. For example, try this Qualtrics link.
Do you also want to save the trial-by-trial data?
Then, you need a server that can receive your data. See the tutorial Saving data to a web server using PHP.