You can use DataPipe with any online experiment. You can even use it with a laboratory experiment, as long as you have an internet connection. This guide will walk you through the steps for a typical online experiment using tools that are widely available and free.
The first step is to create an OSF project for your experiment. You can create an OSF project at https://osf.io. You will need to create an account if you do not already have one. Once you have created an account, click the Create Project button to create a new project. You can name your project whatever you want.
In order for DataPipe to have permission to send files to your OSF account, you need to create an authorization token on the OSF and add the token to your DataPipe account. To create an authorization token, go to your OSF account settings by clicking your name in the top right corner of the screen and selecting Settings. Then click the Personal Access Tokens tab. Click the Create Token button. Give the token a name (we recommend a name that is specific to DataPipe so that you can easily disable the token when you are done using DataPipe) and select "osf.full_write" as the scope. Click the Create Token button to finish creating the token. You will be shown the token value. Copy the token value.
On DataPipe, click the Account button in the top right corner and select Settings. Click the Set OSF Token button and paste the token value into the box. Click Change Token to finish. You should see the icon become a green checkmark to indicate that you have a valid token.
The next step is to create an experiment on DataPipe. Click the New Experiment button in the top right corner. Give your experiment a name and enter the OSF project ID. This ID is part of the URL of the OSF project. For example, if the URL of your OSF project is https://osf.io/abcde/, then the project ID is abcde.
When you create an experiment, DataPipe will automatically create a new Data component on the OSF project. The Data component is where DataPipe will store the data files that it sends to the OSF project. Enter the name you would like to use for the Data component.
Click the create experiment button to finish. You will be sent to the experiment dashboard where you can edit the experiment settings.
There are three optional features that you can enable for your experiment via the experiment dashboard.
Condition assignment will allow you to request the next sequential condition number from DataPipe. For example, if you have 4 conditions in the experiment, DataPipe will respond to the first request with 0, the next request with 1, then 2, then 3, and then cycle back to 0.
Data validation will check the data as it is sent to DataPipe. If the data are invalid, then DataPipe will not send the data to the OSF. The basic data validation features are to check if the data file is a valid JSON or CSV file. Once you have created the experiment, you can also specify a list of required fields that the JSON or CSV must have in order to be considered valid. This is a useful feature to enable because it limits the potential for malicious use of DataPipe. One risk of using DataPipe is that it creates an open path to create files in your OSF project. A malicious and technically savvy user could potentially create spam data and send the files to your OSF account. Turning on validation makes it harder to do this.
The session limit will cap the number of data files that can be sent to your OSF project. This is another way to limit the potential for malicious use. If you set the session limit to 100, then DataPipe will only send the first 100 data files that it receives. You can adjust the session limit later if you need to increase it.
In order to send data to DataPipe, you need to add code to your experiment to communicate with DataPipe. If you are using jsPsych, then you can use the jsPsychPipe plugin. If you are not using jsPsych, then you can use the DataPipe API directly via fetch requests.
After creating an experiment in the previous step, you will be sent to the experiment page. You can also get to the experiment page by clicking My Experiments in the top menu and selecting the experiment you want to view. On this page, there are code snippets for sending data to DataPipe. Select the code snippet for the language that you are using in your experiment and follow the instructions provided on the dashboard to add the code to your experiment.
The next step is to publish your experiment online so that participants can view it. You can use any tool that allows you to publish a web page, such as university web hosting, GitHub Pages, or Netlify. We will describe how to use GitHub Pages, since it is free, accessible, and relatively easy to use. This guide will assume no familiarity with GitHub or git version control. We will describe the easiest way to get started for someone with no experience using GitHub, but if you are already familiar with GitHub, the approach we take here is probably not the best way to do things and you should feel free to follow your own preferred workflow.
First, create a GitHub account at https://github.com if you do not already have one. Then go to https://github.com/new to create a new repository. You can name it whatever you want, but the name that you give it will become part of the URL that you use to access your experiment. Therefore, you may want to avoid names that reveal information that you want to keep hidden from the participants. Check the box to add a README file. The rest of the settings can be left at their default values. Click the "Create repository" button to create the repository.
Next, configure the repository to share its content as a webpage. Go to the Settings tab of your repository. Select the Pages menu item on the left side. For Source, leave it as deploy from a branch. Under branch select main as the source. Click the save button to finish this step.
Now add the experiment files to the repository. In your GitHub repository, click the Add Files button near the top of the screen and select Upload Files. Drag and drop your experiment files into the upload box. You can also click the upload box to select the files from your computer. Once you have uploaded all of your experiment files, click the Commit Changes button.
That's it! Your experiment is now published on the web. You can view it by going to https://[your username].github.io/[your repository name]. If your HTML file is not named index.html, then you need to add the name of the HTML file to the end of the URL. For example, if your HTML file is called experiment.html, then the URL would be https://[your username].github.io/[your repository name]/experiment.html. It may take a few minutes for the uploaded files to be available as a website.
The final step is to activate your experiment. On the experiment dashboard, you can activate three different features of DataPipe for each experiment.
Enable data collection will activate the standard data collection feature. This enables sending text files (e.g., JSON or CSV) to your OSF project.
Enable base64 data collection activates base64-based data collection. Base 64 is a way to encode files as strings, and can be used for collecting data like audio recordings, video recordings, or images.
Enable condition assignment activates the condition assignment feature. This allow you to request the next condition from DataPipe.
We strongly recommend that you only activate the minimum features that you need for your experiment and that you only activate features during active data collection. This will reduce the risk of malicious use of DataPipe.
At this point you should be ready to collect data. We recommend testing data collection carefully at this point to ensure everything is properly configured. You should see data files created on your OSF data component immediately after you finish the experiment.