To link your tasks and questionnaires together and send them out to participants, you’ll need to use our experiment tree.
This walkthrough is designed to follow the Task Builder 2 Step-By-Step Guide, so we recommend completing that first!
If you haven’t followed the step-by-step guide to Task Builder 2, you’ll need to create a project before continuing this walkthrough. Click here for a video demonstrating how to do this.
If you have followed the step-by-step guide to Task Builder 2, go ahead and open your Relational Reasoning project in a new tab and you're ready to start! You can find your projects here.
First, create a new experiment inside your project. New experiments come with a start node and a finish node. Nodes are the building blocks of Gorilla experiments: you can find a full list of nodes on our support pages.
Most experiments start with a consent form. Gorilla has a pre-made consent form example, which we’re going to add here. You can of course also create your own questionnaire, or clone and edit our example.
Next, we want to check that our participants are real people who are paying attention to our experiment! To do this, we’re going to use one of Gorilla’s pre-made bot checks: an anagram task, where participants have to use an onscreen keyboard to solve a simple word puzzle. We’ve called this a bot check, but it also serves as an attention check for genuine participants. You can find more ready-to-use bot checks on our samples page.
Once participants have completed the bot check task, we want to branch them based on whether or not they answered correctly. Gorilla makes it easy to branch participants - you just insert a branch node! The bot check task is already set up to record correct answers, using a field in the Store called 'correct', and it includes 3 questions.
All we need to do here is set up one branch for participants who got 3 answers correct, and another for participants who got less than 3 answers correct.
If you want to store any information during a Gorilla experiment to use later in the experiment, you'll want to use the Store. Learn more about how to use the Store in Task Builder 2.
If participants fail the bot check, they’re either bots or people who just aren’t paying attention. Either way, we don’t want them to take part in our experiment, so we’re going to use a reject node to reject them! If you reject a participant before they finish your experiment, you’ll get their participant token back so that another participant can use it. If you'd like, the reject node will also let you redirect rejected participants to another webpage.
If you want to screen your participants in other ways, for example recruiting only over 50s, or only people who are bilingual, you can use the branch and reject nodes in the same way.
Now we’re happy that all our participants are genuine and engaged with our experiment, we’re ready to move on to the main part of our experiment, the Relational Reasoning task.
We’re going to investigate whether there are differences in performance between 2 versions of the Relational Reasoning task: one with a long fixation time, one with a short fixation time. To do this, we want to randomly assign each participant to one of two conditions, using Gorilla’s randomiser node.
We want to create 2 versions of our Relational Reasoning task, with different presentation times for the fixation cross. To do this, we need to set fixation cross presentation time up as a manipulation in the Relational Reasoning task. Use the Relational Reasoning task you created in the Ten Step Relational Reasoning Task Walkthrough.
To learn more about manipulations in Gorilla tasks, see our Task Builder 2 How To Guide.
If you haven’t completed the task walkthrough, click here for a video demonstrating how to add a pre-made Relational Reasoning task to your project.
Now we’ve set fixation time up as a manipulation to our Relational Reasoning task, we want to add two versions of the same task to our experiment. One will have the default fixation time we set and one will have a longer fixation time.
And that’s our experiment completed! Time to preview it to check we’re happy, and then commit it.
Note: at the end of the preview, you'll have the option to download your data. We strongly advise that you check you're happy with the data you receive before starting to recruit participants.
Once you’ve committed your experiment, you can send it out to participants. There are lots of different recruitment policies you can use in Gorilla, but we’re going to use the most straightforward: the simple link.
For more information about recruiting participants and collecting data, read our guide to launching your study on Gorilla.
In order for participants to enter your experiment, you'll need to have purchased participant tokens. We offer different pricing options to suit everyone, from pay-as-you-go to department and institutional subscriptions. For more information, check out our pricing page.
If your university already has a subscription, find out how to join a subscription, and how to request tokens from your subscription.
Once you've added tokens to your account, you're all ready to start recruiting participants. Happy experimenting!
