Construal Level International Multilab Replication Project
Join a global network of labs in the largest ever collaborative effort
to evaluate construal level theory.
Status update: The study was submitted as a Registered Replication Report to Advances in Methods and Practices in Psychological Science (AMPPS) on June 23 2020. All labs who registered before June 30 should have received an email with the manuscript and the full procedures to use for ethics applications.
What’s the project’s aim? To replicate key experimental findings underpinning Construal Level Theory (CLT)—a theory that suggests that psychologically distant things are more abstractly represented than psychologically close ones.
Which studies are we aiming to replicate? CLT suggests that there are four forms of psychological distance: temporal, spatial, social, and likelihood. We will replicate one key study examining each of these four forms of psychological distance. The four studies we plan to replicate can be found here.
How will we do it? We will use a multi-lab approach where each lab collects data on all four studies. Rather than replicating a single empirical finding, this approach allows us to subject CLT to multiple testing from different angles.
What’s the timeline? We submitted a Registered Replication Report to Advances in Methods and Practices in Psychological Science (AMPPS) on June 23 2020. The plan is that data collection will take place from the beginning of 2021 until the beginning of 2022.
What’s required of you?
· Collect data from a minimum of 100 participants.
· Complete data collection within the data collection window, which we estimate will be from early 2021 until spring 2022.
· Have access to computer(s) on which participants complete the study.
· If necessary, contribute to the translation and/or language-appropriate adjustment of the study materials (applies to non-Swedish and non-English speaking labs only).
What about statistical power? As of June 30 2020, 95 labs have confirmed that they will contribute to the project. This will give us a total N of at least 9500. With our design, we’ll have 2375 participants for each experiment. This sample size will provide 99.17% power to detect an effect of d = .20 (assuming 1% heterogeneity)—less than half the size of the effect estimated in the latest meta-analysis of CLT. You can read more about power here.
What’s in this for you? Co-authorship on one of the first large-scale multi-lab replications on Construal Level Theory! This is an ideal project for an ambitious student.