## The Analysis Scripts ## They are very long, we know, and the commenting isn't all that great, honestly. ---------- The scripts in this component do all the work of taking raw data and turning it into statistics and tables. Of these, the **All-Indicator** file is the more daunting. This one takes the [*raw data*][1] in and creates all the models, objects, and refined material that a researcher would need to test hypotheses. A warning, though - it takes a *really* long time to run. We're lucky enough to have access to an institutional computing cluster, and even there it takes about six days from start to finish. Maybe things would be better if Nick were better at parallelizing his code. Not to worry, though - you can still use our work, even if you don't have a week to spare. The **All-Indicator** script kicks out a large number of objects and datafiles, stored in the [*Server Outputs*][1] folder which are the things that we use in our analyses. Those analyses are all run by the **Follow-Up Analyses** script, which only takes a few minutes to run. Much easier to work with. We have both the preregistered version here, as well as the (slightly-modified) version that we used for the actual paper. The changelog tracking the differences between the two versions is [here][3] After the paper went through review, our reviewers asked for an additional set of analyses. The **Reviewer Requested** script takes care of all that. And, if you want, we have a **How-To** script for running your own versions of this sort of analysis. This one's better commented, and should provide you with a decent baseline to follow if you too want to test the incremental predictive validity of your own implicit attitude tests. ---------- The registration for this project can be found [here][2] [1]: https://osf.io/pzmhr/ [2]: https://osf.io/vxu4r/ [3]: https://osf.io/zpuf4/