Here is the website that hosts the linear regression tutorial. This tutorial has 2 conditions, and plenty of problems.

It is a beginning prototype for how we might teach linear regression using the user's own input and their own data. It i a guided tour through your own personally provided data with prompts to think about and demonstrations of gradient descent.

This includes an input table in JS that the user will input their own information. Nothing is automatically curated, and the current problem involves subjective Likert-scale judgments on their own "interest level" in a class they are taking. In addition to interest level, they report the grade they received in that class. These can be plotted on a 2D scatter plot and used to teach linear regression. The data is input into the table, and that becomes the data for the remainder of the tutorial.

This condition does not include an input table, and does not use language that speaks about the user as an agent of this tutorial. Instead, a hypothetical student is introduced, with hypothetical data, and that data is used for the tutorial. Everything else is the same between Impersonal and Personal conditions.

There is also a version where users can look at their scatter plot of Facebook Likes to Comments on their own profiles, collecting manually for now.

There is also a condition where a list of heuristic facts are listed simply, in a PSA kind of format for the user. The purpose of that condition is bare-minimum control.

Following the tutorial are critiques of models in the world, in addition to several measures on how generative and thoughtful the user is after a given condition intervention.