discuss restructuring of autora variables

[musslick]

• how to treat IVs that are high-level (n-arms) but not relevant as input for the theorist
• how to turn high-level IVs into low-level IVs (experimental design)
• how to treat observations (e.g., participant age) which may be inputs for the theorist (and treated as IVs)
• high-level IVs: should they get their own data table since they are the only ones used in the experimentalist

[musslick]

Consider experimentalist function that needs to identify IVs based on predictions from theorist based on different IVs