@KMKing_Psych @JoshuaDFoster1 @LisaDeBruine @sTeamTraen @jamesheathers @geoffreylitt @MarcusCrede I also have a bunch of R code up on guthub to do these methods on common types of data. https://t.co/chinSn1ym8
@KMKing_Psych @JoshuaDFoster1 @LisaDeBruine @sTeamTraen @jamesheathers @geoffreylitt @MarcusCrede That I did. In answer to the start of this thread, I also have another piece from 2016 that outlines a bunch of psychometric ways to do this. https://t.co/Qj1MS23nU7
@BananasPls Personally, I'd probably simulate some data to give myself a concrete example, and then work through it with the capability of checking the general case derivations against actual specific case concrete values.
@BananasPls If you were able to find the mean of y in the first problem you could translate those x values to a predicted y, I suppose, and see how close to the mean it is, perhaps in conjunction with those residuals to get at the variance of y values relative to prediction.
@BananasPls Well, you have the regression line, and the bounds on the residuals, so if there was a way I would think that might get where to start. If you can use that to get to the variance of your prediction about the line that might allow you to get at the variance in y?