Paul G Curran

[Paul G Curran @PGC8]

@KMKing_Psych @JoshuaDFoster1 @LisaDeBruine @sTeamTraen @jamesheathers @geoffreylitt @MarcusCrede Feel free to ask me anything. 👍

[Paul G Curran @PGC8]

@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

[Paul G Curran @PGC8]

@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

[Paul G Curran @PGC8]

Almost time for the @MrTASBot run at #agdq !

[TheFlyingMarlin @TheFlyingMarlin]

The #AGDQ2019 stream room

[Celeste @celeste_game]

We'll have @TGH_srl speedrunning #Celeste later tonight during #AGDQ2019!
Let's support this awesome event! 🍓✨

https://t.co/uCw88gibBl

[THE FREAM ✦ @FREAM]

#DancingDad is the hero we all needed at #AGDQ2019

[Paul G Curran @PGC8]

If you're not watching #AGDQ2019 yet, now is a great time to tune in for Cuphead, then Celeste, then Windwaker! https://t.co/DtIYvfT0JT

[Paul G Curran @PGC8]

#desertbusforhope #teamorder #400400

[Clinton Lexa @halfcoordinated]

Hi there

thank you for being you

[Paul G Curran @PGC8]

@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.

[Paul G Curran @PGC8]

@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.

[Paul G Curran @PGC8]

@BananasPls The predicted value of y at x1=x2=0 is just the y-intercept, alpha, so that would be justified.

[Paul G Curran @PGC8]

@BananasPls If you find an answer I'd love to hear it, but I don't think I'm much more help without really sitting down with pen and paper. Sorry!

[Paul G Curran @PGC8]

@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?

[Paul G Curran @PGC8]

@BananasPls That said, even the variance wouldn't tell you the point the variance is varying about (the mean), so still not sure that would work.

[Paul G Curran @PGC8]

@BananasPls This seems to be the best I can find, though it's a different case than yours: https://t.co/i3BvZvJg7B

[Paul G Curran @PGC8]

@BananasPls Is there a reason you can't just calculate y bar from the original data? Is the regression output all you have?