Skip Navigation

Data is Beautiful @mander.xyz Blaze @reddthat.com 3 days ago

Parent/Child Height Relationships - Regression toward the Mean

Source of data: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/T0HSJ1

Edit: removed OC as it's not (sorry)

17 comments

Pepsi lookin' ass chart
I don't really understand how to read this, so only 96% of 5' 4" boys are taller than their 5' 2" father? What?
- 96% of males with a 5'4" mother and a 5'2" father are taller than 5'2"
- Y=mothers height. X=Fathers height
  
  %=considering the height of both parents what % of girl/boy are taller than their mother/father
  
  Ah, got it got it, sorry for my confusion lol.
How about daughters taller than father and sons taller than mother?
I might miss the point, but the height is dependent on both parents genetically, so just comparing mothers with daughters is a bit like the usual "correlation does not equal causation" thingie, or not?
- The X and Y are just labeled weird, both graphs reference father's height has the X and mother's height as the Y
  
  Yeah, but there is no graph comparing son vs mother and daughter vs father.
  
  And it seems like an odd thing to omit.
  
  Oh, completely missed that, thanks!
- That's exactly what this is showing. The x axis is the fathers height and the y-axis the mothers height so you see daughters change of being taller going up when their dads are bigger. For sons the chance of being bigger than their father goes up with tall mothers.
Wouldn't the most determining factor here be the height of the chosen partner?
.dta

~~Are they series?~~ Edit: are they serious?

You've viewed 17 comments.