The benefits of smart friends: Among 715 adolescents, their middle school best friend's IQ predicted increases in their own intelligence 3 years later (controlling for adolescents' prior own IQ) Fascinating work by @FSNole57! https://psyarxiv.com/t7pdx/
-
precise set of steps you could go through to correct for this, but intuitively it seems like it should be possible. Seems like it would be a question for Andrew Gelman or someone who specializes in statistical methods.
A simple approach here is to adjust the correlation matrix for measurement error and then use SEM on the corrected correlation matrix.
