@jon_mellon This is a mistake even leading political scientists make. I think of the absurd interpretation of https://t.co/QRVEWaAfZA which is compatible with Facebook ads having a higher ROI than any other intervention
And that scroll-by ability has been why I've been skeptical of FB effects. Also @dbroockman had an early study where he found FB ads couldn't even raise candidate name recognition(!) https://t.co/PlzTnnJIjQ https://t.co/4VNA43iq1r
@mcherenson @aecoppock @seth_j_hill @vavreck Though there's some strong evidence that even dozens of exposures don't seem to have an effect. From what I've seen these political DVs are less malleable, so I think your point applies more to specific issues t
@NateSilver538 I am writing my thesis on this very topic. Nate could be correct. This paper showed that online ads didn’t increase name recognition for US political candidates. Consistent with Ridout’s hypothesis that online ads only have influence for <