Ok so the correlation coefficient for vote share is .048, meaning the R squared or explained variance is .0028 i.e. around three percent. No, monkeys cannot predict vote share.
That said, for a paper with such an obvious bullshit design the text itself is not that bad, at least they seriously investigate the facial features that seem to be relevant to these cases.
If you must remember this paper (please don’t), do it as “monkey ganze is a better predictor for election races in the US than a coin throw”.
The clickbaity title is misleading. The article does not argue that monkeys have a strong election prediction ability. It merely argues that there is some gentle correlation between monkey gaze behavior and election outcomes. That’s not surprising. Almost any reasonable hypothesized factor—number of letters in candidate’s name, for example—will have some correlation. The title and tone of the article are overblown.
This highlights one major problem with democracy - it's usually the loudest/most charismatic/tallest/most publicised person that will win elections. Unfortunately, our system also selects for the best liars - people who are adept at fooling people.
This article is not peer reviewed. It says monkeys look less at masculine features than feminine ones and thus they often look more at the loser than winner of a US election. The monkeys always fail at predicting female
winners vs male despite voters selecting female candidates about half the time. I think the title of both paper and post are funny but wrong. A monkey predicts election outcomes as well as a coin with “the manly one wins” written on both sides (not well).
Let's look at the traits of "successful" politicans; so (like everything not at the top of the food chain) monkeys respond to loud, in your face, potentially threatening stimulus, or alternatively monkeys are not color blind.
That said, for a paper with such an obvious bullshit design the text itself is not that bad, at least they seriously investigate the facial features that seem to be relevant to these cases.
If you must remember this paper (please don’t), do it as “monkey ganze is a better predictor for election races in the US than a coin throw”.