,
13 tweets,
8 min read
Read on Twitter
@dutch163 @PetraRoesner 1/ Yes, me too. Once and for all: most phenomena cannot be directly measured. We deal with "by proxy" data. "There is a correlation between A and B" does NOT mean that "A causes B" or that "A and B share a common cause".
@dutch163 @PetraRoesner 2/ That is one of the fundamental things we teach in methodology of science classes. We can have amazing data sets and sadly, horrendous publications because we haven't trained the researchers in logic, epistemology and method.
@dutch163 @PetraRoesner 3/ Researchers are not supposed to "believe" anything. They collect and interpret data. With that, they offer a MODEL (hypotheses, theories, etc are MODELS). That model will be subject to empirical testing.
@dutch163 @PetraRoesner 4/ The longer it resists the empirical challenge without being pierced by anomaly holes, forcing proponents to add ad hoc hypotheses to it, the stronger it gets.
@dutch163 @PetraRoesner 5/ However, no matter how long it resists, if the model is based on correlations, it remains a correlational model. Intellectual honesty and methodology strictness require that we NEVER, ever, infer causative statements from it.
@dutch163 @PetraRoesner 6/ Scientists don't "believe" anything. Either they know or they don't. They may suggest a few explanatory models and it is their duty to point out each one's strengths and weaknesses.
@dutch163 @PetraRoesner 7/ So, NO, we don't know why authoritarianism correlates with poor cognitive abilities; we don't know why MRI's from high scoring authoritarians differ from low scoring ones in respect to certain stimuli and areas of the brain;
@dutch163 @PetraRoesner 8/ We don't know why, in industrial societies of European descent (or properly so), about 30% of the population scores high in authoritarianism. Correlation with cultural variables still seem stronger than any other variable...
@dutch163 @PetraRoesner 9/ Because it is long accepted that significant brain changes occur as a response to environmental stimuli, particularly in the first 7 years of child development.
@dutch163 @PetraRoesner 10/ I am disappointed on social psychologists, who should be more knowledgeable in disciplines directly implicated in interpreting these data, such as anthropology, evolutionary biology (and not evolutionary psychology), sociology and History, and propose weak and faulty models.
@dutch163 @PetraRoesner 11/ The ones doing that are no better than folks like Jordan Peterson or Gad Saad, who are known to commit scientific sacrilege, having no respect to method at all. They are also pretty close to eugenics.
@dutch163 @PetraRoesner 12/ There is a reason why psychiatry and neurology researchers about suicide, for example, are going back to the classics and reading Durkheim. A century of correlations did not land in causation. It is frustrating, but honesty requires that we go back to the drawing board.
