Hypotheses derived from someone's observations: Group X is less intelligent than Group Y. It's trivial to demonstrate how poor our observations are generally. (Though often, in rarefied circumstances, based on our expertise, we probably can have useful obsertations). - ThreadSky

rkapitany.bsky.social • 6 days ago

Hypotheses derived from someone's observations: Group X is less intelligent than Group Y.

It's trivial to demonstrate how poor our observations are generally. (Though often, in rarefied circumstances, based on our expertise, we probably can have useful obsertations).

Comments

rkapitany.bsky.social•6 days ago

Expertise and deep knowledge is something students definitionally lack. Any hypothesis on ritual, for example, is something I could give a pretty good and hopefully accurate intuition on, even in the absence of a direct literature on the topic.

rkapitany.bsky.social•6 days ago

But a student who has generated an 'hypothesis' without a decade of wide and deep reading is... just testing their own biases (implicit, explicit, cognitive, social, whatever). It's epistemically less valuable and valid than more grounded hypotheses based in transparent and rigorous documentation.

rkapitany.bsky.social•6 days ago

That doesn't necessarily mean the observation itself is invalid, tho. It's probably just muddled.

I've had supervision students who are attracted to my research because they have personal religious convictions.

rkapitany.bsky.social•6 days ago

This lead to a project where the 'hypothesis' was: Compared to non-Muslims, adolescent Muslims have better grades in school (because their faith encourages frequent textual study). Observation. Sure. Obviously wrong though.

And wrong because the observation was biased.

rkapitany.bsky.social•6 days ago

Rather than building up from literature (say, "does frequency of reading generalize to understanding across domains", or something similar), they assumed that *Islamic practices* were core to the hypothesis, because it was core to their observation.

rkapitany.bsky.social•6 days ago

They proposed a causal relationship (Islamic Doctrine -> better students). If their observation had been, something like, "I've noticed Islamic students outperform their non-Islamic peers on standardized tests"... such an observation, especially without an implied causality, is open for exploration.

matti.vuorre.com•6 days ago

Indeed. We've an issue with what counts as an observation / lack of proper observational research. It's not sufficient for one to 'see' that "people sometimes R if S" and then test it with empirical methods. One must start with empirical observation and test it within the same framework.

Comments

Posting Rules

Reply