Does anyone by chance have a reference for the process of theory testing and development? Maybe @devezer.bsky.social ? I'm trying to argue that a basic tenet of theory testing is that it is incremental (ie isolates one feature to test at a time) #AcademicSky #PsychSciSky - ThreadSky

raechel.bsky.social • 104 days ago

Does anyone by chance have a reference for the process of theory testing and development? Maybe @devezer.bsky.social ? I'm trying to argue that a basic tenet of theory testing is that it is incremental (ie isolates one feature to test at a time) #AcademicSky #PsychSciSky

Comments

devezer.bsky.social•104 days ago

I cannot think of a single process or reference. The closest that comes to mind is our recent special issue but I doubt that it will give any specific answers. I'll repost in case others have better suited references.

raechel.bsky.social•104 days ago

Thanks, Berna!

smellosopher.bsky.social•104 days ago

Hm. There's sufficient reason not to think this is how theories or frameworks have been worked with historically. Partly bc theories are not coherent wholes, partly bc of the underdetermination of scientific theories by empirical evidence (quite a mouthful):
https://plato.stanford.edu/entries/scientific-underdetermination/#HolUndChaSciRat

raechel.bsky.social•104 days ago

Thank you, that is really helpful

iflis.bsky.social•102 days ago

Your question made me think of this paper and their idea of the "theoretical cycle" (inspired and build to contrast the so called " empirical cycle"). @irisvanrooij.bsky.social https://journals.sagepub.com/doi/full/10.1177/1745691620970604

boryslaw.bsky.social•102 days ago

If I may, it is, I think, hard to see how this could not be an empirical claim (about how people actually try to test theories in practice), although the (puzzling!) word "tenet" suggests a normative character. As an empirical claim, it requires data that do not seem to be available...

boryslaw.bsky.social•102 days ago

That being said, an empirical test of a theory is always a kind of inference with a substantial number of premises. In that sense, it cannot be a test of an isolated property...

boryslaw.bsky.social•102 days ago

On the other hand, predictions are tested one at a time; in this (trivial!) sense, the test is concerned with an isolated property. Often, for the purpose of testing one premise/assumption is viewed as more problematic than every other premise, but this may be a property of the *researcher*.

boryslaw.bsky.social•102 days ago

In psychology, for instance, every process model that I can think of seems to consist of a number of more or less equally problematic assumptions.

Comments

Posting Rules

Reply