Date: August 25th, 2026 12:02 AM Author: startling pit
If you design experiments correctly and do the math right everything always comes out like trash. you can't "prove" anything unless you basically hit the lottery. No way 90% of these "experiments" are designed properly and. have measurment validity and resolution and reproducibility. it is always way too clean. noone would be able to publish anythihng. they are just going fishing and making a series of defensible looking "choices" until they get an interesting result.
Date: August 25th, 2026 12:28 AM Author: anal excitant boltzmann
i don't think it's limited to social sciences. 'scientists' can't even decide whether statins and vitamin c are good or bad for you, or whether black holes exist/do not exist.
Date: August 25th, 2026 12:34 AM Author: Haunting Blood Rage Milk
the social sciences are the reason you obey authority in this country. are entire legal apparatus is constructed on social science from the 80s supplemented by Obama-era enrichment.
Date: August 25th, 2026 12:42 AM Author: startling pit
You changed it first. The unemployment rate is an official statistical estimate, not the sort of hypothesis driven experimental result I was talking about in the OP.
Date: August 25th, 2026 1:04 AM Author: startling pit
I acknowledge that unemployment statistics are part of the social sciences. It's not a big deal. I don't really disagree with you. I don't think social science is totally useless. I just think a lot of the research isn't reproducible.
Date: August 25th, 2026 1:14 AM Author: Haunting Blood Rage Milk
within the social sciences there are people who quantitative methodologies and people who use qualitative methodologies. You probably heard about the qualitative methodology bullshit and didn't realize the quantitative side exists. You probably also didn't realize that they do statistics better than researchers in biology and medicine
Date: August 25th, 2026 1:29 AM Author: startling pit
I am talking about quantitative research. That's what reproducibility problems apply to. I agree econ has strong quantitative methodology, idk that you can generalize that to sociology and psychology. I think econ researchers generally know more math than bio and medical researchers, but the research generally isn't more empirically valid. Bio research is generally grounded in physical reality. Econ you can have sophisticated math executed perfectly floating on weak assumptions.
Date: August 25th, 2026 10:51 AM Author: Haunting Blood Rage Milk
all 28k studies at that ICPSR link are reproducible. I dare you to find one that isn't. I get that you're a crank who worships Karlstack, but you're obviously a lightweight with no idea what the fuck you're talking about. No one takes you seriously and you're obviously not a functioning lawyer.
Date: August 25th, 2026 12:07 PM Author: startling pit
I don't even know who "karlstack" is and I couldn't care less about the "ICPSR". However, I appreciate your passion, and I sincerely hope you find an effective treatment plan for whatever you are currently experiencing.
Date: August 25th, 2026 11:00 AM Author: Richard Ames
Basically the only thing from the social sciences that has proven to be replicable and have meaningful predictive value is IQ testing.
The issue with "social sciences" is that they're filled with shitlibs who do anything they can to "prove" that IQ isn't the innate driving force that it is. Hence nothing they do is replicable.
Date: August 25th, 2026 8:34 PM Author: arrogant burgundy death wish
I'm a geologist and can attest to the fact that geology is an "investigative" scienc. Most of it cannot be reproduced, it only has patterns. It's fake and gay.
no one ITT has linked to an irreproducibile social science study. It's weird because I provided 28,000 social science studies for poasters to select from. How hard is it to find ONE?