Wow, there's so much wrong with this "statistical analysis" to the point of absurdity. Goes to show that even a 30-year economics professor can be duped by AI.

▲

jeffrey_t_b 3 hours ago | parent | next [-]

It looks like his results are taken from the Federal Reserve 2022 Survey of Consumer Sentiments. https://www.federalreserve.gov/econres/scfindex.htm The data is publicly available. I don't think that this was hallucinated.

	▲	text0404 3 hours ago \| parent [-]
		Yes, he seems to be using that dataset. Can you tell me how the table of numbers was generated? All it says is that it was "calculated" by ChatGPT. This guy is an econ professor; where's his reproducible methodology? Further, he's treating this 2022 snapshot as a timeseries as if the various categories are tracking the same people over time. Just because 80 year olds in the data are wealthier than the younger brackets doesn't mean that those are the same people (in fact, those 80 year olds are boomers who grew up during a time of unprecedented financial growth and stability in the US, and who currently own a massive amount of housing in the US and are refusing to retire and allow younger people to take their high-paying jobs). I can continue but those are the most egregious issues to begin with and I'm tired of reading this slop.

▲

ebb_earl_co 3 hours ago | parent | prev | next [-]

I was going to say something similar: invoking one’s credentials as a professor only counts for legitimacy if that 30-year legacy of thinking on the topic is applied to the problem under discussion.

In my opinion, if the analysis is offloaded to the plagiarism machine, it loses all epistemological value.

▲

crisnoble 3 hours ago | parent | prev [-]

Source: Chat GPTs calculations.