I had such high regard for academia while I was in school. Glad I didn’t end up going down that path, it seems like every day recently there have been stories like this posted to Hacker News. Behavioral economics especially seems hard hit.
In my experience, this is absolutely not true. I mostly only have first-hand observations in the economics, finance, and marketing disciplines. They left me EXTREMELY cynical about all academics.
I have seen:
1) A well-known professor at an Ivy-league B-school openly talking to his class about how the NAR funded his latest study on the real estate market, which (of course) found that the market was not overvalued. This was 2007. The co-author on the paper was the dean of the B-school. He is no longer the dean, but only because he seems to have moved on to more lucrative consulting roles.
2) An older professor, who was one of the first to be in the group labeled "behavioral economics," when asked why he wasn't writing a book focused on more mainstream readers like some other such big names were at the time, STRONGLY hinting if you filled in the blanks although not outright saying that their ethics were questionable, they were all forming hypotheses that were useful for for consulting dollars and dredging up data to support it, and he was too old school (and too old) to be interested in playing that game.
3) A full cohort of marketing grad students (6-7) talking openly about how the system is messed up, and to get a job, you absolutely have to have a good dissertation with a strong results, but to avoid p-hacking allegations, you have to shoot in the dark. So you're left in this situation where you commit financial and career suicide if you DON'T make up data to support your thesis. It is not rare. It is an open secret. No one likes it. They hate it. But they also feel like they are the stupid idiot if they try to fight against it alone. It is how the system works. Only a true saint would accept data that actually does NOT support their hypothesis because it makes it unlikely they will get a job in a very competitive market. Some people do report non-findings, but they are rare. IMO, academic departments should specifically seek out graduates who were willing to report that and stuck by it. But that is not the way the world works. As another poster pointed out, we all like stories, and no significance means no good story to tell.
1) Yeah, lots of people missed the crisis.
2) Yes, when funding doesn't come from the government it comes from industry. And then you need to find hypotheses that are interesting for industry.
3) That's how science works! You don't know what is true or false. You need to take your shot, see if the experiment you run works out. Sometimes you can be the smartest person in the world, put in the most work, but be in an area asking a question that doesn't pan out. It happens!
There are many ways to mitigate this. For example, you can often make experiments where it doesn't matter what the answer is. It's always interesting and publishable regardless of how the statistics work out. But, there is definitely luck involved, not just skill.
> Only a true saint would accept data that actually does NOT support their hypothesis because it makes it unlikely they will get a job in a very competitive market.
Only a scientist would! That's the game we play to discover new truths. And the vast majority of scientists do this.
Pretty much every idea that I have ever had is bad! Yet, somehow, I managed to get a PhD and now graduate many of my own PhD students. The trick is to learn to fail quickly, to ask the kinds of questions that are interesting no matter what, and to develop a sense for the types of questions that are more fertile.
Your take on the last point is not so different from mine. I think you misunderstood. The incentive to change data to avoid a finding of no significance is very great. The risks are small. Yes, good science would be to report it accurately. It almost NEVER happens. It is widespread. More so than people think. These two dishonesty researchers are just the tip. A lot of people in academia are staying silent because they don't want people looking into their own research. False data is not the exception. It is endemic.
That's objectively false. Batteries keep getting better. Cancer survival keeps going up. Computers keep getting faster. By almost any metric you pick, research in almost every single discipline is a massive life-changing success! Not just a success, a wild success.
> Your take on the last point is not so different from mine. I think you misunderstood. The incentive to change data to avoid a finding of no significance is very great. The risks are small. Yes, good science would be to report it accurately. It almost NEVER happens.
Nonsense. I've been a scientist for two decades at multiple universities. I've collaborated with hundreds of people. Never has anyone tweaked data on anything we've ever worked with. We usually do the opposite, find that something works and work as hard as possible to trash our own work so we can't publish it. We spend more time trying to disprove our own work before publishing it than getting any positive results.
> It is widespread. More so than people think. These two dishonesty researchers are just the tip. A lot of people in academia are staying silent because they don't want people looking into their own research. False data is not the exception. It is endemic.
That's a conspiracy theory just like aliens landing on the moon and vaccines causing autism. Somehow, hundreds of thousands of people work together to stay silent and not leak. People who frankly hate each other's guts and want nothing more to do than to disprove each other's theories and show they're superior. Please, this doesn't stand up to even the most basic scrutiny or thought.
Every scientist that I know wants to do science. Wants to be remembered for contributing to our understanding of the universe hundreds of years after they die. You don't do this by lying, you do it by doing the best work you possibly can.
You can even do ethical virtual experiments with the data of 3 billion people. Just specify the situation, the person and what's to happen.. Voila. All things happen regularly at some point..
You really need to clarify this.
> The government is trying to control your mind through the internet
Group of people who both have the theoretical training and practical experience is probably pretty small so private/public crossover isn't unreasonable.
I'm still glad I didn't go into academia for many other reasons, but I don't think I dodged an ethics bullet.
Reputation probably matters more for academia (people go into it because they think the research is important, not for the money, and they think they're standing on the shoulder of giants and not amiable bullshitters taking the credit for findings actually created by their overworked research assistants kludging the data), but if you think academia is full of lies and half truths relative to other professions, you haven't encountered many startups!
Just wait for people to dive into the false equivalence of scales and measures used in political research. It's very much "make it up and see if it sticks." The goal appears to be Psychology Today, not scientific rigor. I wrote about this in graduate school, as part of my hefty statistics coursework, it bothered me so much.
I would think instances of fabrication are quite low — it’s high-risk, medium reward. Even if it goes undiscovered, academia is a marketing game, and you can’t control what other people feel is impactful or not. It’s very easy to do bad science without falsifying data, which is probably the deadliest sin in academia.
Yes, which is why I said the frequency of discovery is low. Seeing a story in the paper about fraud every day doesn't really tell us anything about the rate of fraudulence, which I was hoping would become transparent with my point.