3 ms·

Oh boy. Very trustworthy "science" right here. There are so many problems with this... Study 5+7: online survey via Mechanical Turk (n=90/108) really? Study

Oh boy. Very trustworthy "science" right here. There are so many problems with this... Study 5+7: online survey via Mechanical Turk (n=90/108) really? Study 6: online survey via Craigslist ad, voucher reward Study 3+4: survey with undergrad students (n=120/150) Studies 1+2: rating behaviour of about 200 drivers at an intersection * Not even a definition for social class or ethical behaviour. * "Social class" was assigned via self-ranking or car value, both very problematic. * Almost no numbers or info on methodology. * Online surveys with self ranking for social class, and via paid Mechanical Turk participants or Craigslist ad with vouchers are sure to give a very accurate and representative sample with "higher class" individuals... * In 1+2, very low n considering how easy data collection would be. Driver age and gender were ostensibly controlled for, but I'm doubtful this was done properly. The hypothesis might very well be true, but this "research" is about as valuable as a tabloid opinion piece.

alexgmcm 7y ago
Yeah, I'm sure it'll get plenty of citations though and in modern science that's the figure of merit.
- cryptica 7y ago
  The irony of it is that when you consider the reality of increasing wealth inequality (and the power imbalances that this creates), the claims made by the article are consistent with the ethical degradation of modern science and business in general.
thrwaway69 7y ago
I don't even bother reading papers that are politically charged these days if any at all. There is no end to this kind of papers if you check and I am really tired of seeing the media spin up of these, too. Is exaggerating a claim to the point it makes it false or not verifying the papers before promoting "fake information"? Here's why you need to stop doing x. It's bad for you, research confirms. Why research thinks the type of food you eat has influenced your election candidate Why we are going to be extinct in the next 50 years and scientists can't figure it out. This new material can change the way how processing works in a matter of few years, research shows (date - 2005) Most AI researchers claim this and that (survey cut down to a single event with undergraduates and journalists mixed in.) I would be curious if someone has any papers confirming that the number of shit papers has increased over time or has remained the same whilst being scrutinized more heavily.
puranjay 7y ago
I've made it a point to look up any study mentioned in the media. In almost all cases, dramatic conclusions are drawn from sample sizes in the low xx digits. There is absolutely no way such tiny sample sizes can be conclusive

alexgmcm 7y ago

Yeah, I'm sure it'll get plenty of citations though and in modern science that's the figure of merit.

cryptica 7y ago
The irony of it is that when you consider the reality of increasing wealth inequality (and the power imbalances that this creates), the claims made by the article are consistent with the ethical degradation of modern science and business in general.

thrwaway69 7y ago

I don't even bother reading papers that are politically charged these days if any at all. There is no end to this kind of papers if you check and I am really tired of seeing the media spin up of these, too. Is exaggerating a claim to the point it makes it false or not verifying the papers before promoting "fake information"? Here's why you need to stop doing x. It's bad for you, research confirms. Why research thinks the type of food you eat has influenced your election candidate Why we are going to be extinct in the next 50 years and scientists can't figure it out. This new material can change the way how processing works in a matter of few years, research shows (date - 2005) Most AI researchers claim this and that (survey cut down to a single event with undergraduates and journalists mixed in.) I would be curious if someone has any papers confirming that the number of shit papers has increased over time or has remained the same whilst being scrutinized more heavily.

puranjay 7y ago

I've made it a point to look up any study mentioned in the media. In almost all cases, dramatic conclusions are drawn from sample sizes in the low xx digits. There is absolutely no way such tiny sample sizes can be conclusive