Showing all posts tagged academia:

Rational choice strikes back?

Behavioral Economists in Trouble

Dan Ariely and Francesca Gino (and most likely many others) are in trouble, the kind of trouble that involves data manipulation (by the looks of it) and paper retractions. Data Colada has all the details, but the story was also told by an Israeli investigative TV show, transcribed and translated here.

Episodes of Data Manipulation

From a standard rational choice perspective, episodes like these are only to be expected, given that:
  • Opportunities exist (e.g. adding or dropping observations in the raw data file from an anonymous websurvey or experimental platform)
  • The stakes (fame, fortune and/or positions at attractive universities) are high
  • The probability of getting caught is low (or has been, at least)
These facts add to an already significant pile of spectacular results in behavioral economics and social psychology that do not replicate. Sometimes the results were genuinely shaky (e.g. power posing), and sometimes the findings were simply made up, as in the case of Diederik Stapel (wikipedia, NY-Times). See further:

The Standard in Behavioral Economics and Social Psychology

The standard in behavioral economics and social psychology has long been that good papers are empirical articles that show that humans behave strangely and irrationally in ways that rational choice theory and neoclassical economics cannot account for. Mainstream economics' mathematical models were frequently criticized as having little to do with reality. The irony is that behavioral economists respond to incentives in a predictable rational manner – and some of their papers fail to describe the real world.

Perhaps we are witnessing the retaliation of neoclassical economics?


Update aug 17:

Gino is suing Data Colada. This is not the scientific exchange we wanted. Donate to their legal defense here.


Update


Wanted: Google scholar embedding

Citations are increasingly being counted in academia, and it is becoming increasingly common to put for example google scholar citation counts on your CV. The weird thing is that to get something like this on a webpage...
...I had to copy my citation count from my google scholar page , which means that it is probably (hopefully...) out of date when you read this.
A little research revealed that apparently it can be done using "R, scholar, ggplot2 and cron", but surely someone at google could let users do it simply by pasting some code the way it is done with youtube videos and whatever.

The fact that this yet does not exist is a bit worrying, but much more worrying is the fact that I seem to be almost the only one looking for an easy way to "embed google scholar citations" - googling this results in exactly one hit [oct 4, 2015].