Sad Irony: Author of book on morality accused of scientific misconduct. What can this teach us?
Marc Hauser, evolutionary bioligist at Harvard has recently been found guilty of scientific misconduct. Ironically, he is the author of Moral Minds: How Nature Designed Our Universal Sense of Right and Wrong. A blog at nature.com says:
In the 2002 Cognition paper, Hauser and his colleagues reportedly trained cotton-top tamarins to recognize two different “grammars”. These grammars were patterns in the sequence of syllables, for instance “wi wi di” (AAB) vs. “le we we” (ABB). One group of monkeys was trained on the first pattern, and the other group trained on the second pattern.
The investigators then played these sounds on a hidden loudspeaker, and watched the monkeys to see if they turned to look in the direction of the sound more often when they heard a different “grammar” than the one to which they were accustomed. Hauser and his colleagues claimed that they did, suggesting that the monkeys were able to distinguish between two underlying grammars. But Altmann says that, according to the Harvard investigation, Hauser lacked the critical control data showing how often the monkeys turned toward the loudspeakers when hearing their familiar grammar. “Perhaps they would turn round as often if they heard anything coming from that speaker,” Altmann wrote. “The experiment as run did not allow any conclusions to be drawn regarding monkeys’ ability to distinguish between different grammatical patterns.”
As someone who has worked in a lab doing meticulous video tape coding myself, this is pretty crazy. I think they’re suggesting that Hauser didn’t code the videos at all, and that he just had the outcomes of the data without the actual data. I have no idea if this is true or not, but it highlights for me how important it is for employees to understand the connection of what they do to the bigger picture and, for my research-oriented readers, the importance of looking at operationalization. As a coder, you sit there staring at videos probably for hours straight. In this case, it sounds like you’d be watching for the chimps to look toward the speaker. You’d probably be recording the specific time that they looked over, how long they looked for maybe… I’m not sure what else. There would probably be several coders to make sure everyone is agreeing. But it’s these little details that everything else is built on. This is the operationalization of the hypothesis. If these details are wrong, then other research in the future that’s built on this is going to be flawed and probably useless. As a coder, you’ve really got to understand this, and you have to be aware of this as your blood shot eyes are trying to focus. Maybe you zoned out while you were watching. You have to rewind and watch it again.
Anyway, maybe Marc just lost the data, but it is a good reminder of the connections from the little to the large.
