Saturday, August 03, 2013

Positivity ratio of 2.9 debunked

I often denounce the idea that psychologists can say much useful about the best interest of the child (BIOTCh), and that even the top universities are overrun with quacks. Here is a good example.

Neuroskeptic blog writes:
British psychology student Nick Brown and two co-authors have just published an astonishing demolition of a top-ranked paper in the field of positive psychology: The Complex Dynamics of Wishful Thinking ...

The ‘critical positivity ratio’ is a popular idea. Fredrickson and Losada’s 2005 paper on it has been cited a massive 964 times on Google Scholar, just for starters.

And yet – that paper is complete rubbish. As are Losada’s previous papers on the issue. I criticize a lot of papers mysef, but this one really takes the biscuit. It’s an open and shut case.

As Brown et al write, the idea of a single ‘critical ratio’ that determines success or failure everywhere and for everyone is absurd in itself:
The idea that any aspect of human behavior or experience should be universally and reproducibly constant to five significant digits would, if proven, constitute a unique moment in the history of the social sciences.
But even were there a magic ratio, it wouldn’t be 2.9013. The whole analysis in the 2005 paper was based on taking a poorly-described dataset and then making it fit a mathematical model, purely by means of elementary misunderstandings.
I am not aware of this research being used for child custody decisions, but it could be. The parents and kids could be assessed to see if their ratio is above or below 2.9. It could be as valid as the inkblots I had to take.

(In fairness to the paper, it uses 2.9 and not 2.9013.)

Such a procedure might be an improvement over what psychologists like Ken Perlmutter, Faren Akins, and Bret Johnson use. They ignore the published research, and make recommendations directly contrary to established knowledge.

1 comment:

Anonymous said...

On the topic of shrinks, this bit deserves an honorable mention: