#experimental bias

[thoughtsfromunderarock]

The power of good looks

I have no doubt that being attractive is a beneficial characteristic in our extremely materialistic society.  However, this article at fast company describing an experiment where an attractive woman changed her appearance to see how her looks affected her treatment by strangers, does not really make the point well. The problem, it is hard to tease looks apart from all of the other social cues we are constantly bathed in.  Note that while experiment was presented as an exploration of how appearance ( i.e. looks) affect social interactions, the experiment was described as one where the woman changed here appearance by

dressing down, changing her posture, and adopting a surly, scowling demeanor. 

Of course, this means much more than simply looks was being tested.  Behavior and attitude where also changed during this experiment.  

I suspect that self confidence ( which may covary with looks) is also important for getting ahead in this world.  

The Facebook experiment is more interesting but must come with the caution that the type of interaction involved is quite superficial, and may not extend to more sustained interactions.  

[madlori]

Science Term of the Day

CLEVER HANS EFFECT

In the early 1900s, there was this German guy who had a horse named Hans.  He had trained Hans to add, subtract, read, identify music, and do all kinds of amazing cognitive tricks.  It was a big sensation.  Big enough that the German government conveyed a panel to determine if there was fraud.  They decided that there wasn't.

But then a psychologist figured out that Hans was trained all right: to watch his trainer.  He answered questions by tapping his hoof, and the trainer gave off totally unconscious signals with his body and face that signaled Hans when to stop tapping.  The psychologist was amazed to discover that he himself was giving off the same signals when working with the horse, even when he knew he was doing it and was trying not to.

This effect is very important for research.  Unconscious signals from experimenters can vastly affect the outcome of studies.  This is largely overcome by double-blind studies, but those aren't possible in cognitive or psychological research involving animals.  The Clever Hans effect must be taken into account with, say, sign-language-using apes or animals who do other seemingly-amazing things.

The Clever Hans Effect is part of a larger aspect of research methodology known as experimenter effect, wherein the expectations of the experimenter bias the results.