Some days ago, while going through a Readers Digest, I came across an interesting experiment carried out in 1997 by Daniel Levin and Daniel Simons from Cornell University. This experiment now joins the mini-list of clever, crisp, and elegant social psychological experiments that I would like to remember.
The experiment involved a 'stranger' who stops a pedestrian on a university campus to ask for directions. All fine and good. The pedestrian starts giving directions and the stranger is listening and nodding for good measure. All going well. Some seconds later, two men carrying a door (yes, a door) cut across the pedestrian and the 'stranger'. All is well still inspite of the rather rude intrusion. Or is it? The talking 'stranger' quickly sneaks into the spot of one of the 'door-men', while one of the 'door-men' stays back to talk with the pedestrian. Out of the 15 people they tried this little 'prank of an experiment', 7 of them noticed the change. The 'door-man' does look similar to the 'stranger' - even though in this set of experiments the two were wearing different clothes. Not huge numbers here - but rather amusing nonetheless. I'd found a neat video of the experiment but now I can't find the link any more. It's either been removed within the last couple of days or maybe I'm typing in the wrong string of words.
In a second set of experiments, Levin and Simons made the 'stranger' and the 'door-man' wear construction clothes and construction hats. Here, only 4 out of 12 people (college students) noticed the 'new-stranger/door-man' was now a different person.
While this experiment comes across as being funny and neat and even odd for many people interested in psychological and social-psychological experiments, it seems that some (crabby people?) don't like this experiment at all. They think it was too frivolous or something quite obvious, and that it's silly to think of this experiment as being of any use.....
P.S: One thing I've never quite understood: for experiments involving small numbers, such as forty people (which is probably a fairly average number when doing experiments) or 65 people - why do studies and reviews of these studies report percentages? Just reporting the raw numbers would make much better sense, I'd think.
For instance, if 24 out of 40 people while strolling around a waiting room leap up into the air upon finding a person on T.V casually talking back to them while in the middle of reading the evening news - what's the point in saying:
60% displayed 'leaping behaviour'.
15% exited the premises displaying 'fleeing behaviour'.
12.5 % displayed 'yelping behaviour' .
5% displayed 'peering behaviour' (i.e peering into the monitor, peering into the back of the monitor, and peering all around).
5% displayed 'sideways glancing behaviour'.
and 2.5 % displayed 'perfectly normal behaviour' (given the circumstances) but was still carried off by the men in the white coats....?
Why not say -
Out of the 40 subjects 'tested' -
24 leapt.
6 fled.
5 yelped.
2 peered.
2 leered.
And 1 person was carried off by the men in white coats (after he started chatting pleasantly with the 'newscaster' as if talking with people in T.V sets were something quite normal and something that he did everyday.... The 'newscaster' had to end the conversation before walking off the sets, and refuses to participate in any future psychological experiments...)
The Readers Digest did not use percentages, thankfully enough. They simply reported the numbers for the Levin and Simons study.