If you've never taken the "Selective Attention Test"
try it now (1:22)
(No, I did not get to take the test, because I had already read about it first)
I sent the following email to co-author, Christopher Chabris:
"I have been unable to locate any mention of,
what I would consider to be,
the most important information of all:
1) What was the average score of those who saw the gorilla?
2) What was the average score of those who missed the gorilla?
I was unable to locate the info on the internet.
I read the abstract
but could not find the info I wanted."
Here is the reply from co-author Christopher Chabris:
"There's an analysis like this on page 1069 of the article, but we did it by correlating noticing (1 for noticing, 0 for missing) with the error magnitude (# of passes away from correct count), averaged across conditions. The correlation was tiny. In our more recent paper in I-Perception we found that people who made more counting mistakes were more likely to notice the unexpected event, but only when the counting task was difficult.
Here is the excerpt from page 1069 of the article:
"Across these fifteen conditions the correlations averaged to r . 0:15, suggesting that noticing was not strongly associated with counting poorly or inattentively.
As stated in the abstract, those who noticed the gorilla
did miss more passes on average than those who didn't.
So why were the authors raving about the abilities
of those who noticed the gorilla?
I reached the opposite conclusion:
Viewers were given a task.
Those whose powers of concentration were so strong
that they couldn't be distracted, even by someone in a gorilla suit,
should have been praised for passing the test with flying colors;
while those whose powers of concentration were not as strong,
and got distracted from their assigned task, and therefore missed more passes,
should have been the ones who were viewed as having performed less efficiently.
I formed these opinions about the authors:
1) When the results revealed that those,
who could not be distracted by the gorilla, scored better,
the authors tried to bury the accomplishment of those subjects deep in the abstract.
2) Knowing that they could not omit that fact
without being accused of hiding data,
they briefly mentioned it, but even then,
they tried to minimize the importance of the difference in performance.
Why did the authors take a positive quality:
the ability to focus and successfully complete a task,
and turn it around completely backwards,
and accuse the successful subjects of "inattentional blindness?"
I see at least two possibilities:
1) If they highlighted the results
which showed that people who were distracted by the gorilla
scored lower than those whose powers of concentration were superior,
and who therefore, were able to see more passes;
everyone would have yawned and said "So what; why is that news?"
But, then a light bulb appeared over the researcher's heads
and they realized that if they criticized those who performed better,
and gave it an exotic name, like "inattentional blindness,"
then they could give the impression that those who performed more poorly,
possessed some special gift.
Then, as we all saw, instead of yawn-makers, they became celebrity researchers!
2) It is also possible that the researchers fell prey
to one of the most common weaknesses of all humans,
one which all scientists must constantly be on guard against ... confirmation bias.
The authors started with a conclusion,
and when the results produced a contrary result,
they simply "interpreted" the results
in a way that supported their original hypothesis.
Creationists do that all the time.
They start with the answers and try to force science to conform to their myths.
But we expect more from our scientists.
Here is an interesting study that puts the "Selective Attention Test"
in proper perspective (that would be the one that I proposed above).
This is what can happen when you try to hide data and force interpretations
to produce a result that you need
to be true.
Eventually, scientists with more integrity will begin to appear;
and they might be bringing ... embarrassing results
(click here first, then copy from address bar above)