Earlier this year, I made a video
about whether or not you should change
your answers during exams if you're unsure on them.
Now the prevailing wisdom has always been
to go with your gut
but the research we looked at
presented this as what they call
the First-Instinct Fallacy
and showed that more often than not
changing an answer you're unsure of
is more often the better choice.
Now the amazing thing about science
is that at any point in time
new data can come and make you wrong
and that is exactly what has happened to me
because now my initial recommendation for you
And due to that wonderful quality of science,
I now have a new and improved method
to suggest to you
and since I used an ambiguous title for this video,
I'm just going to lay it straight out for you
so I don't waste your time.
Next time you're taking an exam
whether it's your final exams
if you're watching this video near its publish date
or any exam in the future,
right after you answer each question,
you should rate your confidence in that answer
using a scale of one to five.
One being super unconfident I'm not really sure
and five being I definitely know the answer
to this question.
Now this technique takes advantage of something
which is essentially thinking about thinking.
It's our ability to analyze our own beliefs
To make an analogy,
it illustrates how the brain is a lens
that sees its own flaws.
One of the core aspects of metacognition
is our ability to judge our confidence in our knowledge.
We can feel uncertain
that is we know when we do not know.
I first started learning about metacognition
a couple of months ago
when I talked to a guy named Justin Couchman
who's a professor of psychology
at Albright College in Pennsylvania.
And Couchman's first forays
into the study of metacognition
took the form of research on Rhesus monkeys.
No, not Reese's monkeys but yeah you know.
In a study he helped conduct called
and bear with me because this is a mouthful,
the highs and lows of theoretical interpretation
in animal-metacognition research.
He and two other psychology professors
set up to see if animals have the same
metacognitive capabilities that we have.
In their study, the monkeys were given
questions of varying levels of difficulty
and they had to either give an answer
or indicate that they didn't know the answer.
And Couchman and the other professors
were surprised to find out that the monkeys
were able to accurately judge their confidence
and indicate when they didn't know.
They were able to look inside their brains
and analyze their own thinking much like we do.
With the results of this study in hand,
Couchman started thinking about
the metacognitive abilities of his own students
who were often surprised that the grades on their exams
were often much higher or lower
than they initially predicted.
And it turns out there is a reason for this.
The problem with metacognition
is that it isn't perfect.
I use the analogy of a lens that sees its own flaws
for a reason.
Our brains are subject to all kinds of bugs,
cognitive biases, heuristics,
flawed modes of thinking
like to quote the AI researcher,
"The brain is a flawed lens through which to see reality.
"This is true of both mouse brains and human brains
"but a human brain is a flawed lens
"that can understand its own flaws-
"its systematic errors, its biases-
"and apply second-order corrections to them."
One of the biggest flaws is that
our memories are notoriously unreliable
and as a result our metacognitive capabilities
decay as we try to use them
to analyze thoughts that we had
further and further in the past.
So to learn more about this,
Couchman decided to conduct another study.
This time on humans rather than monkeys
and they set up two different tests.
In the first tests,
they had students take a real life multiple choice exam
but after each question,
they asked them to rate their confidence
on a binary scale.
Either writing G or K for Guess or Known.
Additionally the asked the participants
to indicate whether or not
they had revised each answer
after giving an initial answer.
And for this first test,
they found that revisions were more
often than not, correct
especially for the answers that were marked
Guess rather than Known.
After this first test,
they decided to conduct a second test
and the only difference here
was that instead of using a binary rating system,
Guess or Known,
they decided to have the students
rate their confidence on each question
using a 1-5 Scale.
One being super unconfident
and five being almost certain.
And the results here were interesting
because in this case
the initial answers were more often correct
than the revisions.
Now these results would seem contradictory
and that would be the case
if there were only one rule of thumb,
either stick with your initial answer
go with your gut
or as the previous research showed,
revisions are better.
But using confidence tracking,
there's no longer just one rule of thumb.
When you assign a confidence score
to each question in the moment
that you answer it,
you're utilizing your brain's metacognitive abilities
at their most accurate point
giving your brain a more sensitive tool
with which to make a decision.
So to wrap up here,
there were two key findings to that study.
beliefs formed about the exams
after the exams were over
were very very poor predictors of actual performance
and this shouldn't come as a surprise.
I remember myself going into many exams
as a student super confident
that I was going to do awesome
and then i got a bunch of questions wrong
or on the other hand,
being really worried that I wasn't ready
and then just absolutely acing it
and I'm sure you've had these experiences as well
but more importantly,
metacognitive tracking in the moment
was a much more accurate predictor of success
and gave the students a more accurate tool
for judging whether or not
they should make a revision to an answer.
So there you have it.
When you go into your next final exam
or any exam in the future,
try assigning a confidence score
next to each answer
as you answer that question.
Doing so will help you to more intelligently
make revisions and hopefully get better grades.
Now if you're curious and you'd like to know more
about the research that went into this video,
I did an hour long podcast with Justin Couchman
where we went more in depth in his study
and you can click the card right now
wherever it is or down below to listen to it
and I've also linked all the relevant studies
in the blog post for this video.
That's all I've got for you this week
so if you're watching this
and you've got a final coming up pretty soon,
best of luck on it
and hopefully this technique will help you out
and in next week's video,
we'll be talking about an often requested topic,
how to make a study schedule for finals.
So look forward to that
and as always, thanks for watching.
Hey guys, thanks so much for watching this video.
If you enjoyed it,
giving it a like can definitely help this channel out
and also if you want to get
more study tips every single week,
you can hit that big red subscribe button right there.
In addition, I also wrote a book
on how to get better grades and I made three
so if you'd like a copy of it
click the picture right there.
If you want to read the summary
and get links to all those studies,
hit the orange button to go to the blog post
and if you missed last week's video,
we talked about 15 writing apps
that can help you become a better writer.
Lastly, if you want to connect
I'm @TomFrankly on Instagram and Twitter
or you can leave a comment down below.