Transcript for Tom Cech Interview - How does a scientist know that what he or she has found is correct?

Yes, that’s an excellent question. How do we know we’re correct?

Well, of course you repeat an experiment several times...

...to make sure that the result is reproducible.

That turns out to be not enough...

...because there could be an inherent mistake in an experiment and every time you repeat it,

if you repeat it exactly the same way, you make exactly that same mistake...

...and then you get the same result, so...

...although we repeat our experiments, that’s insufficient.

Ultimately what happens is that you try to find a whole different approach to...

...validate what you think is happening.

We used to have a rule in our laboratory that you need to find three independent,

completely different kinds of experiments to test an idea.

Sometimes it’s hard to find three, so we settle for two.

But one usually is not enough.

And then, ultimately what happens is that you publish your...

...information and your explanation in a scientific journal...

...and then, if it is exciting to other people,

they will try to extend your work in directions that are interesting to them,

and that’s when you find out whether your results are correct or not,

because if other people extend them in new directions and can build on them,

then everyone says: “That work was really well done – important work “.

If other people can’t extend it, then they maybe try to just repeat your experiments.

If they can’t repeat it, it doesn’t necessarily mean that you were dishonest...

...or that there was any fraud involved, but if the experiment is not repeatable,

for whatever reason, people just quit thinking about it.

They just push it to the side, so those things...

...may never get corrected in the journal,

but still they don’t become part of the fabric of science...

...and so the process has a wonderful self-correcting ability...

...that those things that are right get amplified and extended in many directions,

...and then that validates them.

Those things that were wrong, or sloppy or perhaps in a few cases even dishonest,

those things are not extendable by independent laboratories...

...so they get pushed sort of under the rug and they don’t do anyone any harm anymore,

because nobody is using them as the basis for thinking about...

...how the human body works, for example.