"We (should) only care about discovering the truth; we shouldn't care ahead of time which truth it is... if we want something to be true and not something else then we pay attention to evidence which favours that theory and discount evidence that doesn't" @seanmcarroll
🧵👇
🧵👇
Describes a huge problem in sports science/sports medicine that is endemic in the social sciences.
There should be no 'proposing a theory'.
There should be no 'proposing a theory'.
(For one thing scientific method states that we can only begin with a hypothesis or an observation.
It only attains theory status once it has been validated with experimentation and repeatedly shown to predict what happens in reality.)
It only attains theory status once it has been validated with experimentation and repeatedly shown to predict what happens in reality.)
(PLUS no theory is ever fully proven - new knowledge can always come to light that disproves it.
No theory or model is sacrosanct.
Nothing and nobody should be above being questioned.
We must always leave room for about and remain open to alternatives)
No theory or model is sacrosanct.
Nothing and nobody should be above being questioned.
We must always leave room for about and remain open to alternatives)
But the more immediate problem is having a pet 'theory' or something we have a stake in being true inevitably leads a tendency to rig things to create findings that support it.
Plus we will tend to spin the results in the most positive way and overstate the implications.
If we can fit any data to our theory we have a problem (see Karl Popper on the difference in mindset between physicists and psychologists).
If we can fit any data to our theory we have a problem (see Karl Popper on the difference in mindset between physicists and psychologists).
(As things stand, we need to look closely at the methods section and examine the data rather than paying too much attention to the conclusions.
None of which we get from just looking at the abstract)
None of which we get from just looking at the abstract)
We also need to be open to being surprised and stumbling on discoveries we didn't expect.
Having a bias in what we want to see going in can make us blind to actual discovery and downplay real findings.
Having a bias in what we want to see going in can make us blind to actual discovery and downplay real findings.
Beyond biasing how investigations are conducted and written up, being too attached to a theory or position on a topic leads to confirmation bias.
It is human nature to cherry pick findings that support our position and dismiss or discredit data that do not fit our narrative
It is human nature to cherry pick findings that support our position and dismiss or discredit data that do not fit our narrative
Science is meant to be about coming with models and frameworks that fit the data and what we observe in practice - not the other way around.
Perhaps time to reinstate the science in sport science..?
END.
Perhaps time to reinstate the science in sport science..?
END.
Post script:
I encourage everybody to listen to the excellent recent @sci_history podcast episode with @seanmcarroll that inspired this thread.
I encourage everybody to listen to the excellent recent @sci_history podcast episode with @seanmcarroll that inspired this thread.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
