Read on Twitter
Bob is a misogynist who subconsciously avoids intelligent women. Bob randomly selects a list of men and women he knows. He uses statistical best practices to analyze their cognitive abilities and concludes women are less intelligent than men.
Is Bob's analysis scientific?
Is Bob's analysis scientific?
Reasons to say Bob's analysis is scientific:
1. Anybody else analyzing the same data will get a similar answer
2. The model is predictive of future data. Bob will continue to avoid intelligent women in the future and so the model will accurately predict his future experiences.
1. Anybody else analyzing the same data will get a similar answer
2. The model is predictive of future data. Bob will continue to avoid intelligent women in the future and so the model will accurately predict his future experiences.
A second misogynist Tom replicates Bob's study using people he knows. Tom confirms Bob's findings.
Additionally, a group of 1000 scientifically-inclined misogynists pool together all of the people they know and a third party data scientist finds a similar result.
Additionally, a group of 1000 scientifically-inclined misogynists pool together all of the people they know and a third party data scientist finds a similar result.
At this point, the finding is seemingly robust and highly replicable.
Should we now be more convinced than ever that Bob's study is objective and scientific?
Should we now be more convinced than ever that Bob's study is objective and scientific?
If you object that this isn't science, what standard scientific norm could you apply to disqualify the finding?
Note that:
- the data (as collected) are accurate
- the models are predictive of future data
- the finding has been replicated multiple times by multiple groups
Note that:
- the data (as collected) are accurate
- the models are predictive of future data
- the finding has been replicated multiple times by multiple groups
• • •
Missing some Tweet in this thread? You can try to
force a refresh
