You know when newspapers write health advice, "drinking tea with milk can reduce the effects of the calming agents in the tea by 27% because proteins in the milk bind with the calming agents so they aren't absorbed into the blood stream." You know that the 27% is a BS figure.
AP should change their style guides so that instead it is written, simply, "drinking tea with milk can reduce the effects of the calming agents in the tea" or maybe "by a percentage in the range of 0% to 50%".
3 comments:
Is your objection that it is impossible to measure health factors with such precision or that over time other studies will show the results to be different under different circumstances?
The objection is to the precision. Because, as you suggest, no study is perfect.
Just read the intro to "Identification for Prediction and Decision" by Manski, and he addresses this very topic!
Post a Comment