19/09/2017
It's as if I was living through scenes from The Emperor' New Clothes. Someone describing recent work progress comparing two techniques with r2 varying between 0.3 and 0.5. "It's looking good", said the presenter. And other people started to join, nodding or murmuring "it's looking really good" one after another.
I was stunned.
We were looking at two independent techniques which have good reproducibility in the measured values. And we are dealing with chemicals, not animals, weather objects or Donald Trump. There is no way an r2 of 0.3 can be of any good use for routine scientific studies.
I strongly objected during the meeting and someone else defended the r2 by saying it's better than before. Well, "better than before" is relative and "looking really good" is absolute. I don't claim to have excellent logics but it just beggars belief to see some senior scientists' logics.
No comments:
Post a Comment