- #1
Ma Xie Er
- 12
- 0
I'm fresh out of school, working for a department where there's some office politics. A co-worker, call this person worker X, is friends with the boss. Worker X is using logistic regression to build a model, and whenever I question anything, the boss tells me worker X is too busy to answer questions.
I find out later, after reading a paper from worker X, that some fishy things are going on. Specifically I'm reading worker X's model revision procedures and...
Step 10) Keep variables which were statistically significant in samples A and B.
...
Step 19) Conduct cross validation study by applying model coefficients derived from sample A to sample B.
I am scratching my head here and thinking "wtf is worker X doing?".
If I call worker X out, I risk a lot. My boss is not mathematically inclined, and makes decisions based on concensus.
What to do?
I find out later, after reading a paper from worker X, that some fishy things are going on. Specifically I'm reading worker X's model revision procedures and...
Step 10) Keep variables which were statistically significant in samples A and B.
...
Step 19) Conduct cross validation study by applying model coefficients derived from sample A to sample B.
I am scratching my head here and thinking "wtf is worker X doing?".
If I call worker X out, I risk a lot. My boss is not mathematically inclined, and makes decisions based on concensus.
What to do?