Appetite and obesity · Cell Rep · 2019
Paper: two appetite signals cooperate. Its own score: chance gives such results nearly half the time.
How often chance alone would make the pair look this much stronger together
The paper says
Twelve rats each got, in turn, leptin, IL-6 (two appetite-curbing signals), both, or neither, into the brain, at doses too small to work alone. The summary says IL-6 “interacted with leptin to reduce feeding”: together they cut eating more than their separate effects added up.
Its own data show
The paper’s results print the test’s score, 0.563, beside a p-value of 0.05. The score measures the gap between the pair’s effect and the separate effects added up, relative to how much rats vary anyway. The standard formula for this setup (12 rats, all 4 treatments each) turns 0.563 into a p-value of 0.47.
A p-value is how often chance alone would give a gap this big if there were none; the paper’s bar is below 0.05 (1 time in 20). The score and the 0.05 can’t both be right, and 0.05 itself misses that bar. The claim rests on this one test.