Viewing profile — clouden

Logistic regression is similar to his example on fitting a line to an exponential trend. The differences is that in logistic regression, the equation would be ln(p)/ln(1 - p) = α +…

A p-value of 0.04 is not very strong evidence of an effect, though it is significant at the 0.05 level. This study says at most that the effect should continue to be studied.