In a rat experiment, five groups of animals are exposed to placebo, low, middle, high, and very high dose. There appears to be an increase of death rates with increasing dose. The experimenters perform a logistic regression of death on dose, treating dose as a categorical variable (like in a one - way anova). They obtain p=0.06.

What is your interpretation of the result?

Do you think they used the right testing approach?