Which option would you select to compare treatments?

Option 1: “This database is a big mess, let’s just analyze for treated vs. non-treated.”

Option 2: “Marketing is only interested in 400 mg, since 800 mg is too toxic. Let’s do four t-tests, ‘is dose X signifi-cantly different from placebo.”

Option 3: “Why don’t we apply ANOVA to use all data simultaneously for the comparison between doses and placebo.”

Option 4: “Maybe we should treat dose as a continuous variable and plot it vs. effect.”

Option 5: “Some people like to plot AUC vs. dose. Does it really make a difference to plot AUC?”

Option 6: A different form of analysis?