Why is it useful to plot ratio estimates and their corresponding confidence intervals on the natural log (ln) scale?

a. Because non-statistically significant results on the ratio scale become statistically significant on the natural log scale.

b. Because most readers of journal articles quantify everything on the natural log (ln) scale.

c. Because the magnitude of an association, and the corresponding 95% confidence interval are comparable in size and width, respectively, regardless of the direction of association (Group1/Group2 versus Group2/Group1), on the natural log (ln) scale.

d. Because the association quantified by a ratio is more likely to be statistically significant if presented on the natural log (ln) scale.