How do you interpret a logistic regression coefficient as an odds ratio?

• $\operatorname{exp}(w_{j})$ is the change in probability per unit of $x_{j}$
• $\operatorname{exp}(w_{j})$ is the odds ratio: multiplier on the odds for a one-unit increase in $x_{j}$
• $\operatorname{exp}(w_{j})$ equals $R^{2}$
• $w_{j}$ is the direct change in probability

$\operatorname{exp}(w_{j})\; = \;\mathrm{odds}$ ratio (multiplier on odds), not on the probability directly.