We used similar custom-made optically correct eye models in our study; this was shown in the figures. We varied pupil sizes by changing the size of the hole in the canister lid (see the second figure of the referenced article), and the 16 diopter convex lens requires students to maneuver their heads in a similar fashion to performing direct ophthalmoscopy on patients. However, as emphasized in our study, we do not believe that efforts should be directed at building better eye models. Instead, we would like to underscore that our study suggests that educators should instead provide greater emphasis on interpreting fundus photographs, because students are more accurate with and prefer using fundus photography both immediately after training and at 1 year. The importance of the fundus examination in patient care should not be hidebound by an attachment to direct ophthalmoscopy, which has resisted sustained skill improvements even with extremely ardent education efforts by our group and others.