In recent years, the ability to use mathematics flexibly in everyday situations is increasingly accentuated, and mathematics education gives more and more attention to problems with realistic contexts. However, it is reported very often that learners have substantial difficulties in connecting mathematics to real situations. In order to solve complex and authentic problems embedded in real situations it is necessary to find translations from verbal descriptions to mathematical notations and to interpret mathematical results with respect to the real situation. In the present work, it was tested experimentally whether translation and interpretation competencies are fostered by presenting multiple choice options from which the learners have to select the correct one. While benefits of translation answer choices emerged, interpretation answer choices did not support learning. However, when the learners were asked to rate how much they liked the problems, opposed results were obtained, indicating that objective learning outcomes and subjective scores did not correspond. Possible reasons for the inconsistent and to some extent unexpected learning outcomes and for the divergence between objective results and subjective statements are discussed.