Giving and receiving peer feedback is seen as an important vehicle for deep learning. Defining assessment criteria is a first step in giving feedback to peers and can play an important role in feedback providers' learning. However, there is no consensus about whether it is better to ask students to think about assessment criteria themselves or to provide them with ready-made assessment criteria. The current experimental study aims at answering this question in a secondary school STEM educational context, during a physics lesson in an online inquiry learning environment. As a part of their lesson, participants (n = 93) had to give feedback on two concept maps, and were randomly assigned to one of two conditions--being provided or not being provided with assessment criteria. Students' post-test scores, the quality of feedback given, and the quality of students' own concept maps were analyzed to determine if there was an effect of condition on feedback providers' learning. Results did not reveal an advantage of one condition over the other in terms of learning gains. Possible implications for practice and directions for further research are discussed.