Virtual Reality learning environments (VR-LEs) are increasingly used to support learning. However, experimental research on feedback strategies in VR-LEs remains scarce. Drawing on productive failure research, generative learning approaches, and the Interactive-Two-Feedback- Loops (ITFL) model, this experimental study addresses this gap by comparing two feedback conditions in a VR-LE on lunar phases: (a) knowledge of correct-response feedback (KCR), presenting the correct solution after task completion, and (b) informative tutoring feedback (ITF), providing stepwise hints and explanations to scaffold learners toward correct solutions. Participants (N = 182; currently n = 135) complete pre-post-surveys and engage in comprehension tasks of increasing complexity across three phases in VR. The first phase, the productive failure phase, requires solving the tasks without support. In the second phase, the feedback phase, participants receive either KCR or ITF while working on the same tasks again. The third phase uses parallel near-transfer tasks to assess learning gains. Outcome measures include task performance across the VR-phases, perceived competence, intrinsic and extrinsic task value, and metacognitive judgements assessed before and after the VR feedback phase. In addition, prior and post-VR knowledge are assessed. Following data collection, a repeated measures MANOVA will test effects of feedback strategy on repeated measures, and a one-way ANCOVA will examine effects on post-VR knowledge controlling for prior knowledge. We hypothesize more positive changes in task performance, motivational outcomes and metacognitive judgements and higher post-VR knowledge for ITF compared to KCR. Overall, the study provides insights into feedback strategies in VR-LEs to derive evidence-based instructional design principles.