In this study, we explored a new experimental methodology for investigating children's (ages 10 to 14) stances with respect to the ethics of online identity play. We used a scenario about peer identity misrepresentation embedded in a 3D virtual game environment and randomly assigned 265 elementary students (162 female, 103 male) to three conditions that differ in how they present the issue. In the abstracted condition, students were asked to answer five open-ended survey questions. In the immersive condition, the same survey questions were embedded in a narrative where the participants were positioned as giving advice to a non-player character (NPC) who was confronted with the issue. In the consequential condition, in addition to advising, participants were asked to act on behalf of the NPC. We found that in the abstracted condition children displayed a more absolutist disposition, casting judgment without interrogating the context. In contrast, in consequential conditions that involved higher levels of role-playing they tended to be more evaluative, considering contextual clues in making judgment. In general, children were bothered by misrepresentations and expected offline and online identity to be more consistent than fractured. (Contains 6 figures.)