Interaction is a key element in turning public displays into a platform for social interaction, making them more engaging and valuable. However, interactive features are still rare in public displays, due to the lack of generalised abstractions for incorporating interactivity. In our work, we explore to what extent the concept of interaction widget, which was so successful on desktop computers, could also be used as an abstraction for remote, shared interaction with public displays. A particular challenge is presenting input feedback in this shared, multi-user, and indirect interaction setting. In this paper, we present a study on the feedback mechanisms of these widgets, to determine if users are able to understand the results of interactions in single and multi-user settings. We have evaluated three feedback mechanisms and the results indicate that the general mechanism provides an appropriate sense of what is happening and could in fact provide general awareness of the interaction alternatives and current status, even in multi-user scenarios.