Peripheral interaction, like ambient information systems (AIS), aims at leveraging the periphery of our attention. While ambient information systems address the perception of information, peripheral interaction targets lightweight interaction outside of the current focus of attention. A number of prototypes have demonstrated the value of peripheral interaction through long-term in-situ deployments. Such studies are particularly suited to evaluate peripheral interaction since they enable the integration of devices into daily routines and thereby move interaction to the periphery of attention. However, they do not lend themselves well to early design phases. In fact, the design process completely lacks early evaluation tools to assess design choices. We propose an experimental method for the evaluation of peripheral interaction in early design phases. In a case study, we compared the results of an eight-week in-situ deployment with the results of this laboratory experiment. We carried out the study with both, novice and experienced users (who had participated in the in-situ), and found comparable results across all three situations (in-situ and lab with novice and experienced users).