This paper follows on from a group of internationally published papers in the area of investigating anthropomorphic user interface feedback. In this paper an experiment is described which aims to examine the effectiveness and user approval issues of anthropomorphic versus non-anthropomorphic user interface feedback. This experiment is highly relevant to the research community as Computer Scientists are not in agreement concerning the effectiveness and user approval of anthropomorphic user interfaces. The experiment described in this paper is in the context of Online Factual Delivery. Specifically the area of direction finding was used to test anthropomorphic feedback against nonanthropomorphic feedback. Statistically significant results are presented for the benefit of the research community, where it is clear that in this context, the non-anthropomorphic user interface feedback was more effective and users tended to prefer this feedback.