In a multi-user distributed haptic virtual environment (DHVE), object state update messages are necessary to maintain a consistent virtual environment at every entity involved. Traffic control schemes applied to reduce the usually high number of update packets lead to inconsistencies among the participants. Whenever a new state update is received, the discrepancy between the old and new object state has to be resolved. For this purpose, various interpolation methods can be used to smoothly correct the state error. In this paper, we investigate and compare three such methods using subjective experiments to determine a haptically preferred method that leads to a convincing force feedback. Besides basic linear interpolation, we test two other methods from literature which are based on cubic polynomials. Our experimental evaluation, based on a pursuit-tracking task and pairwise comparisons, shows that linear interpolation performs better than the two cubic approaches and is subjectively preferred.