I do not agree that you automatically calculated for the mean score and used it to fill in the missing value in the data set. What about looking at the answers from other observations with similar conditions and using these info to guess what the missing value should be.
The critical Z score values when using a 95% confidence level are -1.96 and +1.96 standard deviations. resources.esri.com/help/9.3/arcgisdesktop/com/gp_toolref/spatial_statistics_toolbox/what_is_a_z_score_what_is_a_p_value.htm