A factor analysis indicates that the survey and system data may measure multiple independent behaviors. Only two questions concerning any aspect of feedback were shown to measure the same underlying behavior. For this reason I could not calculate the reliability of the measures using this pilot study.

Concerning my pilot hypothesis that more feedback would lead to more system use, I found correlations that indicate this may be the case, although the relationship between the feedback activities and system use may be interdependent rather than causal. A concatenation of the weighted browsing variables was significant at the .01 level with a coefficient of .441 as an indication of number of links. It is possible, however, that it could be simple interdependence causing the correlations.