This paper reports the first results of the extension of citation analysis to 'non-source' items, which is one strand of an
extensive study of quantitative performance indicators used in the assessment of research. It would be presumptuous to draw
firm conclusions from this first foray into the realm of non-source citations, however our analysis is based on an extensive
experimental database of over 30,000 publications, so the results can be viewed as strong pointers to possible generalised
outcomes. We show that it is possible to mine ISI databases for references to a comprehensive oeuvre of items from whole institutions.
Many types of publications are visible in the ISI data - books, book chapters, journals not indexed by ISI, and some conference
publications. When applied to the assessment of university departments, they can have a significant effect on rankings, though
this does not follow in all cases. The investment of time, effort, and money in a significantly extended analysis will not
be equally beneficial in all fields. However, a considerable amount of testing is required to confirm our initial results.