Search

This paper points out that, in spite of the potential power of the Lieberman model for integrated business systems, (4) its application to practical cases is hampered by the fact that most systems can be described only in a series of matrices which are not compatible for multiplication. Two methods of finding the "product" in such cases are described. The second method offers the advantages that only one matrix need be constructed to describe the entire system, that only elementary column operations are required to yield a solution, that the ordering of the system components need not be considered, and that components which are system inputs, system outputs, and those which are outside the scope of the study are immediately identified. A proof of the mathematical validity of the second method is given. [ABSTRACT FROM AUTHOR]

The creation of social science data libraries has reemphasized many important methodological issues associated with contemporary research techniques. Since documentation of the data-gathering and coding processes is vital for the secondary analyst, considerable attention is focused upon these activities. As research workers become increasingly sensitized to the potential exploitation of their data by others, there is a tendency for them to document their activities in greater detail. Since data libraries are energetic in their attempts to obtain such documentation, this activity is being emphasized repeatedly. While it is often submerged somewhat by individuals working with information collected by them personally in a familiar milieu with an opportunity to establish an unequivocal shared understanding with the source of the information regarding each specific datum, this understanding is problematic for the secondary analyst.