It is axiomatic that research in DSS should be grounded in quality judgement and decision-making (JDM) research since it is focused on supporting and improving management decision making. In coding and analysing this JDM grounding, special care was taken to distinguish between merely citing reference theory in introductory or focussing discussion and using reference theory in the design of the research and interpretation of results. Only the second, integral, use of reference theory was coded in this project. Surprisingly, the result was that 47.8% of papers did not use any reference research in judgement and decision-making. Further, the percentage of papers that explicitly use JDM reference research is falling slightly over time. Table 4 shows the mean number of citations to JDM reference research per paper for each type of DSS. Group and Negotiation Support, and Personal DSS have the most reference citations, with the current professional mainstream of data warehousing having the poorest grounding.

Another aspect of the intellectual structure of a field is the degree of coherence between different sub-fields. Arnott and Pervan (2005), using an historical analysis, characterised DSS as a set of partially connected sub-fields. The ‘partially connected’ descriptor hints at a field that may not be as coherent as may be imagined. Table 5 shows the top five judgement and decision-making reference articles for each DSS type. The total number of references per type is shown in the left column and the right column shows the reference ranking and reference frequency for each type. The ‘many’ classification in KM-based DSS and NSS indicates a large number of reference articles with one or two citations in the sample. These ‘many’ articles are different to those cited elsewhere in the table. This analysis of the JDM foundation citations can provide an indication of the coherence of the field. If the key references across different DSS types are similar then the discipline has a high level of coherence.

What immediately stands out in the table is the major disconnect between the grouping of GSS and Negotiation Support Systems, and the other DSS types; there are no common key references between these two groupings. This suggests that they may even be considered as separate academic fields, a notion that is supported by the conduct of separate specialist conferences and the publishing of separate specialist journals. The lack of any JDM references in data warehousing research indicates that it could also be regarded as a separate academic field. Data structures, data quality, and information delivery seems to be this DSS type’s core concerns. Another interesting observation is the integrating nature of Simon’s behavioural theory of decision making across Personal DSS, EIS/BI, Intelligent DSS, and KM-based DSS. The strength of this referencing does indicate intellectual coherence among these DSS types.

This analysis of Table 5 indicates that the DSS field is fragmented with marked disconnects between important sub-fields. In terms of judgement and decision-making reference theory, there appears to be three disjoint groups of DSS research:

A further aspect of Table 5 is the relative age and scope of the reference research. Although the article sample spans 1990 to 2004, the major references are quite old and many are from the 1970s; only two frequently cited references are from the 1990s. This is another aspect of discipline conservatism. In particular, the early behavioural decision theory associated with Herbert Simon is dominant. This could be a negative consequence of Simon’s Nobel Prize and it could be that the academic standing of Nobel Prize winning theory has prevented or discouraged the search for other reference theory. An author is unlikely to be criticised for basing their research on Nobel Prize winning theory. More contemporary theory is often the subject of vigorous debate and can be a riskier prospect with journal editors and reviewers. Daniel Kahneman won a Nobel Prize in 2002 for his behavioural decision theory based on heuristics and biases and the effect of this more recent prize may be to counteract the evident Simon-based conservatism of DSS research.