An earlier section considered the drivers and scope of the IS discipline primarily from the teaching perspective. This sub-section considers the related, but somewhat different, question of what IS academics have considered appropriate areas of research.

Early endeavours to define the scope included Mason and Mitroff’s (1973) program for research on MIS, Ives et al.’s (1980) framework for research in computer-based MIS, Galliers’ manifesto for Australian-based research (1987) and Jeffery and Lawrence’s special issue on current research directions in IS (1986). Reviews of the research undertaken in IS include Culnan (1986, 1987), Alavi et al. (1989), Alavi and Carlson (1992), Glass (1992), Avgerou et al. (1999), Galliers and Whitley (2002), Vessey et al. (2002) and Banker and Kauffman (2004). Each of these draws attention to the enormous breadth of the topics addressed. The diversity arises in at least two dimensions:

A few attempts have been made to adopt the encyclopaedists’ approach of enumerating the topics that are within the scope of the IS discipline. More adventurously, a few have attempted taxonomies, in order to impose some order on the chaos. The most successful work of this kind is that by Barki et al. (1988, 1993). The second paper reported that articles published in just seven major journals in 1987–92 identified about 2000 different keywords. Barki et al.’s revised classification scheme of 1993 included 1300 keywords under nine major and 56 minor groupings—an increase of 175 on their original 1988 version. It appears that the Herculean task has not been repeated since. Moreover, the use of the Barki scheme appears to have subsided, as reliance on brute-force, free-text search engines has increased. Nonetheless, it is a highly valuable tool of historical analysis.

As Table 2.4 shows, about only half of the Barki et al. (1993) keywords were concerned directly with the core areas of the IS discipline; one-quarter were associated with reference disciplines, and one-quarter with external drivers and constraints. The discipline could be described, kindly, as being strongly professional in its orientation and sensitive to its environment and the needs of its clientele. Alternatively, it could be depicted more critically, as lacking confidence, being derivative, lacking in fundamentals and driven mercilessly by its rapidly changing context.

A later analysis examined articles published in Information & Management and MIS Quarterly from 1981 to 1997, using the Barki high-level structure. Claver et al. (2000) found that the largest concentrations of publications were IS development (13.2 per cent of 1121 papers), DSS (8.9 per cent) and IS evaluation (7.8 per cent). Avgerou et al. (1999) evaluate research foci and methods in Europe, and Galliers and Whitley (2002) analyse the papers accepted at ECIS conferences.

Studies of this nature conducted in Australia include Galliers (1987), Ridley et al. (1998), Pervan and Cecez-Kecmanovic (2001) and Pervan and Shanks (2004, 2006). Pervan and Cecez-Kecmanovic (2001) reported on the results of a survey of heads of IS groups regarding the research profiles of their groups. The heads of 21 of the 36 targeted IS groups responded. This represented more than 400 of the approximately 700 IS academics thought to be active in IS in Australia. The responses confirmed that Australian IS reflects the enormous breadth of scope elsewhere. Similar diversity was detected in relation to the unit of analysis of the research conducted. The primary beneficiaries of the research were identified as being predominantly IS professionals and managers—consistent with the notion of being a professionally oriented discipline—although the subsequent data in Pervan and Shanks (2004, 2006) suggest a strong focus on writing for other academics as well. The average publication count disclosed was about two for each staff member per annum, of which one-third was in journals and two-thirds in conferences. The research funding available was generally small, but Pervan and Shanks (2004) suggested that it was growing.

The diversity apparent in research topics is just as evident in IS academics’ choice of research methods. Taxonomies of research techniques include Alavi and Carlson (1992) and Palvia et al. (2003, 2004). The 1980s saw an extended period of intolerance and mutual distrust and dislike between groups who adopted particular research techniques. The tensions were variously methodological, philosophical and transatlantic. While differences remain, there is sufficient mutual respect and ‘agreement to disagree’ that little energy has been wasted during the past decade. The discipline has become catholic, in one of the positive senses of the expression.

In the IS community internationally, Claver et al. (2000) found that ‘theoretical studies’ (as defined by Alavi and Carlson 1992) fell from 56 per cent to 20 per cent between 1981 and 1983 and 1996 and 1997, while empirical studies rose from 44 per cent to 80 per cent. ‘Field studies’ (although in many cases mere questionnaire-based surveys) rose from 18 per cent to 52 per cent, while case studies rose to a high of 23 per cent but fell back to their original 18 per cent.

In Europe, Avgerou et al. (1999) found that the techniques used varied widely between countries, and differed from those prevalent in the United States. A large proportion of German researchers focused on technology development and testing, whereas those in many other countries conducted a great deal more qualitative analysis.

In Australia, Pervan and Cecez-Kecmanovic (2001) reported that ‘responses revealed dominance of a positivist paradigm, but the interpretivist paradigm was also often used’. Further, ‘the full range of research methods are being used, from survey to action research, to technology development and testing’. Pervan and Shanks (2004, 2006) suggest that interpretivist approaches have been growing in popularity. Critical-theory approaches remain little used.