Types and Levels of Collaboration in Interdisciplinary Research in the Sciences

It is common today for scientists to conduct research in 20th century, significant research on this phenomenon collaboration with their colleagues from different institudid not begin until the 1950s. Researchers in sociology, tions and disciplines. This study collected a sample of psychology, and the history of science have initiated in846 scientific research papers published in 1992 and vestigations into how IDR in science is organized (Barmtested three hypotheses on the relationship between reark & Wallen, 1980, 1986; Birnbaum, 1981; Caudill & search collaboration and interdisciplinarity. Collaboration was measured by the number of authors, number of Roberts, 1951; Robertson, 1983), how scientists behave institutional affiliations, number of affiliation disciplines, in interdisciplinary collaboration (Crane, 1976; Crow, and type of collaboration. Interdisciplinarity was meaLevine, & Nager, 1992; Luszki, 1958; Meechan, 1978), sured by the number of disciplines represented in the and how such activities could be facilitated through better journals cited. The results showed significant differmanagement (King, 1964). Information scientists began ences in degrees of interdisciplinarity among different levels of collaboration and among different disciplines. to investigate this area about two decades ago. Although Some disciplines were shown to be highly collaborative, the literature on this topic is still somewhat sparse, rewhile others were not. This analysis led to the conclusion search has been conducted by information scientists on that collaboration contributed significantly to the degree how scientists used information in interdisciplinary reof interdisciplinarity in some disciplines and not in othsearch (Choi, 1988; Chubin, Porter, & Rossini, 1984; ers. In addition to an analysis of publications, this investiClark & Kinyon, 1989), the extent to which the interdiscigation used a survey that asked authors about their forms of collaboration, channels of communication, and plinarity has influenced the use of information (Hurd, use of information. The survey provided some qualitative 1992; McCain, & Bobick, 1981), and the implications explanation for the bibliometric findings. Findings are IDR has for information systems and services (Allen, discussed from the perspective of scientist-scientist in1980). Research on such matters has practical signifiteraction, scientist-information interaction, and informacance for the design of information systems and services tion-information interaction. to support IDR. IDR is an imprecise concept and difficult to define. INTRODUCTION Varied definitions can be found in works by Lowell H. While interdisciplinary research (IDR) has been comHattery (1986), Michael Anbar (1986), Frederick A. mon practice in the scientific community since the early Rossini et al. (1981), and others. IDR has the following characteristics as generalized by numerous writers (Al* To whom all correspondence should be addressed. pert, 1969; Birnbaum 1981; Blackwell, 1955; Caudill & Roberts, 1951; Luszki, 1958): (1) different bodies of Received June 13, 1996; revised September 19, 1996; accepted Septemknowledge are represented in the research group, (2) ber 19, 1996. group members use different approaches in attempting to solve problems, (3) members of the group perform differq 1997 John Wiley & Sons, Inc.