Introduction
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135
136 137 138 139 140 141 142 143 144 145 146
The purpose of this chapter is to provide an overview of some of the key problems that
researchers and policy makers using causal mapping techniques have wrestled with over
the ensuing years, both in order to illustrate the range of choices confronting the wouldbe
user of these techniques and to highlight the strengths and limitations of particular
approaches. Despite the fact that causal and other forms of cognitive mapping techniques
are generally more labor-intensive and time-consuming than other research
methods, in recent years the emerging field of managerial and organizational cognition
has developed dramatically (e.g., Eden & Spender, 1998; Hodgkinson & Thomas, 1997;
Meindl, Stubbart & Porac, 1994; Narayanan & Kemmerer, 2001; Porac & Thomas, 1989),
to the extent that its reach is now extending across virtually all of the major sub-fields
of management and organization studies, including information technology-related
applications (Nelson, Nadkarni, Narayanan & Ghods, 2000a; Swan, 1997). Researchers
have employed a rich variety of methods in an attempt to gain insights into actors’ belief
systems, ranging from the relatively simple process of having participants list basic
concepts (de Chernatony, Daniels & Johnson, 1993; Gripsrud & Gronhaug, 1985) to more
sophisticated procedures such as the development and multivariate analysis of questionnaire
items (Fombrun & Zajac, 1987) and repertory grid and related multidimensional
scaling and related clustering techniques (Daniels, de Chernatony & Johnson, 1995;
Daniels, Johnson & de Chernatony, 2002; Fournier, 1996; Ginsberg, 1989; Hodgkinson,
1997a; Hodgkinson, Padmore & Tomes, 1991; Hodgkinson, Tomes & Padmore, 1996;
Reger & Huff, 1993). Fortunately, a number of comprehensive reviews of the many diverse
methods for accessing thinking in organizational settings have been published elsewhere
(e.g., Fiol & Huff, 1992; Hodgkinson, 2001; Hodgkinson & Sparrow, 2002; Huff,
1990; Jenkins, 1998; Lant & Shapira, 2001; Mohammed, Klimoski & Rentsch, 2000; J.
Sparrow, 1998; Walsh, 1995).
In this chapter we shall confine our attention to a consideration of one particular class
of cognitive mapping techniques — causal mapping — that has risen in popularity in
research domains as diverse as strategic management (e.g., Fahey & Narayanan, 1989;
Hodgkinson, Bown, Maule, Glaister & Pearman, 1999; Hodgkinson & Maule, 2002;
Maule, Hodgkinson & Bown, 2003; Narayanan & Fahey, 1990), human resource management
(Budhwar, 2000; Budhwar & Sparrow, 2002), and technological innovation (Swan,
1995; Swan & Newell, 1998). In the words of Huff (1990, p.16):
“Causal maps allow the map maker to focus on action — for example, how the
respondent explains the current situation in terms of previous events, and what
changes he or she expects in the future.”
It is the direct links to action implicit within this approach that make it such a powerful
method, applicable across a wide range of contexts. However, as noted in this volume
by Narayanan (2005), causal mapping techniques have been much under-utilized within
the inter-related domains of information systems (IS) and information technology (IT).
This is highly surprising, given the obvious parallels with general systems theory and
the potential of these techniques to shed light on systems-designers’ and users’
understanding of a range of hardware and software capabilities and limitations (cf.,
Nelson et al., 2000a), thereby extending the repertoire of cognitive engineering tools and
techniques available for use in these domains (Schraagen, Chipman & Shalin, 2000;
Seamster, Redding & Kaempf, 1997). However, if this potential is to be realized, it is vital
that important methodological insights already gained in the context of other domains,
where causal mapping techniques have enjoyed widespread prominence, are brought to
bear in the context of IS and IT applications. Since Axelrod (1976) produced his classic
book that introduced causal mapping to the field of policy analysis, a number of
significant methodological issues have risen to the fore across a range of fields, which
in turn has stimulated much thinking and further advances.
In this chapter we map out some of the key methodological choices confronting the
would-be user of causal mapping techniques, drawing upon the wider body of research
that has been conducted using these techniques in other domains, over almost a 30-year
period, both in order to illuminate the nature of those choices and to accelerate progress
in these new, inter-related focal areas of application, by distilling the very valuable
lessons that have emerged from extensive prior usage in these other domains. In so doing,
our purpose is to accomplish three principal aims: (1) to illustrate the range of methodological
choices associated with causal mapping techniques; (2) to highlight the
strengths and limitations of the particular approaches identified; and (3) to offer some
general guidelines to aid the would-be user of these techniques. Our recommendations
are not intended to be prescriptive, but to assist potential users of causal mapping
techniques in making methodological choices that are appropriate in particular contexts
of application.
In Figure 1 we present a schematic overview of the principal stages involved in the causal
mapping process, as discussed in this chapter. Undoubtedly, this representation
oversimplifies the complex realities involved. (In practice, for example, the mapping
process is often an iterative one, with feedback sought from participants during or soon
after the construction and analysis stages.) Nevertheless, it serves as a useful framework
to guide those new to the process of causal mapping and provides a clear overview of
the organizing logic we have employed in structuring our chapter.
The chapter is organized in seven principal sections. Following this introduction, we alert
the reader to ongoing philosophical debates concerning the ontological status of causal
maps (also known as cause maps), outlining our own position in respect of these. In the
third section, we identify a number of issues concerning knowledge elicitation that
researchers need to address if they are to make well-informed mapping choices and we
highlight a number of strengths and limitations associated with particular approaches.
Next, we turn our attention to basic metrics for the analysis of individual cause maps. In
the fifth section we discuss issues associated with the aggregation and comparative
analysis of causal maps, while in section six we consider some measurement issues which
are fundamental to the entire mapping process. Finally, we draw together our key
recommendations and overall conclusions. In an accompanying appendix we provide a
brief overview of some of the available computer software systems for supporting users
throughout the various stages of the mapping process, from data collection/elicitation
to analysis and comparison.