Causal Mapping Background

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
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 

Causal maps have been used to represent managerial cognition at both the individual and

group levels (Axelrod, 1976; Eden & Ackermann, 1998a; Huff, 1990; Meindl, Stubbart, &

Porac, 1996). From a managerial and organizational cognition perspective, five causal

mapping approaches have been used to produce collective causal maps (see Table 1).

Most collective causal map approaches capture the data for the collective maps using

individual maps. The individual maps tend to be either created using a participant-driven

interview, such as the Self Q interview (Bougon, 1983), or a negotiated researcher and

participant interview (Eden & Ackermann, 1998a). The advantage of a participant-driven

approach is the minimization of the possibility of researcher bias impacting the creation

of the individual maps (Nicolini, 1999).

All of the approaches for creating collective maps from individual maps require that the

concepts used in individual maps be standardized in order to create collective maps. The

use of congregate labels created by the researcher to group similar concepts used across

individuals is common to all approaches that merge individual maps into collective maps

(Bougon, 1992; Eden & Ackermann, 1998a, 1998b). In the merging processes associated

with the first four approaches in Table 1 (congregate, shared, group, and oval maps), this

standardization process occurs after the individual maps are created. Congregate labels

are based on researcher’s and possibly participant’s identification of similarities of

beliefs contained in the individual maps. The congregate labels are then substituted in

the individual maps. Once the congregate labels have been placed into the individual

maps, the individual maps can then be merged based on the common nodes (congregate

labels) contained in the individual maps. As a result, the process of merging individual

maps into a collective map is both time consuming and results in a loss of information

regarding idiosyncratic differences among individual belief structures. In addition,

researcher bias may be present as the research/facilitator usually determines the

congregate labels across individual maps (Nicolini, 1999). In contrast, our process

(Tegarden & Sheetz, 2003) enables the individuals in a decision-making team to agree

upon the congregate labels so that researcher intervention and bias is minimized.

Type Data Capture Approach Data Merging Approach

Congregate Map

(Bougon, 1992)

Participant-Driven Researcher-Driven

Shared Map

(Langfield-Smith, 1992)

Researcher- and Participant-Driven Researcher- and Participant-

Driven

Group Map

(Eden & Ackermann, 1998a)

Researcher- and Participant-Driven Researcher-Driven

Oval Map

(Eden & Ackermann, 1998a)

Researcher- and Participant-Driven Researcher- and Participant-

Driven

Group Map

(Tegarden & Sheetz, 2003)

Participant-Driven Participant-Driven

Table 1. Types of collective cause maps

To identify cognitive diversity or what we call cognitive factions in a top management

team, we cluster the causal maps created by the members of the top management team

based on the similarity of the cause-effect linkages between the nodes (congregating

labels). A similar approach was employed by Reger and Huff (1993) to compare cognitive

similarities and differences of industry maps across top managers within an industry. In

the next section we describe the methodology used to identify the cognitive factions.