Narayanan V.K., Armstrong D.J. - Causal Mapping for Research in Information Technology (2005)(en)

270 Ackermann & Eden

group. Moreover in terms of operations, it was important to consider the views not only of those working in the unit but also those operating in the different sites across the globe. Their views needed to be included. Management reviewed the budget and it was agreed that $350,000 would be available for software development with the total budget for staff time and training aiming at a maximum of $750,000 total costs. They were keen to ensure that the new system would be within budget and on time.

As a means of involving the different perspectives and also ensuring a sustainable and achievable product, the unit decided to run a workshop using the causal mapping process and associated software5. In this manner, it would be possible to capture all the different views along with their explanations and consequences, thus building up what was hoped to be a full picture of requirements, aspirations, issues, etc. Thus, the workshop was expected to ensure that those attending felt an involvement in the system development and that all users and the computer support staff understood the rationale for the intended system.

The management team spent some time considering specifically who to involve, as they wanted to ensure that two members of staff represented each constituency. They also wanted to employ the services of a neutral party to facilitate, allowing those attending to concentrate purely on contributing. Once the composition of the group was settled, and a facilitator chosen, a date for the workshop meeting was set.

The day commenced with an introduction from the facilitator, who explained the process to be adopted. Each participant had in front of them a networked laptop. Through this medium they would be able to contribute their concerns, aspirations and requirements simultaneously and anonymously. It was believed that this would be helpful, as firstly there was a lot to cover in the day, and secondly there were some concerns that some participants might feel constrained due to the presence of management. The intention was to start with possible reasons for wanting a new system, before eliciting their aspirations for the new system.

The group began by surfacing their concerns about the current situation — the reasons why a new system was needed. They did this by typing in the statement and seeing it appear both on their laptop workstation and on the public screen (which enabled them to see the views of others and so build on them). It soon became clear that many found this a cathartic process — they were obviously very fed up with the present system! After about 10 minutes participants felt that they had covered this topic.

While they had been doing this, the facilitator had been working to position the statements so that they were roughly clustered and hierarchically positioned. This allowed her to then, with help from the group, begin the process of linking them6 together to develop the causal network where the links represented causal relationships. For example, the group believed that 5 “rigid structure of the systems rather than flexible structure” led to 6 “unable to get decent statistical information” which in turn meant that 7 “they experienced difficulties with managing exam board decisions” (See Figure 2). It was explained that the numbers associated with each statement were for reference only and simply made manipulation easier, and that the three dots (an ellipsis) was a short hand for “rather than” — enabling the capture of contrasting situations.

During the discussion of the triggers for a new system it became clear to the group that they were already surfacing implicitly some of their aspirations for the system (or at least

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

Using Causal Mapping to Support Information Systems Development 271

the contrast of the goal). Two concerns, namely “drop out rates” and “inefficient management of student progress” were, if managed, important (negatively expressed) goals. To reflect this insight, the facilitator edited the original wording so that these concerns now read in a more aspirational manner (for example, instead of stating “drop out rates too high,” the statement now stated: “reduce drop out rates rather than drop out rates too high”). In addition, to reflect the change in status of the statement, a new style (representing through the use of a different font and colour a new category) was created called goals and these two statements given this attribute.

Continuing, the group then spent time considering what might be other goals of the system. Again using their ability to directly type in their contributions along with any links (by now they were familiar with the process) a draft goal system emerged (see Figure 3). Not surprisingly, money featured prominently! However, as the group began to explore the links between the goals, it became apparent that a number of participants felt that they needed to consider the issues more fully — as they were unsure whether they had surfaced all of the goals, or whether they fully understood them. It was time to open up the discussion further.

Using a new view (similar to sheets in Excel) the group then began to consider the new system more deeply. Contributions came rapidly. Within 20 minutes, another 50 or so statements were surfaced. The group began the process of examining this new material (see Figure 4) — suggesting possible links, amending statements to make their meaning more clear, and adding new material. To ensure that there was equal representation, the facilitator created a number of new categories representing the different stakeholder

Figure 2. The first stage: expressing concerns or triggers for a new system (note those statements that are boxed are those that were converted to goals)

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

272 Ackermann & Eden

Figure 3. The developing emergent goals system

constituencies and with the help of the group applied these to the contributions. Having done this, she was then able to note that both marketing and operations had contributed the most (11 and 20 respectively) with computer services raising 10, academics only raising four statements and management five. Not a surprising result given that both marketing and operations would use the system most, with computer services having to maintain and upgrade it.

This insight prompted further reflection and analysis. By examining the tails (those statements that had no statement linking in/supporting it), it was possible to ensure that all systems implications were addressed by ensuring that each tail was a statement of system requirement, or of system characteristic. Where necessary, this meant adding a statement from computer services staff regarding the particular function or action required. Moreover the group were able to check that each system implication led to a goal, thus ensuring that all system characteristics or requirements did contribute to the stated overall goals. Both these analysis prompted the group to identify further goals

— increasing the system of goals from nine goals to 14 goals.

In addition, through using the software’s analytical features7 (discussed above) it was possible to do some logic checks. Firstly, an examination of the “busy” points of the map was undertaken (essentially comprising a count of the number of statements linking in and out). The group were pleased and not surprised that six “unable to get decent statistical information” emerged as the most busy (having eight statements linked to it) with 7 “experience difficulties with managing exam board decisions,” 17 “efficient

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

Using Causal Mapping to Support Information Systems Development 273

Figure 4. Developing the map further

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

274 Ackermann & Eden

management of student progress rather than inefficient,” and 38 “academics provide more appropriate advice to students” each having links to seven statements. The group reflected that this seemed correct.

When examining the most potent statements (those statements that supported the greatest number of goals and were therefore important action points) they were also reassured to note that three of the four most potent statements were IT oriented: 28 “use from dial-in facility,” 27 “make it machine operating system independent,” which was supported by the third 39 “ensure all are able to use the system rather than those with the relevant version of Windows.” All of these supported the final potent statement, which was 21 “have real-time access to student progress.”

The workshop ended with a clear agreement of what was wanted from the system, which addressed the needs and concerns of those who would be using it. As a result, it meant that computer services staff were able to put together the document that would be used for the tendering process. In addition the map acted as an organization memory, enabling renegotiation as the development process unfolded and a number of the agreed actions had to be reviewed.

Postscript

The system was seen as very successful. Eighty percent of the users were pleased by the result, the total cost was within budget and the system was delivered on time. The software is the sole record-keeping system in use, substantially reducing errors, and preventing the need for data having to be re-entered. The system, to date, has never failed and application software updates, data migration issues and the occasional bug causing run-time errors on individual desktops are the only causes of system downtime.

However as with all systems, now that the system exists there are additional features desired by users. As with most information systems, these are learned from using the system — with users not aware of them until after the system has been completed — unfortunately mapping can’t pick up these (echoing the findings reported by Orlikowski, Walsham, Jones & DeGross, 1995).

This case provides a brief illustration of how the mapping process supports multiple stakeholders in determining an IS development strategy for which all felt a high degree of ownership. It reflected the views of all there: both users and developers — quiet members of staff as well as those who were socially confident. The resultant model helped members to develop a common language and based upon this negotiate an agreed way forward. Most importantly it enabled participants to understand the reasons for requirements and how they causally linked to supporting goals.

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

Using Causal Mapping to Support Information Systems Development 275

How Mapping Can Support the IS

Development Process

One of the most difficult aspects in IS development is that of enabling a sensible conversation between IS developers and users, in the manner described above. Managers use a different language from IS developers — one that is driven by the changing needs of the business from day-to-day. With the exception of techno-enthusiasts, managers are unimpressed by the inflexibility of information systems and the requirement to learn new system interfaces beyond those of Microsoft (this was true in the above case where academics who would only infrequently use the system struggled to remember how to access options). Moreover, in addition to managing the different views of managers (who were taking a “sponsor”-like role), a system has to take into account the fact that the users are not universal in their requirements. In the above case, operations staff wanted different capabilities from Marketing, with those staff overseas facing different problems to those based locally. Similarly IS developers use their own particular jargon in a “taken-for-granted” way. Few IS developers have direct business or managerial experience. Consequently causal mapping encourages IS staff to understand the business reasons for their developments, and to justify changes with respect to the mapped goals.

There is little to be achieved by arguing that one group or the other should be different. As with the differences between marketing people and operations people, these differences reflect important specialist expertise particularly so in the case of IS developers, where their expertise can often be more opaque than for other experts in the organization. Organizations function well when multiple perspectives can be fruitfully harnessed. As we have argued above, facilitating mutual understanding and mutual respect depends upon establishing and negotiating linkages between user roles and development possibilities. The assertion that users do know, or should know, their requirements, is not helpful. Users know something about their role in the organization but can only express requirements when they have an understanding of what can be done, and developers can only express what can be done when they have a better understanding of users requirements (each category having a different set of understandings).

By its very nature, causal mapping is about linkages. Linking statements made by both users and developers in a manner that gradually builds a visual artefact (a map) aids different stakeholders in moving from divergent positions to one of convergence and encourages joint ownership. The meaning of any statement is enhanced through its linkages rather than simply relying on the words in the statement. For example, in Figure 4 the statement regarding “unable to get decent statistical information” is given further meaning through the statements explaining it and the six purposes (consequences) realised if it were delivered as “having decent information.” Those supporting it are: “rigid structure of the systems rather than flexible structure” and “[NOT]8 have realtime access to student progress.” This means that a statement made by a user that is linked through an explanation or consequence to a statement made by a developer elaborates the meaning of both statements to both parties. This is a form of psychological negotiation. Within cognitive psychology this is known as the elaboration of a Personal

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

276 Ackermann & Eden

Constructs System (Kelly, 1955). The map, as an artefact and a “visual interactive model,” is thus a device to aid psychological as well as social negotiation.

The map also serves two other mundane but important outcomes. Whether it is recorded through the use of Oval Mapping on the wall, or through the use of computer-aided representation, it is both a continually developing and changing set of minutes (or organisational memory) of the meeting, and a formally constructed model of means to ends (record of the consequences of purposeful action).

The formalities of this type of causal mapping require statements to be made in an actionable format, encouraging statements to be made by both users and developers that naturally encourage each to think about “so what?” (consequences) and “how?” (explanations). The hierarchical structure of mapping (with the goals at the top, key issues supporting them, with options and assertions at the bottom) that follows from arrows representing “means to ends,” naturally forces thinking about ultimate desired outcomes (goals). The likelihood of multiple consequences and multiple explanations naturally represents goals as being interconnected where each goal supports others and, in turn, is supported by others. Information systems are there to help sensible choices to be made by managers. Therefore information, its timeliness, its setting within the context of other information, and its accessibility are determined by the extent to which it can support goals or help avert negative goals. Inevitably some design requirements for information provision are more potent than others. Because the map shows causal linkages, it becomes easier to make judgments about the relative potency of particular aspects of information system design. Thus, some propositions within the map may have consequences for many goals and also provide for many causal chains to the same goal (and so robustness) and so be more potent. Typically the degree of potency provides important clues for prioritising requirements.

The structure of a map, as nodes (actionable statements) and arrows (causality) invites categorization. Thus, each node can be attributed a particular “style” (usually a particular font and/or colour) indicating, for example, different types of requirements, different resource demands, different a delivery time lines, and different priorities. The categorisation may occur during a workshop, or as the model is used as a part of project management (see Eden & Ackermann, 1998). As categorization emerges the software permits analysis of the categories, for example, which statements occur in specified categories but not in others. This means it is possible to determine which proposals are, for example, both inexpensive and high priorities.

Conclusion

Although there are other problem-structuring approaches that help to develop effective IS requirements — most notably Soft Systems Methodology (SSM) — causal mapping and associated software offers a degree of formal modelling that links easily to more traditional IS development methods. It is also, in practice, a transparent method.

In particular, and as with Soft Systems Methodology, the process of mapping within the context of a workshop enables several stakeholders to see their views within the context

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

Using Causal Mapping to Support Information Systems Development 277

of others, develop a richer understanding of possibilities, and avoid miscommunication and possible dysfunctional conflict. Using mapping as a device to help negotiation is also likely to build a more collaborative approach to information systems design. Those developing Information Systems therefore will benefit from using mapping, not only through the added benefit of having a shared agreement for the resultant system but also through being able to involve a wide range of stakeholders.

Mapping provides a natural format for ensuring linkage and integration between IS potential and the business needs of the organisation. Through building up a shared understanding, more effective use of the information systems is likely.

Moreover, as noted by Nelson, Nadkarni, Narayanan and Ghods (2000), mapping can reveal patterns of behaviour, theories in use, which help researchers understand the process of mapping and how it can contribute towards better outcomes. This may help researchers in developing methods for information system design and analysis, that more effectively assist the organization.

Alongside causal mapping, the specially designed Group Support System (Group Explorer) software provides the potential for high productivity meetings as well as an online capture that becomes a natural organisational memory and project-planning tool.

References

Ackermann, F., & Eden, C. (2004). Using causal mapping: individual and group; traditional and new. In M. Pidd (Ed.), Systems modeling: Theory and practice (pp. 127145). Chichester: Wiley.

Ackermann, F., & Eden, C. (2003). Powerful and interested stakeholders matter: their identification and management. Academy of Management Proceedings of the Best Papers, Seattle, WA.

Ackermann, F., & Eden, C. (2001a). Contrasting single user and networked group decision support systems for strategy making. Group Decision and Negotiation, 10,47-66.

Ackermann, F., & Eden, C. (2001b). SODA - Journey making and mapping in practice. In J. Rosenhead & J. Mingers (Eds.), Rational analysis in a problematic world revisited (pp. 43-60). London: Wiley.

Ackermann, F., Eden, C., & Brown, I. (2004). The practice of making strategy. London: Sage.

Ackermann, F., Eden, C., & Williams, T. (1997). Modeling for litigation: Mixing qualitative and quantitative approaches. Interfaces, 27, 48-65.

Boddy, D., Boonstra, A., & Kennedy, G. (2002). Managing information systems: An organizational perspective. Harlow, Essex: Financial Times Prentice Hall.

Boland, R. J., Tenkasi, R.V., & Te’eni, D. (1994). Designing information technology to support distributed cognition. Organization Science, 5, 456-475.

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

278 Ackermann & Eden

Bonarius, H., Holland, R., & Rosenburg, S. (Eds.) (1981). Personal construct psychology. London: Macmillan.

Brown, S. (1992). Cognitive mapping and repertory grids for qualitative survey research: some comparative observations. Journal of Management Studies, 29, 287-308.

Bryson, J., Ackermann, F., Eden, C., & Finn, C. (2004). Visible thinking: unlocking causal mapping for practical business results. Chichester: Wiley.

Bryson, J. M., Ackermann, F., Eden, C., & Finn, C. (1995). Using the ‘Oval Mapping Process’ to identify strategic issues and formulate effective strategies. In J. Bryson (Ed.), Strategic Planning for Public and Nonprofit Organisations. San Francisco: Jossey Bass.

Bryson, J. M., Cunningham, G. L., & Lokkesmoe, K. J. (2001). What to do when stakeholders matter: the case of problem formulation for the African American men project of Hennepin county, Minnesota. Public Management Review, 62, 568-584.

Checkland, P., & Holwell, S. (1998). Information, systems and information systems. Chichester: Wiley.

Checkland, P., & Scholes, J (1990). Soft systems methodology in action. Chichester: Wiley.

Downs, E., Clare, P., & Coe, I. (1988). Structured systems analysis and design method (SSADM). London: Prentice Hall.

Eden, C. (1988). Cognitive mapping: A review. European Journal of Operational Research, 36, 1-13.

Eden, C., & Ackermann, F. (2001). SODA - The principles. In J. Rosenhead & J. Mingers (Eds.), Rational analysis in a problematic world revisited (pp. 21-42). Chichester: Wiley.

Eden, C., & Ackermann, F. (1998). Making strategy: The journey of strategic management. London: Sage.

Fransella, F., & Bannister, D. (1977). A manual for repertory grid technique. London: Academic Press.

Hunter, M. G., & Beck, G. (2000). Using repertory grids to conduct cross-cultural research. Information Systems Research, 11, 93-101.

Jayaratna, N. (1994). Understanding and evaluating methodologies. London: McGraw Hill.

Kelly, G. A. (1955). The psychology of personal constructs. New York: Norton. Martin, J. (1991). Rapid application development. Englewood Cliffs, NJ: Prentice-Hall. Martin, J. (1986). Information engineering. Carnforth, UK: Savant.

Mumford, E. (1983). Designing Human Systems: the ETHICS method. Manchester: Manchester Business School.

Nelson, K., Nadkarni, S., Narayanan, V.K., & Ghods, M. (2000). Understanding software operation support expertise: A revealed causal mapping approach. MIS Quarterly, 24,475-508.

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

Using Causal Mapping to Support Information Systems Development 279

Orlikowski, W. J., Walsham, G., Jones, M. R., & DeGross, J. I. (1995). Information technology and changes in organizational work. London: Chapman Hall.

Simon, H. A. (1959, June). Theories of decision making in economics and behavioural science. American Economics Review, 263.

Stowell, F. (1995). Information Systems provision: the contribution of soft systems methodology. London: McGraw Hill.

Tan, F., & Hunter, G. M. (2002). The repertory grid technique: A method for the study of cognition in information systems. MIS Quarterly, 26(1), 39-57.

Zuboff, S. (1988). In the age of the smart machine. New York: Doubleday.

Endnotes

1Decision Explorer is software designed specifically for causal mapping and is available through www.banxia.com.

2Group Explorer is designed to facilitate the fast and anonymous construction of causal maps for participant groups of 5-15 persons or groups and is available through www.phrontis.com.

3It is not untypical to have maps comprising 1000 nodes and 1500 relationships.

4This case has been amended for confidentiality reasons and the report considers the very early stages in the process.

5See Bryson et al. (2004) Visible Thinking: Unlocking causal mapping for practical business results, Wiley, Chichester for more details on the mapping process and additional cases.

6For more detail about different modes of working see Ackermann, F. and Eden, C. (2001) ‘Contrasting Single User and Networked Group Decision Support Systems,”

Group Decision and Negotiation, Vol 10, 1, pp. 47-66.

7The software, Decision Explorer, acts as a relational database allowing users to select which parts of the model they wish to examine, view whether there are linkages to other potentially relevant material, categorise statements according to their status, e.g., key issue, and carry out analysis on the structure.

8The [NOT] represents the fact that there is a negative link namely “having real time access” leads negatively to “unable to get decent statistical data.”

9Based upon the Ackoff and Emery typology (Ackoff, R. L. and Emery, F., On Purposeful Systems. London: Tavistock; 1972).

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.