Project Management Institute

Identifying project risks

a cognitive approach

Eunice Maytorena,
Centre for Research in the Management of Projects
University of Manchester

Sharon Clarke,
Centre for Research in the Management of Projects
University of Manchester

Marie Dalton,
Centre for Research in the Management of Projects
University of Manchester

Tom Kiely
Centre for Research in the Management of Projects
University of Manchester

and

Graham M Winch
Centre for Research in the Management of Projects
University of Manchester

Proceedings of the PMI Research Conference 11-14 July 2004 – London, UK

Risk management has become an important area of interest in project management practice over the past few years. Interest in the management of risk has increased as competition between firms and the size and complexity of projects have grown. This has lead to the development of best practice standards, tools and techniques as well as numerous writings on the subject with a focus on a more effective risk management process. It is considered that both the risk identification and risk analysis phases of the risk management process are the most important, as they can have the biggest impact on the precision of the risk assessment exercise. While the analysis process, its tools and techniques are well developed, it is recognized that such analysis is dependent on risks being accurately identified in the first place. However, compared with the analysis phase, the process of risk identification is poorly understood and the tools and techniques are less well developed. The aim of this research – funded by The Engineering and Physical Sciences Research Council award GR/N51452/01 – has been to provide a better understanding of the risk identification process within a construction context. That is, how do managers in the construction sector move from initial risk perception to risk identification? Initial results suggest that some of the widely held – but untested – assumptions about an individual's judgement influencing the risk identification process may well be overestimated.

Overview

There is no doubt that the concept of risk is an important issue in society today. One only has to look at current newspaper headlines - global warming, genetically modified food, diets and health - to appreciate this. The concept of risk has become a part of current scientific and political debate, and over the past few decades has become central to our way of life. However, over the years the concept of risk has gradually changed from the idea that risk could bring a chance of something good happening to risk being equivalent to danger. Wilkinson, Elahi and Eidinow (2003) state that a number of drivers such as global interconnectivity, atomization of society, ability to measure and monitor, and technology are rapidly transforming the context and our awareness of that context, which inevitably affects our perception, understanding and approach to risk.

The rapidly changing context also influences the way projects develop. According to Hartman (1997), risks seem more intense as new technologies, stricter regulations and changing business practices are introduced. With project size and complexity on the rise and competition increasing, the management of risks has become an ever more important challenge. This point is reinforced by Royer (2000) who suggests that “unmanaged or unmitigated risks are one of the primary causes for project failure.”

As a result, risk management has become an important area of project management practice. This is evident in the development of best practice standards such as A Guide to the Project Management Body of Knowledge (PMBOK®Guide (2001), the Association for Project Management's Project Risk Analysis and Management (PRAM) guide (Simon, Hillson & Newland, 1997), and the Risk Analysis and Management of Project (RAMP) guide (Lewin, 2002) among others (British Standards Institution and International Organization for Standardization). Although there are important differences between these standards, they share a basic risk management process, illustrated in exhibit 1. This consists of four basic sub-processes located in the context of a clear definition of project objectives that are iteratively looped through the project life cycle:

  • identify and classify the risk;
  • analyze the risk through quantitative or qualitative techniques;
  • respond to the risk through insurance, mitigation, transfer and the like;
  • monitor the risk through the life cycle until its probability of occurring has been reduced to zero.

The analysis and identification phases are considered the most important as they can have the biggest impact on the precision of the risk assessment exercise (Chapman, 1998). However, it is well recognized that such analysis is dependent on risks being identified in the first place. Simply put, if risks are not identified, they cannot be analyzed and managed.

The analysis phase, where the seriousness, impact and probability of risks are ascertained, has been well supported by research with both qualitative and quantitative tools and techniques developed. Qualitative tools include probability/impact rating matrices and assumption testing; quantitative tools include sensitivity analysis, decision trees and Monte Carlo simulation. Research on various aspects of the risk response phase has also been carried out, particularly the use of subcontracting and the role of insurance (Winch, 2002; Stinchcombe, 1985). There is also a growing body of research on the process as a whole, with increasing advocacy of risk management maturity models (Hillson, 2003).

However, the risk identification phase, where the question of what might happen is addressed, has been the subject of much less research than the analysis phase. The subjects covered are the effectiveness of risk identification tools at group level (Chapman, 1998) and influences on risk identification and assessment in construction design management (Chapman, 2001). The focus of the literature is on the tools and techniques used for assisting in risk identification, such as risk breakdown structures (RBS), risk registers and brainstorming.

Risk Breakdown Structures (RBS) provide a hierarchical structure of potential risk sources (Hillson, 2002) from which a list of risks can be drawn through a brainstorming session. A risk register, more widely used, is a list of all the risks that have been previously identified, and its development is typically ad hoc. For this to be of practical use, the register has to be filtered for the particular project under scrutiny and the results prioritized. However, it is not clear how this is done, who does it and how reliable the results are; there appears to be a complete lack of connection with the literature on knowledge management as a tool for capturing organizational learning from projects (Ayas and Zenuik, 2001). Gaining such understanding requires the systematic analysis of data for a large number of projects, but such data sets are difficult to acquire. Brainstorming (Chapman and Ward, 2003) is project-specific and requires a group of experts to creatively consider possible sources of risk. This list is then more analytically considered and key risks identified. The problems with brainstorming are issues such as the selection of the appropriate experts and their number, bringing these experts together frequently enough to be of use to a dynamic project lifecycle, and the avoidance of “groupthink” dynamics.

risk management sub-processes

Exhibit 1 risk management sub-processes
source : Winch (2002)

The Problem: understanding risk identification as a cognitive process

The construction sector has taken on board risk management practices since the early 1980s in response to the increasing uncertainties related to the rapid development of the construction industry. Risk is seen in the construction sector as a negative concept and in this respect the construction industry's definition of risk is different from those of other industries, such as finance. Akintoye (1997) put it thus: “risk is perceived as the likelihood of unforeseen factors occurring, which could adversely affect the successful completion of the project in terms of cost, time and quality.”

In construction projects, there are some known standard risk areas that need to be considered and assessed, but each new project brings along specific project-related risks, which need to be identified. The difficulty is the lack of accurate methods for identifying risks in a construction project. The concern is that in practice the distinction between the risk management process phases are blurred and existing tools are often not sufficient. Egbu (2000) points out that there is a lack of knowledge and doubts their applicability in the construction industry. Currently, construction project managers rely on standard risk assessment pro forma for recurring safety risks, company risk registers, brainstorming sessions between team professionals, and interviews.

Many risk management measures used in the construction industry over the past decade attempt to manage safety, quality or procedural risks-on the assumption that the risks have already been identified and codified. However, there are still the operational risks which are not codified in any risk management system and which are faced by untrained, inexperienced managers on a daily basis. Those responsible for dealing with the project risks appear to comprehend the first stage (identification) and quickly move on to the response stage. This reactive way of dealing with risk is not well understood, with most of the literature noting how to quantify risk, using probability theory or other quantitative methods.

Adams's (1995) risk thermostat, a conceptual qualitative model, is useful as it illustrates how our risk behavior is influenced by our perceptions and attitudes. In short, individual risk-taking decisions represent a balancing act, where perceptions of risk are weighed against the propensity to take risks. How individuals respond to risky or uncertain situations therefore requires an understanding of how we intuitively assess the situation we perceive, before formulating (or justifying) a response.

Thus construction organizations, similar to other project-based organizations, rely heavily on historical data and the judgement of the key individual actors involved in the project to identify risks. These judgements are informed by the individual's knowledge, experience, professional training, role, level of responsibility and length of exposure to the construction industry (Chapman, 2001). The key individuals in a construction project are those who are responsible for the project's success. The same individuals would clearly have the motive to identify the risks that may impact on the project. Their presence on the project gives them the opportunity to carry out the identification, but there has been little work developed on the means by which their expertise and knowledge could be rigorously used. Thus risk identification is part of the much broader problem of judgement under uncertainty (Rosenhead and Mingers, 2001).

Theoretical Background

In this paper we are aiming to improve our cognitive understanding of the risk identification process in a construction context. The method examines the first step in the decision process: information gathering and strategies. But first we need to explain briefly the theoretical underpinning of this approach.

The theoretical underpinning for the research approach lies in the area of decision-making under conditions of risk. The main theory in this area has been the Expected Utility Theory –see Schoemaker (1982) for a review. Here the decision-maker rationally evaluates the probabilities against a final asset position before choosing a course of action--this has been applied to the project risk management context in Chapman and Ward's (2003) concept of efficient risk. However, this theory has been criticized for its assumption that rationality is possible under such conditions, because evidence has been found that decision-makers use flawed heuristics in decision-making, which are subject to systematic biases (Kahneman, Slovic & Tversky (1982); Gilovich, Griffin & Kahneman (2002). Within this perspective Kahneman and Tversky (1979) proposed their prospect theory in which decision makers assign values to gains and losses rather than to final assets and decision weights rather than probabilities. This produces the distinctive s-curve value function of the theory. While there have been important debates within the heuristics and biases literature such as Gigerenzer's (1999) fast and frugal heuristics, this probabilistic approach to decision-making has been widely accepted. However, the heuristics and biases critique of expected utility theory has itself been criticized on methodological grounds due to the artificial nature of the decision problems set (Huber, Wider & Huber (1997). In essence, decision makers are presented with well-defined problems with all required probability distributions available. In practice, an active information search is required by decision makers to tease out the nature of the problem situation and assign the appropriate decision weights to the data. This naturalistic approach is much closer to the sort of situation facing project risk decision makers than those of perfect information envisaged by expected utility theory and bounded rationality envisaged by prospect theory. The research methodology used in this research is therefore based on active information search.

An introduction to the method

Both the theoretical knowledge and most recent research in experimental psychology on the cognitive understanding of risk perception was reviewed by Edkins and Millán (2003) to propose a methodology to study risk identification in a construction context. In particular, the work of experimental psychologists over the last three decades has made an important contribution to the methodological proposal. The specific aim of the method presented is to comprehend the process that individual construction project managers follow when identifying risks. It is important to note that the methodology proposed does not aim to obtain any conclusions about the behavior of project teams; “only once conclusions about the individual can be drawn, focus can be turned on to groups or organizations’ dynamics” (Edkins et al., 2003, p. 3). Therefore, the proposal by Edkins et al. (2003) to use Active Information Search (AIS) and cognitive mapping to study how practicing construction managers identify risk was implemented.

Cognitive mapping (Eden, 1989) is an interactive decision support tool used to analyze the complex or messy processes through which decisions emerge. A cognitive map is a graphical model of an individual's thought process represented through concepts (distinct phrases) and links between concepts--- thus creating a system of concepts---that communicate the nature of a problem. Although cognitive mapping has already been used in the risk management area (Williams, Ackerman & Tait, 1995; Williams, Ackerman & Eden, 1997; Eden, Williams, Ackerman & Howick, 2000), its application to the problem of risk identification and its combination with an active information search methodology is novel.

The research also considered the issue of the individual decision-maker's appetite for risk and the phenomenon of risk compensation where decision-makers adjust the overall level of risk to a constant level in the face of risk mitigation measures (Stetzler and Hofmann, 1996). The proposal of Edkins et al. (2003) noted the importance of taking into account personality factors during the analysis in order to cover a broad range of reasons for the observed behavior. However, the relationship between personality traits and risk-taking behavior is complex, and research linking personality traits with decision-making behavior has been unsuccessful. Adams's (1995) “risk thermostat” concept proposes individual differences in risk-taking, but this has not been successfully operationalized to date. We believe that it is important to calibrate our findings in the light of informants’ appetite for risk. Therefore, as there are no suitable risk instruments available, we developed a psychometric instrument, which draws partially on the work of Pablo and her colleagues (Sitkin and Pablo, 1992; Pablo, 1997) in the oil and gas industry.

This paper focuses on the application of Active Information Search (AIS) and cognitive mapping. For a review of the development of the risk propensity measure (psychometric instrument) used in the research, please refer to Clarke, Dalton, Maytorena, Kiely, & Winch (2004).

Active Information Search to study risk identification in construction projects

AIS was developed to study judgement and decision-making in naturalistic tasks. These are ill-structured problems in information-rich domains, where causal relations and attributions and the decision-makers’ control belief are relevant (Huber et al., 1997). In essence, AIS is a process-tracing technique and although several have been developed over the years (Payne, 1976; Engländer, 1980; Huber, 1997), this research utilizes the technique proposed by Ranyard, Williamson, Cuthbert & Hill (1999) and Williamson and Ranyard (2000).

Payne (1976) introduced process-tracing techniques based on Newell and Simon's (1972) work on human problem solving. However, the approach was too structured and it did not allow for generalization to real-world tasks. In a natural decision situation, an individual actively searches for information that they consider relevant. In this sense, a process-tracing approach focuses on information search and acquisition during the decision process and not on the input and output variables that precede the process.

As a response to the lack of a standardized method available for risky decisions, Engländer and Tyszka (1980) used a question-asking method in multi-attribute choice, where the facilitator acted as an expert. Huber et al. (1997) adapted the question-asking method to develop AIS to study whether decision makers consider imprecise probability information as a deficiency and try to get more precise information. At its core, AIS is a process-tracing technique of information search and collection, carried out in a context of an interview where the respondent is presented with a scenario of a problem. After the review, the respondent asks the facilitator questions in order to obtain information. These questions are recorded and answers are provided in printed form. Huber's model of how individuals reach a decision in a naturalistic situation assumes that the decision maker constructs a simple mental representation of the situation, as well as alternatives, which can change in the course of the decision-making process. If there is uncertainty about which outcome results from the choice of a specific alternative, the decision-maker tries to defuse the uncertain situation by applying a defusing operation.

Ranyard et al. (1999) and Williamson and Ranyard (2000) proposed a process-tracing technique for naturalistic tasks based on Huber's (1997) AIS. This is developed further by providing spoken rather than written answers to questions and by including think aloud instructions, so a conversation approach is adopted. The method produces a concurrent verbal protocol (sequential transcript) developed as the respondent goes through the decision-making situation. Williamson and Ranyard's (2000) evaluation of this approach showed that the conversational (thinking aloud plus verbal responses) AIS is a valuable process-tracing tool which can give insights into approaches to information search and collection strategies, but provides little indication on how information is processed. However, the use of think-aloud instructions and summary sessions are useful for the provision of this information.

The AIS interview procedure, between one-and-a-half and two hours long, was structured in three stages: introduction and warm-up exercise, AIS and scenario exercise and summary. The introduction provided the respondent with the aim of the project, the structure of the interview process and what was expected of the interviewee. The aim was to clarify the exercise to the respondent, but at the same time information was kept to a minimum so as not to influence the outcome of the experiment. The warm-up exercise aimed to clarify the dynamics of the main exercise (scenario), such as thinking aloud and using questions and answers.

The scenario exercise aimed to produce a response from the practicing managers that would be as natural or real as possible. The scenario, based on a real construction project, was developed by the research team in collaboration with the project manager of the real project. The scenario described a building project under a Design and Build contract currently in progress with limited information about its location, team, cost, client, and project status, with a focus on schedule and budget risks. The limited information meant that the potential of the scenario to shape the informants’ responses was kept to a minimum and would produce the responses needed for the AIS method to work. The respondent was asked to assume that he/she was part of the contract team and had to take over the project at short notice. He/she then went through the AIS process described above with the aim of identifying the main risks or risk areas in the project.

The aim of the summary exercise was to obtain a retrospective view of the risks that were identified and the reasons why the respondent considered them risks. The facilitator could also ask why certain questions were asked or not. This type of report can be used to review the validity or consistency of the data collected (Williamson and Ranyard 2000).

The structure of the verbal protocols (transcript) obtained from the interviews consisted of a series of statements from the respondent and questions and answers between the respondent and facilitator. These were subsequently coded and analyzed. Maytorena et al. (2004) provides a more detailed description of AIS procedure applied to this context.

The Analysis

Sample

51 (4 female, 47 male) practicing construction managers from four collaborating firms, each with a minimum of five years of experience, were interviewed. The first five interviews constituted a face validity exercise for both the questionnaire and AIS approach. One interview could not be conducted properly due to time constraints; therefore this data has been excluded from the analysis. In total, the data from 45 interviews has been used in the analysis.

Active Information Search

Both the scenario and summary stages were tape-recorded, and transcripts produced. The verbal reports (sequential transcripts) contain data on the lines of reasoning and type of information searched for and used during the scenario exercise. Due to the volume of data gathered (15-20 pages of transcript per respondent) there was a need to go further than doing a content analysis. Therefore, the proposal of Edkins et al. (2003) to use cognitive mapping as suggested by Eden (1989) was initially followed. In the course of this research stage, however, it was found that, because of the linear nature of the AIS data the full potential of Eden's cognitive mapping approach could not be exploited. Nonetheless, Decision ExplorerTM (cognitive mapping software) continued to be used to map or graphically represent the AIS data. This type of graphical representation can still be considered a cognitive map because they represent “people in relation to their information environment” (Fiol and Huff, 1992). But for the purpose of clarity, they will be referred to as information search maps.

Maps were built by transcribing directly into Decision Explorer. Starting at the beginning of the tape, concepts (distinct phrases) sequentially numbered were entered into Decision Explorer and linked to represent a chronological relation (concepts following in time). A concept could be a question, statement or answer from the facilitator. The sequence of concepts and links was broken when a new question was asked about a new or different theme; the new concept then marked the start of a new line of inquiry and reasoning.

During this stage, a coding framework was developed with three distinct variables: type of phrase, process and outcome. The types of phrase variables were coded as answers (facilitator's input), questions and statements. Exhibit 2 shows the coding framework used. The process variables indicate the approach taken to search and collect information; this could be in a linear or feedback mode. A linear style approach was evident when a series of single independent questions without follow-up were asked in order to collect information on what would be considered standard procedures;, for example, safety, permit requirements or logistic issues. A feedback style approach was evident when a series of related questions was asked in an investigative manner. Other approaches used to navigate through the information were: self-orientation, assessment making, further action, and assumption making. A self-orientation approach was used by the respondent to find his/her bearings or to bring about clearly understood relations. Assessment making was used to express an estimate of value, quality or magnitude. Further action involved a statement where the respondent specifically mentioned activities that needed to be done to solve a problem. Assumption making was used to express something taken as being true for the purpose of argument. Finally, the outcome variables are the risks identified by the respondents.

To improve coding reliability, two coders independently examined the information search maps. The comparison between coded maps indicated a low level of inconsistency.

Coding framework

Exhibit 2 Coding framework

Each coded map contained between 200 and 600 concepts. To help manage the data, we used Decision Explorer's cluster analysis option, where individual information search maps were segmented into groups of concepts (clusters). Concepts were grouped based on the strength of linkage. Each cluster describes the sequence and style of information search a respondent went through during a particular theme of the scenario. In this sense, each cluster describes a specific theme and therefore is given a title.

In order for the data to be analyzed, it is necessary to summarize the “information search maps” further. Therefore, “summary information search maps” per individual were created using VISIOTM These contain data about the information that was sought (cluster titles), the sequence of the information search (arrows), the number of questions asked per theme (cluster size), the style of information search used (feedback, linear, assessment, self orientation, further action), the risks identified, and feedback loops (dashed links). Exhibit 3 provides an example of an individual summary information search map.

In order to read the map, start at the bottom left side, and work your way through by following the arrows. We can see that this respondent searched for information on the design team and their scope of work (cluster 1). In the first instance, this information was sought using a feedback approach (arrowed circles), where 13 questions were asked. This led the respondent to identify five risk events (coded as a, b, q, t and u) listed on the right side of the map. As we follow the arrows, we can see what other themes were covered and where assessments were being made (gray clusters) and with how much information (number of questions in parenthesis). In addition, the orphan risks (not linked) indicate the risk events identified without any additional information acquisition or follow-up.

Summary information search map

Exhibit 3 Summary information search map

Thanks to the collaboration of the project manager of the real project on which the scenario was based, it was possible to assess the potential impact on the project of the risks identified, and benefit from hindsight knowledge regarding the risk events on the project. This allowed a risk identification performance measure to be developed. All identified risk events were input into a matrix and rated on a scale of 1/1 (low) to 5/5 (high); that is, a 25-point rating system. Each identified risk event was rated individually following a two-phase approach: first, in terms of its potential impact on cost and time to the project; second, the degree to which the risk event could be managed or controlled. As the risk events were identified based on a scenario of a real project, the control used for this measure were the parameters within the scenario and the real project itself. The rated matrix was independently reviewed for consistency. The idea was to link these results with the summary maps in search of correlations. These are described in the following section.

Results

The individual summary maps provide an overview of the individual risk identification process. They provide data on the information that was sought, the sequence of the information search, the strategy used for information search, which decisions were based on prior experience or training and which were based on information collected during the exercise. These are described next.

The information that was sought was limited to the context of the scenario. Therefore, themes covered by the respondents included: the team, program and sequencing, building and design details, site conditions and facilities, procurement aspects, regulatory aspects, health and safety, subcontractors, costs and budget, demolition aspects, client-related, environmental aspects, construction, logistics, third parties, and contractual. All these can be considered areas of potential risk sources, which could influence the project's performance mainly in terms of time and cost. Quality was not mentioned in any significant way by respondents. The themes covered can be categorized within the construction risk sources identified by Perry and Hayes (1985), and Mustafa (1991) as physical, environmental, design, logistics, financial, legal, political, construction and operation. The information sought can be considered elements within a bounded mental information environment particular to each respondent. This bounded information environment is surveyed to create a mental model of the problem addressed. For example, the majority of site managers’ maps show that the type of information sought was strongly related to site aspects such as accommodation, health and safety, method statements, things on the ground and logistics--while project managers’ maps tended to seek information on all the themes, indicating a larger bounded information environment. The difference among project managers’ maps seems to relate mainly to the focus or attention brought to particular themes over others. That is, information for a particular theme was sought with a higher or lower level of detail (number of questions per theme), and reviewed more often (self-orientation clusters). Although the sample size is small, it appears, perhaps not surprisingly, that the bounded information environment is highly determined by the discipline practiced. The key here is the ability with which individuals’ access and review information.

The order in which information was sought was again limited to the context of the scenario. As the scenario described a design-and-build project in progress, the sequence of information search was closely linked to the program activities. Respondents generally seemed to use the building process as a mental model for information gathering. Although not all the respondents used this approach throughout the exercise, it was common to find three or four themes of sequenced information search in accordance with the program. Three distinct approaches have been identified in the beginning of the information search. First, respondents wanted to know about site-related activities in progress (activity-orientated); 40% of respondents used this approach. Second, respondents wanted to know about the project objectives, the client and the team (human-orientated); 44% of respondents used this approach. Third, respondents wanted to know if all regulatory aspects were addressed (legally-orientated); 16% of respondents used this approach.

The strategy of information search could be feedback and linear. 48.8% of respondents used a high feedback style of information search, while 51.1% used a high linear (low feedback) style approach. When combined with the performance measure, the analysis showed that 82.6% of high performers used a high feedback style approach of information search to identify risks; 90.9% of low performers used a highly linear style of information search to identify risks. This suggests that better quality risk identification is dependent on the way information is acquired on the one hand, but also how it is processed This is an area that needs further research. Pearson's Chi-Square (x2=24.18; df=1; p<0.01) indicates there is a significant association between risk identification performance and level of feedback style.

Assessment, self-orientation and further action approaches, although used, did not play a significant role in the risk identification performance measure. Nonetheless, self-orientation and assumption making were the most commonly used strategies to understand the information provided. 51% of respondents used a high level of self-orientation; 42.2% of respondents used a high-level approach of assumption making from information provided; and 51% of respondents used a high level approach of assumption making from other sources.

The maps also show what risks were identified based on either prior experience or training (orphan risks). These are risks that were mentioned based on drawings, sketches and program only, and no specific questions or follow-up information were asked for. The types of risks identified were mainly related to safety, logistics, and environmental issues. 40% of respondents used this strategy to identify an average of two risks, out of an overall average of 12. This gives an indication that from initial impressions experience plays little part in identification of high impact risks, but plays more of a part in the identification of standard risk areas.

The analysis highlighted that there was no significant correlation between risk identification performance and years in management role (x2=.203; df=1; p=.652) or years in current job title (x2=.021; df=1; p=.384). This indicates that the assumptions made about the significance of an individual's experience to identify high impact risks may well be over-estimated. Results show that risk identification performance measure declined after the age of 45 and with 20 or more years in a management role. This could indicate however, that managers in this position are more comfortable with risks. This is an area for further research. In terms of education, however, the data suggests that managers with a BS/MS education had higher risk identification performance measures. In this case more data needs to be gathered to establish the existence or not of a statistical correlation.

The scenario mentioned that there were some management concerns, but did not mention what they were. It was assumed that respondents would want to know what these were early on. However, only three respondents (6%) specifically asked in the beginning of the exercise and nine other respondents (20%) asked later on. Why ¾ of the respondents did not think this information was necessary is of interest. Are management concerns so embedded in the general context that they become unnoticed? This question still needs answering.

Conclusions

This paper describes a method for studying the way that project managers in the construction sector go about identifying risks. The review of the risk management literature showed that the risk identification phase of the risk management process, although one of the most important phases, is poorly understood, and that the tools and techniques available are less developed than the ones used in the analysis phase, for example. The review also highlighted the construction industry's concerns about the lack of accurate methods to identify risk, as well as their lack of knowledge and doubts about the suitability of the ones available. Understanding how managers identify risks--that is, understand the means by which they use their knowledge, expertise and information--placed the inquiry in the area of judgement under uncertainty.

The review of the development and critiques of key decision-making theories pointed towards the importance of the use of active information search for teasing out the nature of a problem situation. As a result, the methodology used to study the risk identification process is a conversation-based Active Information Search in combination with cognitive mapping.

Preliminary results show that:

  • the type of information search deployed is highly determined by the management discipline practiced;
  • the sequence of information search selected seems to be acquired based on a mental model of the construction process;
  • the information search strategy plays an important part in risk identification performance;
  • assumption making and self-orientation are additional strategies used to understand information gathered;
  • no significant correlation between risk identification performance and age, experience, or role could be found;
  • level of educational attainment seems to contribute to a better risk identification performance measure;
  • at initial view, manager tend to think in terms of impact rather than probability;
  • the transition from risk perception to risk identification is guided in part by the needs for information, acquisition of information at varying levels and information use.

This paper has provided some insights and better understanding into the way that construction project managers identify risk. The method has allowed us to capture the initial thoughts and process. The findings indicate that the preconceptions about the importance of age and experience of actors in the initial phase of risk identification may well be misguided. The strategy of information search, however, is strongly linked to the risk identification performance and this would be an interesting area of further research. In addition, the correlation with the risk propensity scores will provide further in-depth information about risk identification performance.

The research has taken us a step further towards achieving the research aim of suggesting improved training techniques for risk identification and provide recommendations for firms in knowledge management and risk.

Adams, J. (1995). Risk. London: UCL Press.

Akintoye, A.S. & MacLeod, M.J. (1997). Risk analysis and management in construction. International Journal of Project Management 15(1): 31-38.

Ayas, K. & Zeniuk, N. (2001). Project-based learning: Building communities of reflective practitioners. Management Learning 32: 61-76.

Chapman, C & Ward, S. (2002). Project risk management (2nd ed.). Chichester: Wiley.

Chapman, R. (1998). The effectiveness of working group risk identification and assessment techniques. International Journal of Project Management 16(6): 333-343.

Chapman, R. (2001). The controlling influences of effective risk identification and assessment for construction design management. International Journal of Project Management 19: 147-160.

Clarke, S., Dalton, M. Maytorena, E. Kiely, T. & Winch, G. (2004). Development of a risk propensity measure to predict the decision-making behavior of construction project managers. Working paper, Manchester Centre for Civil and Construction Engineering, UMIST.

Eden, C. (1989). Using cognitive mapping for strategic options development and analysis (SODA). In Rosenhead, J. (ed) Rational analysis for a problematic world. West Sussex: John Wiley and Sons.

Eden, C. Williams, T. Ackerman, F. & Howick, S. (2000). The role of feedback dynamics in disruption and delay (D&D) in major projects. Journal of the Operational Research Society 51: 291:300.

Eden, C., Ackerman, F. & Cropper, S. (1990). Cognitive mapping -a user's guide. Working paper 90/2. Department of Management Science, University of Strathclyde, Glasgow.

Edkins, A. & Millán, G. (2003). Construction risk identification: A review of the literature on cognitive understanding of risk perception. Bartlett Research Paper No 16. London: Bartlett School of Graduate Studies, UCL.

Edkins, A., Millán, G. & Maytorena, E. (2003). Construction risk identification: a methodological proposal. Bartlett Research Paper No 17. London: Bartlett School of Graduate Studies, UCL.

Egbu, C. & Serafinska, Z. (2000). Attitudes to risk management in diverse project environments. COBRA conference, University of Greenwich, RICS research foundation.

Engländer, T. & Tyszka, T. (1980). Information seeking strategies in an open decision situation. Acta Psychologica 56: 113-123.

Ericsson, K.A. & Simon, H.A. (1993). Protocol analysis, verbal reports as data. USA: MIT Press.

Fiol, C.M. & Huff, A.S. (1992). Maps for managers: where are we? Where do we go from here? Journal of Management Studies 29(3):267-85.

Gigerenzer, G., Todd P. & The ABC Research Group (Eds.) (1999). Simple heuristics that make us smart. New York : Oxford University Press.

Gilovich,T., Griffin, D. & Kahneman, D. (eds) (2002). Heuristics and biases: the psychology of intuitive judgement. Cambridge: Cambridge University Press.

Hartman, F. (1997). Proactive risk management - myth or reality? In Kähkönen, K. and Artto, K. (ed) Managing Risks in Projects. London: E&FN Spon.

Hillson, D. (2002). The Risk Breakdown Structure (RBS) as an aid to effective risk management. Fifth European Project Management Conference. Cannes, France.

Hillson, D. (2003). Effective opportunity management for projects: Exploiting positive risk New York, Marcel-Dekker.

Huber, O., Wider, R. & Huber O.W. (1997). Active information search and complete information presentation in naturalistic risky decision tasks. Acta Psychologica 95: 15-29.

Kahneman, D. & Tversky A. (1979). Prospect theory: and analysis of decision under risk. Econometrica 47:263-291.

Kahneman, D., Slovic P. & Tversky A. (Eds.) (1982). Judgement under uncertainty: heuristics and biases Cambridge: Cambridge University Press.

Lewin, C. (2002). RAMP: Risk analysis and management for projects. London: Thomas Telford.

Maytorena E., Clarke, S., Dalton, M., Kiely, T. & Winch, G. (2004). Active information search and cognitive mapping to study the risk identification process in construction. Working paper. Manchester Centre for Civil and Construction Engineering. UMIST.

Mustafa, M.A. & Al-Bahar, J.F. (1991). Project risk assessment using the analytical hierarchy process. IEE Transactions of Engineering Management 38: 46-52.

Newell, A. & Simon, H. (1972). Human problem solving. Englewood Cliffs, N.J. USA: Prentice Hall.

Nicholson, N., Fenton-O’Creevy, M., Soane, E., & Willman, P. (2001). Risk propensity and personality, London Business School.

Pablo, A.L. (1997). Reconciling predictions of decision making under risk: Insights from a reconceptualized model of risk behavior. Journal of Managerial Psychology 12(1): 4-20.

Payne, J. (1976). Task complexity and contingent processing in decision making: An information search an protocol analysis. Organizational Behavior and Human Performance 16: 366-387.

Perry, J.G. & Hayes, R.W. (1985). Risk and its management in construction projects. Proceedings of Institution of Civil Engineers. ICE.

Project Management Institute. (2001). A guide to the project management body of knowledge. PMI.

Ranyard, R., Williamson, J., Cuthbert, L. & Hill, E. (1999). Inconsistent decisions involving risk and uncertainty: tracing people's thoughts. EPSRC research program on Risk and Human Behavior. 5: 7-11.

Rosenhead, J. & Mingers, J. (eds.) (2001). Rational analysis for a problematic world revisited: problems structuring methods for complexity, uncertainty and conflict (2nd ed). Chichester: Wiley.

Royer, P.S. (2000). Risk management: the undiscovered dimension or project management. Project Management Journal 13(1).

Schoemaker, J. (1982). The expected utility model: Its variants, purposes, evidence and limitations. Journal of Economic Literature 20: 529-63.

Simon, P. Hillson, D. & Newland, K. (1997). PRAM: Project Risk Analysis and Management Guide. Norwich, Association for Project Management.

Sitkin, S. B., & Pablo, A. L. (1992). Reconceptualising the determinants of risk behavior. Academy of Management Review 17(1): 9-38.

Stetzler, A. & Hofmann, D.A. (1996). Risk compensation: Implications for safety interventions. Organizational Behavior and Human Decision Processes 66: 73-88.

Stinchcombe, A. L. & Heimer, C. A. (eds.) (1985). Organization theory and project management: Administering uncertainty in Norwegian offshore oil. Oslo: Norwegian University Press.

Wilkinson, A., Elahi, S. & Eidinow, E. (2003). Risk world scenarios. Journal of risk research 6(4-6): 297-334.

Williams, T.M. Ackerman, F. & Eden, C. (1997). Project risk: Systemicity, cause mapping and scenario approach. Kähkönen, K. Artto, K.A. (ed) Managing Risks in Projects. London: E&FN Spon.

Williams, T.M. Ackerman, F. & Tait, A. (1995). The effects of design changes and delays on project costs. Journal of the Operational Research Society 46: 809-818.

Williamson, J. & Ranyard, R. (2000). A conversation-based process tracing method for use with naturalistic decisions: an evaluation study. British Journal of Psychology 91: 203-221.

Winch, G. M. (2002). Managing construction projects: An information processing approach. Oxford: Blackwell

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI or any listed author.

Advertisement

Advertisement

Related Content

Advertisement