Writing a critical review of two journal articles about project risk Management

Use and bene®ts of tools for project risk management

T. Raz a,*, E. Michael b

aFaculty of Management, Tel Aviv University, Faculty of Management, Ramat Aviv, 69978 Tel Aviv, Israel bComverse Network Systems, Tel Aviv, Israel

Received 11 January 1999; received in revised form 12 April 1999; accepted 12 May 1999

Abstract

Risk management is one of the key project management processes. Numerous tools are available to support the various phases of the risk management process. We present the results of a study designed to identify the tools that are most widely used and those that are associated with successful project management in general, and with e�ective project risk management in particular. The study is based on a questionnaire administered to a sample of project managers from the software and high-tech industries. The

response data was analyzed in order to ®nd which tools are more likely to be used in those organizations that report better project management performance and in those that value the contribution of risk management processes. # 2000 Elsevier Science Ltd and IPMA. All rights reserved.

Keywords: Project risk management; Risk management process

1. Introduction

The management of risk in projects is currently one of the main topics of interest for researchers and practi- tioners working in the area of project management. A recent survey of research on the topic by Williams [1] includes 241 references. Risk management has been designated as one of the eight main areas of the Project Management Body of Knowledge (PMBOK) by the Project Management Institute, which is the largest pro- fessional organization dedicated to the project manage- ment ®eld. Further, most training programs for project managers include a course on risk management. Within the currently accepted view of project management as a life cycle process, project risk management (PRM) is also seen as a process that accompanies the project from its de®nition through its planning, execution and con- trol phases up to its completion and closure. A number of variations of the PRM process have

been proposed. Boehm [2] suggested a process consist- ing of two main phases: risk assessment, which includes

identi®cation, analysis and prioritization, and risk con- trol, which includes risk management planning, risk resolution and risk monitoring planning, tracking and corrective action. Fairley [3] talks about seven steps: (1) Identify risk factors; (2) Assess risk probabilities and e�ects; (3) Develop strategies to mitigate identi®ed risks; (4) Monitor risk factors; (5) Invoke a contingency plan; (6) Manage the crisis; (7) Recover from the crisis. The Software Engineering Institute [4], a leading

source of methodologies for managing software devel- opment projects, looks at project risk management as consisting of ®ve distinct phases (identi®cation; analysis; response planning; tracking and control) linked by an ongoing risk communications e�ort. In its Guide to the Project Management Body of Knowledge, the Project Management Institute [5] presents four phases of the PRM process: identi®cation; quanti®cation; response development and control. Kliem and Ludin [6] describe a four-phase process

(identi®cation, analysis, control and reporting) that parallels Deming's four steps for quality management (plan, do, check and act). Chapman and Ward [7] out- line a generic PRM process consisting of nine phases: de®ne the key aspects of the project; focus on a strategic approach to risk management; identify where risks might arise; structure the information about risk

0263-7863/00/$20.00 # 2000 Elsevier Science Ltd and IPMA. All rights reserved. PII: S0263-7863(99)00036-8

International Journal of Project Management 19 (2001) 9±17

www.elsevier.com/locate/ijproman

* Corresponding author. Tel.: +972-3-640-8722; fax: +972-3-640-

9560.

E-mail address: [email protected] (T. Raz).

assumptions and relationships; assign ownership of risks and responses; estimate the extent of uncertainty; evaluate the relative magnitude of the various risks; plan responses and manage by monitoring and control- ling execution. It is evident from this brief review of representative PRM processes that there is general agreement regarding what is included in the process, with the di�erences depending on variations in the level of detail and on the assignment of activities to steps and phases. Of course, any PRM process requires tools for its

implementation. The adoption of analysis, planning, control, or management tools involves a certain invest- ment, which in certain cases may be quite signi®cant. This cost represents the e�ort required, both at a per- sonal and at the organizational level, to understand and to learn how to use the tool, and to acquire the necessary infrastructure (technical expertise, computing aids, databases, operating procedures, etc.). A ques- tion of major relevance to any individual or organiza- tion considering the adoption or improvement of a PRM process is: which tools can provide the greatest bene®ts? In this paper we present the results of a study

designed to answer this question. The approach taken consisted of surveying a sample of project managers in order to ®nd out which tools are widely used, which tools are associated with successful project management in general, and with e�ective project risk management in particular. In this context, the term `tool' is given a wide meaning, including not only special purpose tools, but also practices and processes that are likely to con- tribute to the management of risks in projects. In the next section we describe the methodology of

the survey, including the selection of the sample and the design of the questionnaire. This is followed by a sta- tistical analysis of the results, and by a discussion of the practical implications of the results. We conclude with some observations regarding the way the research was carried out.

2. Methodology

The data for this study was obtained by means of a questionnaire. The questionnaire, which was written in Hebrew, was distributed either personally or via e-mail to a random sample of about 400 project managers from the software and high-tech sectors in Israel during April through June 1998. At the end of the survey per- iod there were 84 usable completed questionnaires. The questionnaire consisted of three main sections,

each containing a number of brief questions to be answered on a 0±5 scale. The ®rst section dealt with the extent of the contribution of individual PRM tools to the project success in general. The objective here was to

identify the tools that were perceived as being the most valuable by the respondents. The second section of the questionnaire dealt with the

e�ectiveness and e�ciency of the manner in which pro- jects are managed in the respondent's organization. With these questions we sought to investigate whether there is a relationship between the use of PRM tools and the level of performance of the project management process. The third section addressed the contribution of a risk

management process to overall project success. In par- ticular, we wished to learn about the di�erences in PRM tool usage between those project managers who believe that risk management is a valuable process, and those who do not. The speci®c questions that were included in each of the three sections along with analysis of the responses are presented next.

3. Individual tool contribution

This section of the questionnaire consisted of a list of 38 tools and practices mentioned in the literature as contributing to project risk management. The list was developed as follows. First, an initial compilation of over 100 tools was drawn from the literature. This list was presented to a panel of ®ve individuals who had responsibility for PRM in their organizations: three major software development companies, a company engaged in the development and manufacture of com- munications hardware, and a manufacturer of chemical products. These ®ve individuals eliminated from the lists duplicated tools, combined related tools into a single one, eliminated tools that were not applied in practice and added related tools that were missing from the list, yielding the ®nal list of 38 tools. It is interesting to note that certain tools that are normally associated with risk management, such as decision trees, fault tree analysis and in¯uence diagrams, were reported to be seldom or not used at all, and consequently are not included in the ®nal list. The tools were grouped according to the ®ve stages of

the Software Engineering Institute Risk Management process [4]. We added a sixth group for tools, processes and practices of a general nature. The tools in this group, which we called `Background', are likely to a�ect the manner in which risks are managed without being speci®cally related to one of the ®ve stages in the PRM cycle. The respondents were asked to rate the contribution

of each tool to the project risk management process by indicating a value between 0 (no contribution at all) to 5 (critical contribution). The results for the 38 tools, divi- ded into the six groups, appear in Table 1, along with the mean and standard deviation of the responses. Bib- liographical references mentioning the tools in the risk

10 T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17

management context are noted next to each tool description. The tools without references were added to the list based on suggestions from the panel. The right- most column shows the ranking of the tools sorted in descending order of their mean score. It is interesting to note that out of the ten tools that

received the highest mean score, ®ve belong to the `Background' group, and that all the tools in this group received scores above the overall average of 2.72. This ®nding indicates that risk management is tightly related to other management practices, such as requirements management, subcontractor management, con®guration control, and that the contribution of these types of organization-wide processes to e�ective project man- agement is well recognized. Also, it appears that simu-

lation and prototyping are development practices that are associated with risk management. The tools in the risk control group are perceived as

low contributors. There are two possible explanations for this ®nding. One is that currently there are no e�ec- tive tools for risk control, and that the tools o�ered in the literature are not perceived to be adequate. The other explanation is related to management culture. Project managers might be willing to invest time and e�ort in the earlier phases of risk management, which are carried out in conjunction with other project plan- ning activities. However, during the execution of the project they become busier and are subject to mounting resource and time pressures, and are likely to neglect the risk control phase. Consequently, risk control tools are

Table 1

Descriptive statistics for the PRM tools

Tool Description Group Mean Standard deviation Ranking

T1 Checklists[8] Identi®cation 2.20 0.98 36

T2 Brainstorming[9,10] Identi®cation 3.74 0.79 8

T3 Risk documentation form[4] Identi®cation 2.65 0.94 29

T4 Periodic risk reporting[4] Identi®cation 2.88 1.12 24

T5 Risk probability assessment[2,11,4,8,12] Analysis 3.57 1.01 14

T6 Risk impact assessment[2,11,4,8,12] Analysis 3.86 0.68 3

T7 Risk time frame assessment[4] Analysis 2.58 1.25 30

T8 Risk classi®cation[13,4,14,15] Analysis 2.38 1.17 33

T9 Ranking of risks[2,11,4,8,12] Analysis 3.29 0.91 17

T10 Graphic presentation of risk information Analysis 1.82 1.24 38

T11 Responsibility assignment[4] Planning 3.99 0.84 2

T12 Planning for risk mitigation[4,14] Planning 3.61 0.99 12

T13 Time-limited action-item lists Planning 3.70 0.82 9

T14 Cost-bene®t assessment during risk planning[10,16] Planning 2.69 1.08 26

T15 Cause and e�ect analysis during risk planning[9,10,17] Planning 2.33 1.14 34

T16 Project replanning for risk mitigation[4] Planning 3.17 1.05 21

T17 Revision of risk assessments[4] Tracking 3.30 0.77 16

T18 Periodic document reviews[4,8] Tracking 3.18 1.01 20

T19 Periodic risk status reporting[4] Tracking 3.20 1.05 19

T20 Periodic reporting of risk mitigation plans[4] Tracking 2.80 0.93 25

T21 Periodic trend reporting[4] Tracking 2.58 1.01 31

T22 Critical risk reporting to senior management[4] Tracking 3.75 0.99 6

T23 Analysis of trends, deviations and exceptions[4] Control 2.69 0.93 27

T24 Project replanning[4] Control 2.94 1.02 23

T25 Procedure for closing risks[4] Control 2.20 1.27 37

T26 Contingency plans for risk mitigation failure[4] Control 2.43 1.14 32

T27 Cost-bene®t analysis during risk control[10,16] Control 2.68 1.10 28

T28 Cause and e�ect analysis during risk control[9,10,17] Control 2.27 0.99 35

T29 Prototyping Background 3.75 1.03 7

T30 Simulation Background 4.00 0.85 1

T31 Benchmarking Background 3.58 0.87 13

T32 Requirements management[18] Background 3.69 0.89 10

T33 Subcontractor management[18] Background 3.77 0.96 5

T34 Con®guration control[18] Background 3.81 0.70 4

T35 Quality control[18] Background 3.69 0.76 11

T36 Quality management[18] Background 3.39 0.76 15

T37 Training programs[18] Background 3.11 0.82 22

T38 Customer satisfaction surveys Background 3.27 0.96 18

Average across all tools 2.72 0.64

T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17 11

used sporadically or not at all, and their contribution is rated as low.

4. Project management performance

This part consisted of six questions designed to represent the manner in which projects were managed at the respondent's organization. We addressed two com- plementary aspects of project management: e�ciency and e�ectiveness. By e�ciency we meant that the pro- ject is managed without unnecessary expenditure of e�ort, which we attempted to measure by considering the extent of project plan changes, emergency meetings and e�ort invested. By e�ectiveness we meant meeting the expectations of those involved in the project. Here we asked to rate the satisfaction of the project partici-

pants and of the customer, as well as the quality of the project output, as measured by post-delivery rework. The respondents were asked to indicate the extent to

which the projects they manage exhibit the character- istics listed in Table 2, ranging from very low (0) to very high (5). The responses to questions 1, 2, 3 and 6 were recoded so that a higher value indicated a more favor- able situation. The mean and standard deviation of the (recoded) questions appear in Table 2. We de®ned the Project Management Performance (PMP) index as the average of the scores over the six items. Fig. 1 shows a plot of the PMP index against total

tool score. Each observation corresponds to one respondent. In some cases, two or more observations coincided on the chart and appear as a single point. We can see that there is a positive relationship between the extent of tool use and the PMP score. Indeed, the coef-

Table 2

Descriptive statistics for the project management performance questions

Question: To what extent do the projects you manage exhibit the following characteristics? Mean Standard deviation

1. Extent and frequency of plan changes 1.85 1.02

2. Frequency of emergency meetings 2.85 1.18

3. Agreement between e�ort invested and e�ort required 2.04 0.80

4. Participant satisfaction 3.30 0.85

5. Customer satisfaction 3.60 0.75

6. Number of post-delivery product changes 2.71 1.02

Project Management Performance (PMP) index 2.72 0.64

Fig. 1. Project Management Performance index against total tool score.

12 T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17

®cient of correlation was 0.36 at a level of signi®cance of 0.0008. This suggests that respondents who come from environments conducive to better project management performance are more likely to assign higher scores to project risk management tools. To further explore the relationship between individual

tool scores and the PMP score, we divided the 84 respondents into two sets, based on their PMP index value. The threshold value was 3, resulting in two sets of almost equal size: a LOW set (43 respondents with PMP below 3) and a HIGH set (41 respondents with PMP above 3). Table 3 shows, for each tool, the average score for each of these two sets of respondents. The Student t test was used to test the null hypothesis that the average scores in the populations represented by the two sets of respondents are equal. The rightmost col- umn shows the level of statistical signi®cance. Using

0.05 as the threshold for rejecting the null hypothesis, we highlighted with a star (*) the tools in which there is a statistically signi®cant di�erence in score between the LOW and the HIGH populations. There were four- teen such tools, and in all cases the average tool score for the HIGH population was higher than for the LOW population. Six of the fourteen tools belong to the Background

group. This ®nding suggests that there is a link between the organization-wide application of practices such as Simulation (T30) and processes such as Subcontractor Management (T33), Quality Management (T36), Train- ing (T37), Customer Satisfaction Surveys (T38), and an e�ective and e�cient environment for managing projects. Four tools Ð Risk Probability Assessment (T5), Risk

Impact Assessment (T6), Risk Classi®cation (T8) and

Table 3

Relationship between tool score and PMP index

PMP low PMP high

Tool (n=43) (n=41) T-test signi®cance

T1 Checklists 2.28 2.12 0.466

T2 Brainstorming 3.70 3.78 0.635

T3 Risk documentation form 2.51 2.80 0.153

T4 Periodic risk reporting 2.72 3.05 0.185

T5 Risk probability assessment 3.21 3.95 0.001*

T6 Risk impact assessment 3.65 4.07 0.004*

T7 Risk time frame assessment 2.33 2.85 0.053

T8 Risk classi®cation 2.00 2.78 0.002*

T9 Ranking of risks 3.07 3.51 0.026*

T10 Graphic presentation of risk information 1.77 1.88 0.686

T11 Responsibility assignment 3.63 4.37 0.000*

T12 Planning for risk mitigation 3.44 3.78 0.118

T13 Time-limited action-item lists 3.79 3.61 0.301

T14 Cost-bene®t analysis during risk planning 2.51 2.88 0.117

T15 Cause and e�ect analysis during risk planning 2.21 2.46 0.310

T16 Project replanning for risk mitigation 3.12 3.22 0.656

T17 Revision of risk assessments 3.14 3.46 0.054

T18 Periodic document reviews 2.86 3.51 0.003*

T19 Periodic risk status reporting 3.14 3.27 0.578

T20 Periodic reporting of risk mitigation plans 2.60 3.00 0.051

T21 Periodic trend reporting 2.35 2.83 0.029*

T22 Critical risk reporting to senior management 3.72 3.78 0.785

T23 Analysis of trends, deviations and exceptions 2.44 2.95 0.011*

T24 Project replanning 2.93 2.95 0.925

T25 Procedure for closing risks 2.09 2.32 0.421

T26 Contingency plans for risk mitigation failure 2.28 2.59 0.221

T27 Cost-bene®t analysis during risk control 2.47 2.90 0.067

T28 Cause and e�ect analysis during risk control 2.07 2.49 0.051

T29 Prototyping 3.91 3.59 0.156

T30 Simulation 3.74 4.27 0.004*

T31 Benchmarking 3.42 3.76 0.073

T32 Requirements management 3.58 3.80 0.254

T33 Subcontractor management 3.56 4.00 0.035*

T34 Con®guration control 3.74 3.88 0.389

T35 Quality control 3.67 3.71 0.844

T36 Quality management 3.21 3.59 0.023*

T37 Training programs 2.86 3.37 0.004*

T38 Customer satisfaction surveys 3.07 3.49 0.045*

T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17 13

Ranking of Risks (T9) Ð belong to the Analysis group, suggesting that more extensive use of risk analysis tools is associated with better project management practice. The other tools where there was a marked di�erence between the HIGH and LOW populations are Respon- sibility Assignment (T11), Periodic Document Reviews (T18), Periodic Trend Reporting (T21), and Analysis of Trends, Deviations and Exceptions (T23). Perhaps the common side of these tools is that they provide some structure and periodic discipline to the risk management e�ort, and that they can be implemented in an explicit and systematic manner. It is also interesting to note that there was no sig-

ni®cant di�erence in the scores given to the tools in the Identi®cation group, suggesting that risk identi®cation is carried out in basically the same manner in all organi- zations. Overall it appears that organizations that man- age their projects more e�ciently and more e�ectively tend to attach more value to risk analysis tools, tools

that provide structure and discipline, and organization- wide processes associated with quality processes and practices.

5. Contribution of PRM process.

This part of the questionnaire was designed to ®nd out the extent to which the existence of a process for managing risks was perceived to contribute to various aspects of project success. The premise behind this part of the analysis is that those respondents who perceive the contribution of PRM to project success to be higher are more likely to adopt and implement some type of risk management process than those who gave a low score. One of the reasons we included this part is that we

met many project managers who claimed that risk management was an unnecessary activity, and that the

Table 4

Descriptive statistics for the Risk Management Contribution questions

Question: To what extent does the Risk Management process contribute to the following? Mean Standard deviation

1. Overall project success 3.94 0.57

2. Meeting project schedule 3.76 0.74

3. Meeting project budget 3.58 0.78

4. Meeting planned objectives 3.74 0.70

5. Achieving customer satisfaction 3.36 0.93

6. Success of other projects in your organization 2.89 0.99

Risk Management Contribution index 3.55 0.51

Fig. 2. Risk Management Contribution index against total tool score..

14 T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17

resources it required could be put to better use else- where in the project. The questions in this part included, in addition to the four commonly accepted project suc- cess criteria (schedule, budget, objectives and customer satisfaction), two general items: overall project success and contribution to the success of other projects in the organization. Here we also de®ned a composite index, Risk Management Contribution (RMC), as the average of the responses to these six questions. The results for this part of the questionnaire appear in Table 4. Fig. 2 shows a plot of the RMC index against total

tool score. Each observation corresponds to one respondent. Here again, some points on the chart represent more than one observation. As anticipated, there is a strong positive relationship, with the coe�- cient of correlation equal to 0.67 at a level of sig- ni®cance of 0.0001. This means that those respondents

who believe in the value of the risk management process as a contributor to overall project success are more likely to apply the tools in the list. To further explore the relationship between individual

tool scores and the perceived contribution of the risk management process, we divided the 84 respondents into two sets, based on their RMC index value. The threshold value was 3.5, resulting in two sets, a LOW and a HIGH set, both of equal size (42). Table 5 shows, for each tool, the average score for each of these two groups of respondents. Here again, the Student t test was used to test, at the 0.05 level of signi®cance, the null hypothesis that the average scores in the populations represented by the two sets are equal. Tools in which there is a di�erence in score between the LOW and the HIGH populations are highlighted. Statistically sig- ni®cant di�erences were found in 24 of the 38 tools, and

Table 5

Relationship between tool score and RMC index

Tool RMC low RMC high T-test signi®cance

T1 Checklists 1.83 2.57 0.000*

T2 Brainstorming 3.57 3.90 0.054

T3 Risk documentation form 2.50 2.81 0.131

T4 Periodic risk reporting 2.79 2.98 0.441

T5 Risk probability assessment 3.43 3.71 0.196

T6 Risk impact assessment 3.69 4.02 0.024*

T7 Risk time frame assessment 1.98 3.19 0.000*

T8 Risk classi®cation 1.81 2.95 0.000*

T9 Ranking of risks 3.05 3.52 0.016*

T10 Graphic presentation of risk information 1.62 2.02 0.137

T11 Responsibility assignment 3.98 4.00 0.898

T12 Planning for risk mitigation 3.52 3.69 0.446

T13 Time-limited action-item lists 3.74 3.67 0.692

T14 Cost-bene®t analysis during risk planning 2.26 3.12 0.000*

T15 Cause and e�ect analysis during risk planning 2.02 2.64 0.012*

T16 Project replanning for risk mitigation 2.93 3.41 0.037*

T17 Revision of risk assessments 3.12 3.48 0.033*

T18 Periodic document reviews 2.81 3.55 0.001*

T19 Periodic risk status reporting 3.02 3.38 0.120

T20 Periodic reporting of risk mitigation plans 2.57 3.02 0.025*

T21 Periodic trend reporting 2.31 2.86 0.012*

T22 Critical risk reporting to senior management 3.50 4.00 0.020*

T23 Analysis of trends, deviations and exceptions 2.17 3.21 0.000*

T24 Project replanning 2.71 3.17 0.042*

T25 Procedure for closing risks 1.74 2.67 0.001*

T26 Contingency plans for risk mitigation failure 2.12 2.74 0.012*

T27 Cost-bene®t analysis during risk control 2.33 3.02 0.003*

T28 Cause and e�ect analysis during risk control 2.00 2.55 0.010*

T29 Prototyping 3.76 3.74 0.916

T30 Simulation 4.02 3.98 0.799

T31 Benchmarking 3.43 3.74 0.102

T32 Requirements management 3.60 3.79 0.331

T33 Subcontractor management 3.45 4.10 0.002*

T34 Con®guration control 3.64 3.98 0.029*

T35 Quality control 3.52 3.86 0.044*

T36 Quality management 3.19 3.60 0.014*

T37 Training programs 2.81 3.40 0.001*

T38 Customer satisfaction surveys 3.00 3.55 0.008*

T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17 15

in all cases the average group for the HIGH contribu- tion population was higher than for the LOW con- tribution population. It is worthwhile to note that, similarly to the results

reported in the previous section, there were no di�er- ences in the use of the tools in the Identi®cation group, with one marked exception Ð Checklists. However, there were signi®cant di�erences in four of the six tools in the Analysis group, ®ve of the six tools in the Track- ing group and in all six tools of the Control group. This ®nding is even more striking if we recall that the tools in these groups had the lowest average scores. Our inter- pretation is as follows. The use of risk identi®cation tools is about the same,

whether a risk management process is in place or not. However, those organizations who believe in a process (and, presumably, have adopted one) are more likely to apply tools that assist them in the weakest areas, namely analysis, tracking and control of risks. In other words, it is relatively easy to identify risks, and you may not need a process for that purpose, but you do need one for the more complicated tasks of to analyzing, tracking and controlling the project risks.

As for the Planning group, we notice that there are signi®cant di�erences in three of the less commonly used tools: Cost±Bene®t Analysis (T14), Cause and E�ect Analysis (T15) and Project Replanning for Risk Miti- gation (T16). Here again, it seems that project managers who value the contribution of a project risk manage- ment process are more likely to apply formal risk plan- ning tools that those who do not. We also note that six of the ten tools in the Background group exhibited sig- ni®cant di�erences in use between the two populations, including tools T35 through T38, which received a rela- tively low overall ranking. This ®nding suggests that the existence of organization-wide processes such as those in this group is associated with a higher rate of adoption of a process approach to risk management.

6. Practical implications

The objective of this study was to identify the tools that have the greatest potential for contribution to a project risk management process. The results of the survey provide important guidance with respect to this issue. First, any one facing the design of a risk management process should ®rst consider the tools that are most commonly used. These are obtained from the ranking in Table 1. Adoption of these tools moves the organization towards the current state of the practice in the ®eld. Beyond that, in order to strive for some type of com-

petitive advantage, one should adopt the tools that are associated with better performing project management practices (PMP=HIGH), or that are used by practi- tioners who already have a good risk management pro- cess (RMC=HIGH). These tools are listed in Table 6. Obviously, there are costs and lead times associated with the adoption of tools and practices, and these, as well as other factors should be taken into consideration. These two types of results, the ranking in Table 1 and

the selected tools in Table 6, can also serve for diag- nostic purposes. By comparing the current practice with these two lists, it is possible to identify areas of insu�- cient coverage and to devise a course of action that will enhance the project risk management capabilities in the organization.

7. Concluding remarks

The approach taken in this study is similar to the benchmarking methodology developed by Camp [19], Balm [20] and others: we looked at current practice in the more successful organizations and attempted to identify what it is that they do that others do not. This approach can be extended to other aspects of the risk management process, such as assignment of roles and responsibilities, timing and frequency of risk manage-

Table 6

Tools with signi®cantly higher use in the HIGH populations

Tool Description

Signi®cantly higher use in RMC=HIGH and in PMP=HIGH

T6 Risk impact assessment

T8 Risk classi®cation

T9 Ranking of risks

T18 Periodic document reviews

T21 Periodic trend reporting

T23 Analysis of trends, deviations and exceptions

T33 Subcontractor management

T36 Quality management

T37 Training programs

T38 Customer satisfaction surveys

Signi®cantly higher use in PMP=HIGH

T5 Risk probability assessment

T11 Responsibility assignment

T30 Simulation

Signi®cantly higher use in RMC=HIGH

T1 Checklists

T7 Risk time frame assessment

T14 Cost-bene®t analysis during risk planning

T15 Cause and e�ect analysis during risk planning

T16 Project replanning for risk mitigation

T17 Revision of risk assessments

T20 Periodic reporting of risk mitigation plans

T22 Critical risk reporting to senior management

T24 Project replanning

T25 Procedure for closing risks

T26 Contingency plans for risk mitigation failure

T27 Cost-bene®t analysis during risk control

T28 Cause and e�ect analysis during risk control

T34 Con®guration control

T35 Quality control

16 T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17

ment activities, and amount of e�ort applied. It could also be applied to other processes related to project management, and, as the benchmarking proponents advocate, to many aspects of the organization. Our speci®c methodology was based on a ques-

tionnaire which required a self-assessment of the e�- ciency of the project management practices in the organization and of the contribution of the risk man- agement process. Although the possibility of respondent bias exists, we feel that by setting threshold values rela- tive to the sample rather than in absolute terms, we overcome most of this di�culty. Our survey was limited to the software development

and high-tech industrial sectors. This was done because we suspected that in these sectors, which are character- ized by very short opportunity windows, high uncer- tainty, and rapid decision-making, the need for tools and processes to manage risks ought to be greater than in other areas, such as construction, heavy manufactur- ing or services. Further, our review of the literature revealed that a substantial part of the tools and meth- odologies used in PRM were developed for these two speci®c areas. Another limitation of this research stems from the

unique characteristics of Israeli culture, which places a high value on personal initiative, improvisation and on- the-spot problem-solving, while giving less emphasis to disciplined work processes. It will be interesting to see if there are di�erences across various industrial sectors and di�erent organizational cultures. Our impression, based on personal contacts with numerous project managers in di�erent countries over the last several years, is that indeed this will be the case.

References

[1] Williams TM. A classi®ed bibliography of recent research relat-

ing to project risk management. European Journal of Opera-

tional Research 1995;85:18±38.

[2] Boehm BW. Software risk management: principles and practices.

IEEE Software 1991;8:32±41.

[3] Fairley R. Risk Management for Software Projects. IEEE Soft-

ware 1994;57±67.

[4] Dorofee AJ, Walker JA, Alberts CJ, Higuera RP, Murphy RL,

Williams RC. Continuous Risk Management Guidebook. Pitts-

burgh: Carnegie Mellon University, 1996.

[5] PMI. A Guide to the Project Management Body Of Knowledge.

Project Management Institute, 130 South State Road, Upper

Darby, 1996.

[6] Kliem RL, Ludin IS. Reducing Project Risk. Gower, 1997.

[7] Chapman C, Ward S. Project Risk Management: Processes,

Techniques and Insights. John Wiley, 1997.

[8] Down A, Coleman M, Absolon P. Risk Management for Soft-

ware Projects. McGraw-Hill, 1997.

[9] Lumsdaine E, Lumsdaine M. Creative Problem Solving. New

York: McGraw-Hill, 1990.

[10] Xerox Corporation and Carnegie Mellon University. The Uni-

versity Challenge: Problem Solving Process User Manual. Stam-

ford, CT: Xerox Corporation, 1992.

[11] Sisti FJ, Sujoe J. Software Risk Evaluation Method. Technical

Report Ð CMU/SEI-94-TR-019, Software Engineering Institute,

Carnegie Mellon University, Pittsburgh, Pennsylvania, 1994.

[12] Air Force Systems Command/Air Force Logistics Command.

Pamphlet 800±45, Software Risk Abatement, 1988.

[13] Carr MJ, Konda SL, Monarch, I, Ulrich FC. Taxonomy-Based

Risk Identi®cation. Technical Report Ð CMU/SEI-93-TR-183,

Software Engineering Institute, Carnegie Mellon University,

Pittsburgh, Pennsylvania, 1993.

[14] Brassard M. The Memory Jogger: Featuring the Seven Manage-

ment and Planning Tools. MA: Methuen, 1989 (GOAL/QPC).

[15] Brassard M, Ritter D. The Memory Jogger II: A Pocket Guide of

Tools for Continuous Improvement and E�ective Planning. MA:

Methuen 1994 (GOAL/QPC).

[16] Arrow KJ. Behavior under uncertainty and its implications for

policy. In: Decision Making: Descriptive, Normative, and Pre-

scriptive Interactions. Cambridge: Cambridge University Press,

1988. p. 497±507.

[17] Scholtes PR. The Team Handbook: How to Use Teams to

Improve Quality. Madison, WI: Joiner Associates, Inc, 1988.

[18] Paulk MC, Curtis W, Chrissis MB, Weber CV. Capability Matur-

ity Model for Software, Version 1.1. Software Engineering Insti-

tute, Carnegie Mellon University, Pittsburgh, Pennsylvania, 1996.

[19] Camp RC. Benchmarking Ð the search for industry best practices

that lead to superior performance. ASQC Quality Press, 1989.

[20] Balm GJ. Benchmarking: A Practitioner's Guide for becoming

and Staying Best of the Best. QPMA Press, 1992.

T. Raz, E. Michael / International Journal of Project Management 19 (2001) 9±17 17