Too many cooks spoil the broth, or maybe not?

Abstract

Decision making is an integral part of project management. Every key person in any organization faces situations on a daily basis where a decision should be made for resolving problems. Although every problem has multiple solutions, many restrictions and various ways to be addressed, there are also many techniques, methods and approaches proposed by scholars and practitioners that can be implemented to help increase our decision making effectiveness. The Forum will address this issue by engaging participants to reach a group decision using Group Analytic Network Process (GANP). As all participants should be equally familiar, the illustrative case that will be used during the Forum will be the selection of the town that will host one of the forthcoming PMI EMEA congresses. The results of this game will provide significant insights for the group decision making process, explore the proposed techniques and bring out the advantages and disadvantages reported in literature.

Introduction

Decision making is a substantial interpersonal skill that any project manager should possess. A Guide to the Project Management Body of Knowledge (PMBOK® Guide) – Fourth Edition, the most widely accepted project management standard, introduces decision making in Appendix G – Interpersonal Skills / chapter G6 (PMI, 2008, p 420). However, PMBOK® Guide refers also to decisions that have to be taken in many project management processes. Indicative examples are go/no go decisions, initiate a phase or not, make or buy, project prioritization, etc. According to PMBOK® Guide project managers and project teams use a decision making process such as the six-phase model shown in Exhibit 1.

Six-phase decision model as appears in PMBOK® Guide

Exhibit 1 - Six-phase decision model as appears in PMBOK® Guide

The present study focuses on the third step concerning the Ideas to action where projects teams need to define the criteria, rate the pros and cons of the alternatives and select the best among them. Particularly, special attention is given to the group decision making processes since most organizational decisions, especially the complex ones, are made by groups rather than individuals. It is, however, sometimes difficult to know when to involve others in the decision-making process and to what degree. Practice and research have revealed that groups outperform individuals working in isolation since groups’ solutions to problems are typically of higher quality. The advantages of group’s decisions over an individual’s have been widely studied in literature. One obvious explanation is that groups can pool information and abilities and thus gain access to a collection of knowledge that is superior to that of any single individual. This knowledge enables the group to reject incorrect approaches and provides a check on the possibility of committing errors. Being in a group also tends to motivate and inspire group members by enhancing their level of contribution. As a result, they tend to be more accepting of the decision and share the responsibility. Also those who may not have contributed still do not tend to oppose to it as “the group” has come to this decision.

However, there are also several potential disadvantages of group decision making. For example, highly cohesive groups sometimes encourage a restricted view of alternatives, known as groupthink. Groups may also polarize toward extreme points of view if an appreciable element of risk is involved. In other words, group members tend to feed off of each other's fears or enthusiasm, and can make what researchers refer to as risky shifts and cautious shifts. Another potential problem is that group decision making tends to be much more costly than individual decision-making. Given the time and energy that meetings can consume, it is usually best to reserve group decision making for more important decisions that require high-quality solutions. In addition, decision-making in groups tends to be influenced by the relative status of group members. Thus, when a group member who possesses relatively little status offers an objectively good suggestion, it may be rejected. But if the same suggestion is offered by a group member with high status, the likelihood of its being adopted is greatly increased.

Problem Statement and Proposed Approach

Group Decision Making

The difference between making a decision alone and making a group decision is large. The interaction can distill the best out of each member, creating better resonance of ideas and synthesis of view-points. This is mainly because groups can represent a larger and more diverse set of perspectives and constituencies, thus being more “fair” (Tyler & Smith, 1998). Moreover, the idea that “two heads are better than one” is widespread and typically accurate, based on the empirical record (Bonner, 2004). Especially, group decision making in projects can be an effective management tool as people tend to resist what is forced upon them and support what they help to create. On the other hand, group decision processes have proven sometimes elusive, difficult to understand and truly complex. One of the principle difficulties with making decisions in groups is deciding how to make the decision! There are several alternatives to consider as: should the leader just ask the group and make the decision? Should the leader delegate the decision to some members of the group? Should the group make the decision through some form of majority vote? Should all decisions involving the group be made by consensus?

The Vroom-Yetton leadership model (Vroom & Yetton, 1973) suggests that the decision making style should take into consideration three factors:

  • the quality (or correctness) of the decision, for those decisions that involve a high degree of expertise,
  • the required level of commitment to the decision by the group members and
  • the time available to make the decision.

There are four well-known styles for group decision making which can be used within different decision environments, encompassing specific advantages and disadvantages (Exhibit 2). When using the command style the leader makes a decision for a group with little or no input from the members of that group. The group members may provide specific information on request, but are not asked to contribute towards finding a solution. When using the consultative style the leader seeks input and advice from the group before making a decision for a group, but then makes the final decision him/herself. When using the consensus style the leader seeks input and advice from a group and works through the decision making process with the group, until every member of the group can “live with” the final decision that is made. When using the majority vote the decision is made by selecting the alternative which has the more votes from the members of the group.

Styles of group decision making

Exhibit 2 - Styles of group decision making

Multi Criteria Decision Analysis

The problem seems to be even harder when, as usual, the decision involves multiple conflicting criteria. Within the field of Operations Research many Multicriteria Decision Analysis (MCDA) methods have been developed. Some of the most popular are the Analytic Hierarchy and Network Profcess (AHP, ANP), the Outranking methods, such as the ELECTRE, Multiattibute Utility Theory (MAUT), Fuzzy sets and Mathematical Programming models. Although the number of MCDA methods is very high and still increasing, there is no specific, generally accepted, method for every problem as each problem is unique. They all require the definition of options, criteria, and most of them demand a measure for assessing the relative significance of each criterion to the other criteria (Belton & Stewart, 2002; Peniwati, 2005). They differ, however, in terms of how they combine the data (DETR, 2000). Choosing one method out of all the existing ones is itself a multicriteria task, as different aims are involved (appropriateness of the data and the structure of the problem, method applicability, acceptance of the decision etc.). In the context of the present forum the ANP will be used as it is easy to understand, takes into account both tangible and intangible criteria, it is friendly to use and matches the human thinking (Kirytopoulos, D.Voulgaridou & V. Voulgaridou, 2008). It is underlined here that the focus of the Forum is on the group decision making rather than the method itself.

The Analytic Network Process

The Analytic Network Process is a generalization of the Analytic Hierarchy Process (AHP) developed by Thomas Saaty (1996). ANP incorporates feedback and interdependent relationships among decision criteria and alternatives and provides a general framework to deal with decisions without making assumptions about the independence of higher level elements from lower level elements or the independence of the elements within a level as in AHP. In fact, the ANP uses a network of elements without need to specify levels (Saaty, 2005). Technically, the model consists of clusters and elements. As in the AHP, the dominance or relative importance of influence is the central concept and judgments are provided from the fundamental scale of the AHP (Saaty, 2005) by answering the question: Given a criterion X, which of the two elements Y,Z is more dominant with respect to that criterion? In order for the influences among the elements to be meaningful at the final stage of the method (synthesis), a specific element, each time, is used to perform the pairwise comparisons. This element is called control criterion. In short, the ANP approach handles interdependence among elements by obtaining the composite weights through the development of a ‘supermatrix’.

The group decision-making process in the ANP involves the construction of pairwise comparison matrices at each level of network either by consensus voting or by aggregating the individual preferences (Saaty, 1989). In the consensus voting approach all group members agree upon the values for each comparison judgment. If the group is unwilling or unable to vote or cannot achieve a consensus, then a compromise group solution can be obtained by combining the individual preferences into aggregated group preferences. Forman and Peniwati (1998) showed that the group prioritization methods in the ANP/AHP apply two basic techniques for aggregating the individual preferences into a group preference, depending on whether the group wants to act together as a unit or as separate individuals, and specify two aggregation approaches, the Aggregating individual judgements (AIJ) and the Aggregating individual priorities (AIP). The weighted arithmetic mean (WAM) and the geometric mean (GM) mathematical procedures are commonly used to determine group preferences for both the AIJ and AIP aggregation approaches. Aczel and Saaty (1983) claim that only the GM is an appropriate procedure as it preserves the reciprocal properties of the aggregated pairwise comparison matrices. Forman and Peniwati (1998) also state that the GM should be used for AIJ, as for AIP either the WAM or GM are meaningful procedures, satisfying the Pareto optimality principle. Lately, a third approach the weighted geometric mean (WGM) has been considered to be the optimal procedure (Levy and Taji, 2007).

Forum description and implementation

General Description of the Decision Problem

The decision problem that the PMI Global Congress 2010 – EMEA attendees will be asked to analyze concerns the selection of the host city for one of the forthcoming PMI EMEA congresses. The PMI EMEA’s congress is one of the biggest professional gatherings of its type, with a continually growing attendance. As a result, the logistics involved in the organization of the congress are gigantic. Arrangements must usually be made many months in advance. A successful conference is one where a large number of members attend and where all services match the expectances of attendees. The specific decision environment is considered appropriate as it falls within the knowledge and experience of all attendees, while some of them, hopefully, might be able to provide further insights, in case they have themselves participated in any conference organizing committee. The group decision making processes that will be followed are depicted in Exhibit 3 and analyzed hereafter.

Forum implementation steps

Exhibit 3 - Forum implementation steps

Group decision making processes

Creation of four groups of Decision Makers (DM)

According to the number of the attendees four groups of DMs will be formed. Those that have previously organized conferences will join the same group and will be considered as ‘experts’.

Definition of the Alternatives

The alternatives, namely the candidate host cities will be announced to the groups with little input from them, as described in the command style (refer to Exhibit 2).

Definition of Criteria

The attendees will define as a group the decision criteria. In order to do so they will be encouraged to utilize the consultative style of group decision making, as described above (refer to Exhibit 2).

Pairwise Comparisons

The pairwise comparisons are used to obtain the dominance or relative importance of influence by using the fundamental scale of the AHP (Saaty, 2005). The questions that each group will need to answer are formed as follows: Given a criterion X, which of the two elements Y, Z is more dominant with respect to that criterion? The fundamental scale of AHP includes intensity from 1 to 9, with 1 meaning equal importance and 9 extreme importance. For reasons of brevity and economy of time the attendees will be asked to perform the pairwise comparisons with respect only to the alternatives by using the intensities 1 (equal importance), 4 (moderate) and 7 (strong). For instance, the comparison matrix of all the alternatives with respect to the criterion site-seeing could look like Exhibit 4. That means that with respect to the criterion site-seeing City 2 is moderately (4) better than City 1 and so on.

Comparison of Alternatives with respect to Site-seeing

Exhibit 4 – Comparison of Alternatives with respect to Site-seeing

Obtaining Results and Discussion

In order to obtain the final results the mathematical calculation of the ANP will be performed and the final synthesis of all four groups will be made by using the WGM. The formulas are only given here for reasons of completeness since the project manager may use existing software for handling the mathematical burden. More specifically, to the experts’ group a weight of 34% will be assigned, while the results of the rest of the groups will be taken into account weighted by 22%, each, as shown in equation 1.

img

Where, wij,z the final weight for the pair of alternatives i and j with respect to the criterion z and img the relative weight provided by Group 1 (experts’ group), Group 2, Group 3 and Group 4, respectively. The weights of the elements (exhibit 4) of all comparison matrices are arranged both vertically and horizontally according to clusters. This matrix is known as the supermatrix. The supermatrix of an ANP network and detail of the matrices in it are presented in the matrices (2) and (3), respectively.

img

Where:

img

and

ci : cluster i, eij: element j of cluster i, ni the number of elements of cluster i, jk the kth element of cluster j (where k = 1, …, nj).

If needed, the supermatrix is normalized in order to provide the weighted supermatrix. Finally, the weighted supermatrix is transformed into the limit matrix by raising itself to powers. The reason for multiplying the weighted supermatrix is because we wish to capture the transmission of influence along all possible paths of the supermatrix. Raising the weighted supermatrix to the power 2k + 1, where k is an arbitrarily large number, allows convergence of the matrix, which means that all columns are the same (Saaty, 2005). From this limiting matrix the decision maker can derive the final solution. During the presentation, these mathematical steps will be performed by the appropriate software and a discussion will follow during which the participants will provide their opinion concerning the whole group decision making process.

Conclusion-Learning points

A final conclusion will be derived and the main learning points concerning both the group decision process and the method used will be clarified. Among completion of the session the attendees will have learned tips and tricks for group decision making so as not to “spoil the broth” and acquired basic understanding for using the ANP method in the real world project-related decisions.

References

Aczel, J. & Saaty, T. (1983, March), Procedures for synthesising ratio judgements, Journal of Mathematical Psychology, 27(1),93–102.

Belton, V. & Steward, T. (2002), Multiple Criteria Decision Analysis, Kluwer Academic Publishers, Dordrecht, Netherlands

Bonner, B. (2004, April), Expertise in Group Problem Solving: Recognition, Social Combination, Group Dynamics: Theory, Research, and Practice, 8(4) 277-290.

DETR (2002), Multi-Criteria Analysis: A Manual, Department of Environment, Transport and Regions, London. Forman, E. & Peniwati, K. (1998, July), Aggregating individual judgements and priorities with the the analytic hierarchy process, European Journal of Operational Research, 108(1) 165–169.

Kirytopoulos, K., Leopoulos, V. & Voulgaridou, D. (2008, July) Supplier selection in pharmaceutical industry: An analytic network process approach, Benchmarking: An International Journal (BIJ), 15(4) 494-516.

Levy, J. & Taji, K. (2007, October), Group decision support for hazards planning and emergency management: A Group Analytic Network Process (GANP) approach, Mathematical and Computer Modelling, 46(7-8) 906-917

Peniwati, K. (2005), “Criteria for evaluating group decision-making methods” in Saaty, T. and Vargas, L. (eds), Decision Making with the Analytic Network Process, Springer, US, pp. 251-273.

Project Management Institute. (2008) A guide to the project management body of knowledge (PMBOK®) (Fourth ed.). Newtown Square, PA: Project Management Institute.

Saaty T. (1989) Group decision making and the AHP, in B. Golden, E. Wasil and P. Harker (eds.), The analytic hierarchy process: applications and studies, Springer-Verlag, New York.

Saaty, T. (1996), Decision Making with Dependence and Feedback: The Analytic Network Process, RWS Publications, Pittsburgh, PA.

Saaty, T. (2005), Theory and Applications of the Analytic Network Process. Decision Making with Benefits, Opportunities, Costs and Risks, RWS Publications, Pittsburg, PA.

Tyler, T. & Smith, H. (1998), Social justice and social movements, in D. Gilbert, S. Fiske and G. Lindsey (eds.), The handbook of social psychology, McGraw Hill, Boston.

Vroom, V.H. and Yetton, P.W. (1973). Leadership and decision-making. Pittsburg: University of Pittsburg Pres

© 2010, K.Kirytopoulos, D. Voulgaridou, V. Voulgaridou
Originally published as a part of 2010 PMI Global Congress Proceedings – Milan, Italy

Advertisement

Advertisement

Related Content

Advertisement