Impact of information technologies on project management functions in the building design process


The dynamic nature of the design process, the interdependence of various participating entities, the high expectations for high-quality products, processes and services, and the need for teamwork, flexibility and a high degree of coordination suggest that information technology has great potential in the design phase of a building project. The objective of the study presented in this paper is to assess the impact of information technologies on project management functions in the building design process in the next 15 years.

There seems to be no doubt that the quality and productivity achieved in the building design process can be enhanced without additional cost via the application of the rapidly developing information technologies. But, how can executives readjust and if necessary realign the processes involved in building design to synchronize them with the requirements of new information technologies? A major part of the answer to this fundamental question relies on having a good idea of the potential impact that the application of information systems and technologies will have on project management functions used in the building design process in the coming 10 to 15 years.


The computerized model that is the subject of this paper forecasts the impact of major breakthrough developments in seven information technologies (virtual reality, computer–aided design / computer–aided manufacturing (CAD/CAM), artificial intelligence, geographical information systems / global positioning systems (GIS/GPS), Internet/Intranet, project management software, and wireless communication technology) on the maturity levels of the nine project management functions defined by A Guide to the Project Management Body of Knowledge (PMBOK® Guide) (integration, scope, time, cost, quality, human resources, communications, risk, and procurement management) of organizations involved in building design projects.

In the research reported in this paper, a modified Delphi approach was used over the Internet to elicit expert opinion. Three types of information were collected. First, the forecasted (next 15 years) maturity levels of the nine project management functions were obtained from building designers (for the conceptual and design phases). Second, the forecasted (next 15 years) probabilities of occurrence of major breakthrough developments in the seven information technologies were obtained from information technologists. Third, the assessments of the cross-impacts between the variables of the model were obtained from relevant respective parties. The data collected were subjected to a multi-level cross-impact analysis, which is a method typically used for long-term forecasting in financial markets. Cross-impact analysis was used for analyzing the interdependencies between the variables of the model by means of a simulation model built into the system.

Delphi Method

The objective of most Delphi applications is the reliable and creative exploration of ideas or the production of suitable information for decision-making. The Delphi Method is based on a structured process for collecting and distilling knowledge from a group of experts by means of a series of questionnaires interspersed with controlled opinion feedback (Adler & Ziglio, 1996). According to Helmer (1977), Delphi represents a useful communication device among a group of experts and thus facilitates the formation of a group judgment. Wissema (1982) underlines the importance of the Delphi Method as a monovariable exploration technique for technology forecasting. He further states that the Delphi method has been developed in order to make discussion between experts possible without permitting a certain social interactive behavior as happens during a normal group discussion and hampers opinion forming. Baldwin (1975) asserts that lacking full scientific knowledge, decision-makers have to rely on their own intuition or on expert opinion. The Delphi method has been widely used to generate forecasts in technology, education, and other fields (Cornish, 1977; Gatewood & Gatewood, 1983; Huss, 1988). The Delphi method recognizes human judgment as legitimate and useful inputs in generating forecasts. Later on, the notion of cross-impacts was introduced to overcome the shortcomings of this simplistic approach (Helmer, 1977).

In the original Delphi process, the key elements were (1) structuring of information flow, (2) feedback to the participants, and (3) anonymity for the participants. Clearly, these characteristics may offer distinct advantages over the conventional face-to-face conference as a communication tool. Goldschmidt (1975) agrees that there have been many poorly conducted Delphi projects. However, he warns that it is a fundamental mistake to equate the applications of the Delphi method with the Delphi method itself, as too many critics do. On the other hand, there have been several studies (Fowles, 1978; Helmer, 1981; Reuven & Dongchui, 1995; Wissema, 1982) supporting the Delphi method.

In general, the Delphi method is useful in answering one, specific, single-dimension question. An improvement in forecasting reliability over the Delphi method was thought to be attainable by taking into consideration the possibility that the occurrence of one event may cause an increase or decrease in the probability of occurrence of other events included in the survey (Helmer, 1977). Therefore cross-impact analysis has developed as an extension of Delphi techniques.

Cross-Impact Analysis

A basic limitation of many forecasting methods and the Delphi method is that they produce only isolated forecasts; that is, events and trends are projected one by one, without explicit reference to their possible influence on each other. Most events and developments however, are in some way connected to each other. Interdependencies between these events and developments can be taken into consideration for more consistent and accurate forecasts. Cross-impact analysis addresses Delphi's lack of a mechanism for discovering mutually exclusive or conflicting outcomes. Thus, some outcomes forecasted by Delphi could be impossible to obtain simultaneously (Reuven & Dongchui, 1995): for example, full employment and a low rate of inflation. Cross-impact analysis addresses this problem directly by analyzing conditional probabilities—for example, the likelihood that inflation will be low if full employment is achieved. It examines the interactions of forecasted items (Gordon & Hayward, 1968).

The cross-impact concept originated with Olaf Helmer and Theodore Gordon in conjunction with the design of a forecasting game for Kaiser-Aluminum (Helmer, 1977). It represented an effort to extend the forecasting techniques of the Delphi method. In 1968 at UCLA, Gordon and Hayward (1968) developed a computer-based approach to cross-impact analysis and they published their findings in the paper titled “initial experiments with the cross-impact matrix method of forecasting.” In this approach, events were recorded on an orthogonal matrix and at each matrix intersection the question was asked: If the event in the row were to occur, how would it affect the probability of occurrence of the event in the column? The judgments were entered in the matrix cells. Cornish (1977) stated that most forecasting methods may not consider many reactions between forecasted events. Cross-impact analysis, however, attempts to reveal the conditional probability of an event given that various events have or have not occurred. Duval et al. (1975) claim that cross-impact analysis differs from both probability theory and mathematical statistics; a cross-impact analysis is concerned with the identification of possible outcomes rather than with an understanding of what is or what was. They define cross-impact analysis as a systematic way to examine possible future developments and their interactions.

Technology forecasting does not follow a fixed methodological pattern. However, the way in which the study is approached and the choice of methods depend on the individual researcher (Wissema, 1982). Several versions of cross-impact analysis have been developed by researchers (Duval et al., 1975; Fargionne, 1997; Gordon & Hayward, 1968; Hanson & Ramani, 1988; Helmer, 1977; Novak & Lorant, 1978; Sarin, 1978; Wissema & Benes, 1980). The evaluation of the technique has not followed a single path but has produced a variety of different methods for constructing, utilizing, and evaluating cross-impact matrices.

There are several methodologies for different applications. Gordon and Hayward (1968) defined three modes of connection between variables. Assume event E1 occurs. A second event, E2, may be completely unaffected by E1; it may be enhanced by the occurrence of E1; or it may be inhibited by the occurrence of E1. Thus E1 may affect E2 as: Unrelated, Enhancing, and Inhibiting.

Exhibit 1. Conceptual Model of Cross-Impact Analysis

Conceptual Model of Cross-Impact Analysis

The impact of breakthrough developments in IT on the building construction process can be measured in terms of the deviations in the maturity levels of project management (PM) functions. The PM functions that are described in the PMBOK® Guide (Project Management Institute Standards Committee, 1996) are used in this study. The reason for the selection of these functions is their clear definition in terms of processes and their wide acceptance among project management professionals. The life cycle of a building project consists of the conceptual, design, construction, and operation phases. This holistic approach may benefit all stakeholders in building projects including the suppliers, the processors, and finally the customers of the construction industry. In this research three modules are taken into consideration (Exhibit 1):

MODULE 1: Information technology module

MODULE 2: Building processes module

MODULE 3: Project management functions module

In these modules, events, trends and phases are defined for causal cross-impact analysis. Module 1 defines the events; these are the breakthrough developments in information technologies. Module 2 identifies the trends; these are changes in the maturity level of the PM functions of the PMBOK® Guide. Module 3 covers the phases of a building project. Occurrence of the events in Module 1 has impacts on the trends in Module 2; these impacts are assessed within the framework of Module 3.

Exhibit 2. Scale Developed for Cross-Impacts

Scale Developed for Cross-Impacts

The modular design enhances the flexibility of the model. Therefore, the model can be easily modified according to the needs of its user by adding or deleting modules. This multifunctional, multiuser, modular approach enables the model to develop without strict limitations and costly modifications.

Three five-year scenes are proposed for the years 2000–2005, 2005–2010, 2010–2015 to define the time horizon in this study. The reasons for the selection of the time horizon of 15 years and the three time intervals of five years each, namely scenes are as follows:

• They are easy to understand.

• They are long enough to allow experts not to make trendbased forecasting.

• They are not too long for practical and planning purposes.

• They are short enough that inter-event and inter-trend impacts in the same scene can be ignored.

• Time intervals are suitable for long-term forecasting.

Exhibit 3. Forecasted Probability of Major Breakthrough Developments by Delphi Process

Forecasted Probability of Major Breakthrough Developments by Delphi Process

Classifying the causal cross-impacts into four groups makes the model easier to comprehend (see Exhibit 1). These causal cross-impacts between and within the modules are defined as:

• Inter-module impacts

• Inter-event impacts

• Inter-trend impacts

• Inter-phase impacts.

Inter-event impacts assess the impacts of ITs on ITs. The occurrence of an IT may decrease or increase the probability of occurrence of the other ITs in the following scenes. These impacts are elicited from IT experts by means of a Delphi process.

Inter-trend impacts state the interdependencies between the PM functions. Fluctuation of the maturity level of any PM function may have impacts on the others in the next scene. These interdependencies are obtained from PM professionals.

Inter-phase impacts assess the impact of phases on each other. Any fluctuation of the maturity level of any management function may not only have impacts on the same management function in the next scene, but may also have impacts on the same management function in the other phases of a building project. For example any improvement in the quality management function in the design phase may also affect the quality management function in the construction phase in the following scene. In other words, project management functions may be impacted by other project management functions in other phases of the project.

The impacts are classified in terms of intuitive perceptions. The reason for this is to make use of experts in a most convenient way. The following scale is developed to collect expert judgment. Actual cross-impact coefficients to be inserted in the matrix are also listed (see Exhibit 2).

+3 and –1 are not absolute limits; occasionally, for extremely large impacts, numbers could be employed whose absolute value is in excess of 3 or less than –1. This flexible scale lets the experts judge their intuitive interpretations’ effects on the model.

The modules and the causal relations between them are defined in the following sections. In this study, causal cross-impact analysis is inevitably conducted in a domain of what might be called “soft data” and “soft laws.” As Helmer (1977) stated that dependence on intuitive judgment is not just a temporary expedient but in fact a mandatory requirement. Therefore, instead of firm observational data, this model utilizes judgmental inputs; in place of well-confirmed empirical laws, the model makes use of intuitively perceived regularities. Therefore, reliance on expert opinion is essential. Even though the model can be used with data that are elicited from a single expert, a Delphi process is also included for added reliability.

In the Delphi process two types of data are elicited from the experts:

1. Predictions for the future scenes: event probabilities and trend values are elicited in respective scenes. These include:

• Predictions of the probability of occurrence of major breakthrough developments in ITs in respective scenes (Exhibit 3)

• Predictions of the expected maturity levels of the PMBOK® Guide project management functions in the design phase in respective scenes (see Exhibit 4)

Exhibit 4. Maturity Levels of PM Management Functions in the Design Phase of a Building Project Forecasted by Delphi process

Maturity Levels of PM Management Functions in the Design Phase of a Building Project Forecasted by Delphi process

Exhibit 5. Cross-Impacts of Information Technologies on PM Functions in the Design Phase

Cross-Impacts of Information Technologies on PM Functions in the Design Phase

2. Assessments for the cross-impacts: impacts of the variables (events, trends, and phases) on each other are elicited by means of the intuitive scale described above. These include:

• Assessments of the cross-impacts of ITs on PM functions in the design phase (see Exhibit 5)

• Assessments of the cross-impacts of ITs on ITs

• Assessments of the cross-impacts of trends (PM functions) on trends

• Assessments of cross-impacts of building project phases (conceptual, design, construction, and operation phases) on phases.

Data elicitation for all cross-impacts was obtained via Internet forms that were submitted to experts with required feedback for the Delphi process.

Information Technology Module

The IT module is basically a technology forecasting module. From the beginning of this study, IT's rapid development led us to consider a model that does not strictly depend on specific ITs but rather depends on a flexible technology module that can be changed and modified easily. This property of the model makes sure that the model will not be outdated because any development can be easily integrated into this modular model. Seven ITs are defined as events of the module. These ITs will be described later.

The IT module is the ignition point of the model (see Exhibit 1). This module defines the likelihood of occurrence of breakthrough developments in IT. When the model runs, it is this module that defines the scenarios based on the occurrence and nonoccurrence of the events in the respective scenes. Probabilities of occurrence of breakthrough developments in these events are elicited from IT experts (see Exhibit 3). Each scenario consists of three scenes (years 2000–2005, 2005–2010, 2010–2015) and the seven information technologies defined in the following sections. Currently this scenario-generating module considers only the occurrence and nonoccurrence of the events in the defined time horizon; therefore 221 (2,097,152) scenarios can be generated.

Project Management Functions Module

The PM functions module defines and measures the trends in PM functions in the process of building projects (see Exhibit 1). The maturity levels of PM functions are defined and measured by Kwak (1997) on a scale of 1–5 (1—no conformity, 5—complete conformity). Kwak (1997) reports current maturity levels of PM functions in the U.S. construction industry. Future trends in the maturity levels of PM functions are elicited from experts by means of a Delphi method (see Exhibit 4) in the design phase of a building project. The trends’ cross-impacts on each other are also elicited from experts. The nine PMBOK® Guide project management functions are used in this model and defined as trends.

Other trends throughout the building design process can also be simulated for different events. Modules such as productivity, safety, quality, etc., can replace this module or be added to the model. This way, the effects of technological and environmental developments can be forecasted on several aspects of building projects. This modular design makes this methodology a generic forecasting frame for the construction industry.

Building Processes Module

The building processes module includes the phases of the life cycle of a building project. This module basically forms the backbone of the model (see Exhibit 1). All PM functions and ITs are tested in the frame that is defined by this module. One can easily eliminate one or more phases or can easily add subphases. This brings the advantage of multipurpose use; one can focus on any phase or phases as this research focuses on the design phase of a building project. This module can be replaced by modules such as a management levels module (strategic, tactical, and operational) or a management hierarchy module (site or office practices), etc.

The building processes module also analyses the relationships between the phases of a building project. Interdependencies between these phases are defined in terms of cross impacts of the maturity levels of PM functions. If one considers a construction project as a whole system, any development in any phase of the process may have an impact on the other phases. Departments responsible for the processes are in fact internal customers of the company, and they provide and get both services and products to and from other departments. Their sensitivity to the other departments of the company has to be analyzed for their efficiency in addition to the efficiency of their relationships with external customers. This module therefore explicitly defines these interdependencies and helps professionals to understand their organization as a whole system.


The findings of this research are presented in two sections namely, Delphi survey results and cross-impact analysis. In the Delphi survey results, the data that are elicited from the experts are analyzed. In the second section, the multilevel cross-impact model is utilized in order to assess the impacts of IT on the maturity levels of the project management functions the design process of a building project.

The multilevel cross-impact analysis model developed in this research demands data from experts in a variety of disciplines (i.e., designers, construction managers, property managers, and information technologists). Hence the model solely depends on the intuitive judgment of the experts and the model's output can be only as good as this input. Therefore, the selection of these experts is very important. To that end, a Delphi survey of the members of PMI's Midwest Chapter in the U.S. was conducted. The endorsement of the Chapter's board was obtained. Volunteers were sought through an announcement in the Chapter's monthly newsletter. Later, members were invited via e-mail to join the study. A total of 34 experts accepted to participate. Data presented in three of the exhibits were elicited from these experts.

Delphi Results

The results of the Delphi survey show the impacts of ITs on PM functions (see Exhibit 5) and the forecasted maturity levels of the PM functions (see Exhibit 4) in the design phase of a building project. First, the significant interdependencies between ITs and PM functions will be discussed. Then forecasted maturity levels of PM functions for each scene (i.e., years 2000–2005, 2005–2010, 2010–2015) will be presented.

Exhibit 5 is designed to exhibit the impacts of breakthrough developments in ITs on the maturity levels of PM functions. Each cell contains a rating between 0 and 3 where 0 = No impact, 1 = Slight impact, 2 = Moderate impact, and 3 = Significant impact. The sum of the total impacts is recorded at the end of each row (i.e., impacting ITs) and column (i.e., impacted PM functions). The sum of the impacts at the end of a row indicates that particular IT's total impact on the PM functions, whereas, the sum of the impacts at the end of a column indicates the total impact coming from all the ITs to that particular PM function. For example, Exhibit 5 shows that scope management (column 3) is the most impacted PM function by ITs with a total impact rating of 13. On the other hand, artificial intelligence is the IT with the most impact on PM functions, with a total impact rating of 20 (column 11). The last row in Exhibit 5 indicates that the most impacted PM functions by ITs are scope management (rating = 13), time management (rating = 12), integration management (rating = 11), and communications management (rating = 11). The least affected PM functions are cost management (rating = 6), quality management (rating = 6), and human resources management (rating = 6). On the other hand, column 11 shows that the ITs with greatest impact on PM functions are artificial intelligence (rating = 20), PM software (rating = 18), and virtual reality (rating = 13), whereas the ITs with least impact are geographic information systems/geographic positioning systems (rating = 2) and wireless communication technologies (rating = 5).

The results of this Delphi survey indicate that the maturity levels of PM functions can be enhanced by the use of some ITs. Indeed virtual reality, computer aided design, and artificial intelligence technologies can actually help both the customer and designer to understand the actual building better before it is built. This ability can positively impact scope management activities in the design phase of a building project. Artificial intelligence and PM software can be utilized for more accurate time management. On the other hand, integration management is very important in the design phase because many problems related to the integration of the many disciplines and functions throughout the building process can be solved in the design phase. This may enhance constructability and quality of design. Communications management in the design phase is significantly impacted by breakthrough developments in Internet/intranet technologies. The building design process highly depends on information from sources as diverse as material dealers, technology developers, equipment manufacturers, technical consultants, etc. More and more business communications are done via the Internet/Intranet. Blueprints and other information can be transferred between the parties by electronic means. Increasing integration of the software and hardware technologies makes this possible. This creates many opportunities for design firms. It eliminates to some extent the barrier of geography; design firms can work with other professionals with more flexible schedules and with fewer limitations. ITs also support concurrent engineering activities that can according to Buckley (1994), if done correctly, reduce lead times from concept to customer, and increase quality performance. The ITs with most impact on communication management are artificial intelligence, PM software, and Internet/Intranet services.

The IT that has significant and moderate impact on all the four PM functions (scope, time, integration, and communications management) appears to be artificial intelligence. Indeed artificial intelligence may help designers in a variety of ways; for example, artificial intelligence programs given the design constraints can generate design alternatives. Generation and evaluation of the design alternatives may enhance the designer's creativity to address potential design problems. Of course, the other advantage of artificial intelligence is the reduction of the time required for the decisionmaking process. Traditional sketch drawings may take days to generate a few alternatives, whereas, numerous alternatives can be generated and evaluated by the artificial intelligence technologies in matter of minutes.

One of the three PM functions that are least impacted by ITs is cost management (total rating = 6). The IT that impacts cost management most is PM software technologies that seem to have a relatively moderate impact on the maturity level of the cost management function. This is because costs can be tracked and controlled in the design phase by cost control software. Quality management that also has the lowest rating of 6 is not impacted significantly by any of the ITs, except for virtual reality that has a moderate impact on it; it is either slightly or not impacted at all by the remaining ITs. The moderate impact of virtual reality on quality management may come from the importance of virtual reality presentations in order to get more customer input and solve potential conflicts in the design phase before they actually occur. This may increase the quality of the building project. Human resources management is the third function that is least impacted by ITs (total rating = 6). The strongest impact comes from PM software technologies that have a moderate impact on the maturity level of this function. While these functions do not receive direct impacts from ITs, they might be impacted by secondary impacts via other PM functions as discussed later.

On the other hand, artificial intelligence and PM software technologies significantly and moderately impact procurement management, respectively. Indeed the processes required to acquire goods and services from outside the performing organization and decision-making processes may benefit from both artificial intelligence and PM software technologies. Artificial intelligence in this case can help decision-makers in the selection process of sources or identifying the potential risk areas. PM software technologies may help to keep track of procurement planning and contract administration.

Breakthrough developments in artificial intelligence have the highest impact (total rating = 20) on the maturity levels of PM functions in the design phase. Indeed the design process is an iterative process, therefore artificial intelligence technologies may assist designers in their decision making process. Developments in artificial intelligence may enhance the efficiency of CAD and virtual reality technologies.

The second part of the Delphi survey was designed to retrieve the forecasted values of the maturity levels of PM functions in the design phase. The results are presented in Exhibit 4. The total change in the maturity level of each PM function is presented in column 6. The total change indicates the percentage increase in the maturity levels from scene 0 (i.e., year 1997) to scene 3 (i.e., years 2010–2015). The total changes show experts’ perception of potential improvement areas relative to each other. The highest increases can be seen in the maturity levels of quality management (total change 41%) and risk management (total change 31%); Kwak (1997) had measured the maturity levels of quality management and risk management as 2.9 in 1997, the lowest among the others. The maturity levels of scope and cost management on the other hand show little increases of 9% and 11% respectively. The predictions of the experts show that attaining full maturity in the industry (i.e., maturity level of 5) is not reachable within the time horizon of this study. That is why companies that deal with the design of building projects can differentiate themselves from their competitors by enhancing their PM maturity levels.

Cross Impact Analysis

A multilevel cross-impact analysis uses the data that are elicited from experts as described in the preceding sections. A computer run of 10,000 replications of the model produced the results that consist of forecasted maturity levels of PM functions and the probabilities of major breakthrough developments in ITs in the given time horizon (i.e., years 2000–2005, 2005–2010, 2010–2015). Hence, the trigger actions of the model in the first scene can only affect the second scene (i.e., 2005–2010); the forecasted values for the first scene (i.e., 2000–2005) are limited to the Delphi survey results. The second and the third scenes (i.e., 2005–2010, 2010–2015) have cross-impact analysis values along with the Delphi results. Because of the nature of the study, the relationships between the variables are all positive; therefore, the results of the Delphi survey are either positive (i.e., “significant,” “moderate,” and “slight”) impacts or “no” impacts. Lack of negative impacts naturally makes the model generate higher forecasts then the Delphi results. This is because the model generates new estimates based on the Delphi estimates and sums up all the negative and positive impacts around the Delphi results. Indeed, if there were negative impacts, the model might have generated lower values than the estimated Delphi results (in cases where the negative impacts overcame the positive impacts). For example, if we assume that the events (trigger actions) are not occurring at all, then the model should and actually does present the original Delphi estimates (i.e., forecasted maturity levels and forecasted probabilities), since there are no impacts either positive or negative on the estimated trend and event values. The model's performance depends on the Delphi estimates; indeed the results of the model can be as good as the inputs.

The results are presented in two sections. First, the results of the basic computer run are discussed. Then sensitivity analyses are conducted in order to determine the sensitivity of the PM functions in each phase to major breakthrough developments in ITs.

Exhibit 6. Maturity Levels of PM Management Functions in the Design Phase of a Building Project Forecasted by Model

Maturity Levels of PM Management Functions in the Design Phase of a Building Project Forecasted by Model

Results of the Basic Run

A basic computer run of the model consists of 10,000 replications. Computer runs of 1,000, 4,000, and 10,000 replications resulted in standard errors of 0.0038, 0.0019, and 0.0012 respectively for the integration management function with similar results in the other PM functions. For comparative policy analysis and scenario generating purposes, replications of 1,000 and above are reasonable given the low standard errors. The results consist of the forecasted maturity levels of PM functions in the design phase of a building project (see Exhibit 6) and the forecasted probabilities of the major breakthrough developments in ITs. Since the model produces values only for the second and the third scenes, the analysis will be conducted in the second (i.e., years 2005–2010) and the third scenes (i.e., years 2010–2015). The values in the first scene (i.e., years 2000–2005) are the Delphi forecasts. As discussed earlier, the model produces higher estimates than the Delphi forecasts in the last two scenes. A ranking of the mean values of the estimates or a ranking of the percentage increases are useful in this case.

In the design phase, Exhibits 3 and 5 are used for comparative analysis. The model forecasts more increases than the Delphi forecast for reasons discussed earlier. The differences can be calculated easily by subtracting the respective values of total change in column 6 of Exhibit 4 from column 6 of Exhibit 6. For example, while total change in integration management is forecasted by experts as 18%, the model forecasts a 30% total change. The difference of 12% shows that, experts’ estimates are not consistent with the model's forecasts. The same situation can be observed in scope management with an 11% higher increase forecasted by the model. In contrast, the forecasts generated by the model for communications management (difference = 3%) and cost management (difference = 5%) are much closer to the predictions made by experts. The mean of the differences of total changes between the Delphi forecasts and model forecasts is 8% in the design phase.

Sensitivity Analysis

Sensitivity of each PM function to an IT is important in order to understand the impacts of ITs on PM functions. The design of the model enables sensitivity analysis. For example, the sensitivity of the PM functions to major breakthrough developments in the Internet/intranet technologies can be observed by setting all the events’ (i.e., ITs’) occurrence probabilities to “0” except for Internet/Intranet technologies. Then running the model will show the impact of Internet/intranet technologies on the PM functions without the other ITs’ interactions.

Exhibit 7. Sensitivity of the Maturity level of PM Functions to Major Breakthrough Developments in IT in the Design Phase of a Building Project in Terms of Percentage Change

Sensitivity of the Maturity level of PM Functions to Major Breakthrough Developments in IT in the Design Phase of a Building Project in Terms of Percentage Change

For each sensitivity run, the model starts at the same random number entry which guarantees that, at analogous stochastic decision points, the likelihood of major breakthrough developments will be as “lucky” or “unlucky” as in the preceding basic run. Each basic run consists of 10,000 replications. The sensitivity analysis is conducted for each of the seven ITs defined in this study.

The deviations of the results (i.e., results of sensitivity runs) from the Delphi forecasts are used for sensitivity analysis. In order to accomplish the objective of this study, deviations from the Delphi estimates are calculated and summarized in terms of percentages for the design phase (see Exhibit 7). Exhibit 7 shows the sensitivity (% change) of each PM function to the given IT. For example, the maturity level of integration management in the third scene of the design phase is 3.9 (see Exhibit 4). Considering breakthrough developments only in virtual reality, the maturity level of integration management will be 4.0 in the third scene (result of the model). The difference is 0.1 that is 2.6% increase from the Delphi estimate. The highest deviation in this study is found to be 5.7%. This result indicates that breakthrough developments in an IT taken individually do not impact the maturity level of any one PM function significantly, whereas they effect the maturity level of all PM functions through primary (tangible) and secondary (intangible) impacts. This finding can be used to maximize the benefits of IT deployment and therefore decreases the cost and failure rates at the initial phases of reengineering/restructuring efforts. The impact of an individual IT on the maturity level of all PM functions can be observed from the bottom row (total change %) of Exhibit 7.

The sensitivity analyses consider both primary and secondary impacts of IT. The primary impacts are the ones that affect the maturity level of a PM function directly. The secondary impacts are the impacts coming from the other functions and/or the impacts coming from the same function in the different phases. If only the primary impacts are considered, the results should and actually do reflect the Delphi results about the cross-impacts of ITs on PM functions. However, secondary impacts make this model more realistic because it considers the interdependencies between the variables.

The summary of the results for the design phase is presented in Exhibit 7. PM functions are sensitive to breakthrough developments in PM software (total change 25.3%), virtual reality (total change 22.8%), and artificial intelligence (total change 22.8%). The overall maturity level of project management functions in a design firm may be enhanced most by investing in these ITs. Focusing on one PM function will not help the organization to reach higher overall maturity levels. These results indicate that IT investments should be considered in a holistic environment. The secondary impacts of improvements will spread benefits of IT deployment throughout all PM functions.

Studies conducted by Alter (1996) and Callon (1996) conclude that IT investments are useless if the current business processes are not analyzed and redesigned to perform best with the new ITs. The different sensitivity of each PM function to an IT in the different phases of a project support this idea. Indeed, PM functions aim to achieve the same goals in each phase of a building project but the tools to achieve these objectives should be and actually are different from each other. For example, human resources management in the design phase has different procedures then human resources management in the construction phase. While a relatively small number of employees dominate the process in a design firm, a large labor force may be typical in a construction company. To deal with these differences PM functions should utilize different ITs for their specific cases.

The modular design of the model allows the researcher to add other ITs in order to test the sensitivity of the maturity levels of PM functions in the process of a building project for those specific ITs. As the results of this study indicate, the pace of the developments in ITs is very fast. Change breeds change, and each change in IT brings different opportunities for the building design process. Further and more detailed studies in this area may help building construction professionals to redesign their processes for better performance in the light of the new ITs.


The results of the study show that the maturity levels of PM functions in the design process of the building project can be enhanced by major breakthrough developments in some ITs. In the time horizon of this study experts do not foresee the full maturity levels for PM functions in the design phase of a building project. However, steady increases are expected for the PM maturity levels. Increases are expected to be parallel to each other, which suggest strong interdependencies between the maturity levels of PM functions. Therefore improvement efforts to PM functions have to consider not one or few but all of the PM functions together. Likewise IT deployment should target not one but all PM functions in order to increase overall PM maturity level.

Delphi results indicate an increasing pace of technological innovation in some ITs. The increasing rate of major breakthrough developments has potential to change the processes in the design phase of a building project. This suggests that designers in the building construction industry have to be ready to face emerging challenges created by rapid major breakthrough developments in ITs.

Given the cross-impacts between the variables, the model showed inconsistencies of the Delphi survey results (i.e., expert opinions) in some cases. For these cases experts may be asked again to consider their Delphi forecasts in the light of these findings. These inconsistencies are to be expected and probably the results of the multidimensional interdependencies (direct and secondary impacts) which cannot be judged by the experts. In this case, the model serves as a learning tool for professionals.

The maturity levels of PM functions are found to be sensitive to breakthrough developments in PM software, virtual reality, and artificial intelligence. Sensitivity analyses showed that developments in single ITs do not have significant impacts on the maturity levels of PM functions. On the other hand, major breakthrough developments in several ITs with their direct and secondary impacts can enhance the PM maturity levels of a building design organization. Therefore, restructuring/reengineering efforts should address all aspects of process improvement. As the literature survey indicates, IT deployment without addressing the structural issues in the organisation does nothing but automation. This research showed that PM maturity levels in the building design process can be enhanced to some extent by the use of some ITs.

The model's scenario generating capability can be used effectively for planning and analysis purposes. The model can be used for comparative policy analysis. The testing of a variety of policies and their implementation through specific plans and appropriate budget allocations may identify several policies as potentially promising. The integrated approach used in this research both in terms of disciplines and in terms of phases helps professionals to develop robust strategies of IT deployment in their projects.

At a global level, this research developed a unique methodology for understanding ITs impact on the maturity levels of PM functions in the design phase of a building project integrating expert, IT-specific, PM-specific, and process-specific knowledge, and a mathematical structure. It is important to realize that “playing the model” is a learning experience. The greatest payoff may not be the determination of optimal strategies but an increasing understanding of the insights of the subject and the gradual production of an improved model. This model might provide a conceptual framework within which to formulate forecasts, policies, and plans in areas where interdisciplinary approaches are essential and where reliable and comprehensive theories are absent.

Adler, M., & Ziglio, E. (1996). Gazing into the oracle. Bristol, PA: Jessica Kingsley Publishers.

Alter, S. (1996). Information systems: A management perspective (2nd ed.). New York: The Benjamin/Cummings Publishing Company, Inc.

Callon, J.D. (1996). Competitive advantage through information technology. New York: McGraw-Hill.

Cornish, E. (1977). The study of the future. Washington, DC: World Future Society.

Duval, A., Fontela, E., & Gabus, A., Eds. (1975). Portraits of complexity. (Ed. Baldwin, M. M.) Columbus: OH: Battelle Memorial Institute.

Enzer, S., & Alter, S. (1978, June). Cross-impact analysis and classical probability: The question of consistency. Futures, 10 (3), 227–239.

Fowles, J. (1978). Handbook of futures research. Connecticut: Greenwood Press.

Gatewood, R.D., & Gatewood, E.J. (1983). The use of expert data in human resource planning: Guidelines from strategic forecasting. Human Resource Planning, 5 (1), 83–94.

Gordon, T.J., & Hayward, H. (1968, December). Initial experiments with the cross-impact matrix method of forecasting. Futures, 1 (2), 100–116.

Hanson, W.H., & Ramani, N. (1988). Technology forecasting: A hydroelectric company experience. Technology Management Publication, 1 (3), 266–270.

Helmer, O. (1977, Feb.). Problems in futures research: Delphi and causal cross-impact analysis. Futures, 17–31.

Helmer, O. (1981, Oct.). Reassessment of cross-impact analysis. Futures, 389–400.

Huss, W.R. (1988). A move toward scenario analysis. International Journal of Forecasting, 4, 377–388.

Kwak, Y.H. (1997). A systematic approach to evaluate quantitative impacts of project management. Ph.D. dissertation, University of California, Berkeley.

Novak, E., & Lorant, K. (1978). A method for the analysis of interrelationships between mutually connected events: A cross-impact method. Technological Forecasting and Social Change (12), 201–212.

Project Management Institute Standards Committee. (1996). A Guide to the Project Management Body of Knowledge. Upper Darby, PA: Project Management Institute.

Reuven, L., & Dongchui, H. (1995). Choosing a technological forecasting method. Industrial Management, 37 (1), 14–22.

Sarin, R.K. (1978, February). A sequential approach to cross-impact analysis. Futures, 53–62.

Wissema, J.G., & Benes, J. (1980, October). A cross-impact case study: The Dutch construction sector. Futures, 394–404.

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.

Proceedings of PMI Research Conference 2000



Related Content