Impact of communication on success of engineering design projects
Previously unimaginable increases in information technology and innovative new production techniques present the design and construction industry with tremendous potential to improve the program and project delivery process. Fundamental to these changes, however, is the need for improved communication (Gushgari et al., 1997; Kartam, 1997; Liu et al., 1994; Luiten & Tolman, 1997; Thomas et al., 1998). As shown in the Exhibit 1, communication needs within design and construction can be divided into two broad types—information distribution and human understanding. The hypothesis of this research is that each element is essential to success in an engineering design project. Without adequate means to communicate/distribute information it is little value to anyone other than a single person or group. Without a human and organizational framework in which to use the information, it is of no value to anyone. This paper presents the results of an analysis of over 240 completed civil engineering design projects and examines the correlation between both of these types of communication and project success.
Data Collection and Development of Measures
The study data is from a 1,000 employee public works engineering bureau that is responsible for the planning, design, and construction management of capital improvement projects for a large city in the western United States. Two hundred and forty-three projects including municipal facilities, stormwater, sewer, and street projects were studied in this analysis of communication impacts. The construction costs of the projects ranged from $25,000 to $15,000,000, and completion dates were between July 1993 and May 1999. The projects all had multidiscipline design requirements (civil and other) and all established project teams. The teams consisted of the primary discipline design squad, additional discipline designers, oversight program managers, and client agency contacts. The level of communication, however, varied within these project teams from project to project.
In order to test the research hypothesis, measures of communication (in terms of information distribution and human understanding) were established. In this analysis, communication effectiveness was measured across four planes—two related to information distribution and two related to human understanding and teamwork. The two communication measurements defined to measure information distribution were:
• Average number of meetings held per month of the project (varies from 0.25 to 1.0)
• Average number of status reports issued per month (varies from 0.1 to 1.0).
The two communication measures defined to measure human understanding and teamwork were:
• Completion of human relations/communication training by the design squad (Yes or No)
• Organizational structure in which the design squad is working (Matrix or Functional).
Exhibits 2 and 3 show summaries for communication data. Exhibit 2 shows a data summary the two communication measurements related to information distribution. Exhibit 3 shows a data summary the two communication measurements related to human understanding and teamwork.
Engineering project performance is traditionally best measured as adherence to scope, budget, schedule, and quality. Based on the historical data available (243 projects), two engineering design project performance measures were established. One measure of project success was defined as the design phase cost performance index (CPI = BCWP / ACWP) (PMI®, 1996) for design labor-hours spent on the design of each of the 243 projects. The second measure of project success was defined as the construction phase cost performance index (CPI = BCWP / ACWP) (PMI, 1996) for designer labor-hours spent during the construction of each of the 243 projects. This second measure is somewhat of a quality measure as well as a cost measure. If the quality of an engineering design is poor, then additional designer labor-hours (beyond a standard budgeted amount) will be needed to correct design quality errors. It is recognized, however, that not all quality errors will immediately be manifested as additional designer labor-hours during construction (they may become apparent many years into the service life of a project), and many times additional designer hours may be required during construction due in no part to design quality (i.e., contractor rework or substitutions). Unfortunately, no schedule performance data was available for the 243 completed projects. Exhibit 4 shows a data summary for the design and construction phase CPI‘s for designer labor-hours. As shown in the exhibit, for the projects of this study, the average engineering labor costs were over budget in both the design (Design CPI (average) = 1.55) and the construction (Construction CPI (average) = 1.97) phases.
Each of the four communication effectiveness measures is then tested verses the two project success criteria. Based upon this analysis, a relative measure of impact of communication importance to project success can be made, and the hypothesis of the paper is tested.
Human Understanding and Teamwork
In this analysis, the average CPIs for each possible outcome of each of the two measures of human understanding and teamwork are compared across each measure. One theoretical measure of human understanding and teamwork is the organizational structure the design team functions under. The two organizational structure observed in the data set of projects for this study were matrix (arguably assumed to be more communicative) and functional (less communicative). Exhibit 5 shows the results of whether the organizational structure in which the design squad is working (matrix or functional) influences the average design phase CPI measurement for designer labor-hours. Exhibit 6 shows the results of whether the organizational structure in which the design squad is working (matrix or functional) influences the average construction phase CPI measurement for designer labor-hours. In both Exhibits, it is seen that projects completed under a matrix organizational structure for the design team (45 occurrences) had a higher CPI than projects completed under a functional organizational structure for the design team (198 occurrences). Hence, this increased measure of communication did not influence engineering design performance.
The second theoretical measure of human understanding and teamwork is whether or not the design team completed outside consultant training related to teamwork and communications. The two training measurements observed in the data set of projects for this study were yes, the training was completed, or no, the training was not completed. Exhibit 7 shows the results of whether the training of the design squad influences the average design phase CPI measurement for designer labor-hours. Exhibit 8 shows the results of whether the training of the design squad influences the average construction phase CPI measurement for designer labor-hours. In both Exhibits, it is shown that projects completed by teams that did not complete the training (208 occurrences) had a lower CPI than projects completed by teams that did complete the training (35 occurrences). Hence, again, this increased measure of communication did not influence engineering design performance as defined by the study measures.
A similar analysis was conducted for the information distribution measures. In this analysis, the average CPIs for subdivided high and low outcomes of each of the two measures of information are compared across each measure. One theoretical measure of information distribution is the number of meetings held by the design team each month. Exhibit 9 shows the results of whether infrequent meeting (defined as less than 1.000 meetings per month) or frequent meeting (defined as more or equal to 1.000 meetings held per month) influences the average design and construction phase CPI measurements for designer labor-hours. Exhibit 9 shows that projects with frequent meeting had higher CPIs than projects with less frequent meetings. Hence, this increased measure of communication did enhance engineering design performance.
The second theoretical measure of information distribution is the number of reports issued by the design team each month. Exhibit 10 shows the results of whether infrequent reporting (defined as less than 0.75 reports per month) or frequent reporting (defined as more or equal to 0.75 reports issued per month) influences the average design and construction phase CPI measurements for designer labor-hours. Exhibit 11 shows that projects with frequent reporting had higher CPIs than projects with less frequent reporting. Hence, this increased measure of communication did also enhance engineering design performance.
Given these positive performance influences for meeting and reporting frequency, the following studies were done: scatter plots, coefficients of correlation, and regression line plots for CPI measures verses meeting and reporting frequency. Exhibit 11 shows the plot of correlation of design CPI to meeting frequency. Exhibit 12 shows the plot of correlation of construction CPI to meeting frequency. Exhibit 13 shows the plot of correlation of design CPI to reporting frequency. Exhibit 14 shows the plot of correlation of construction CPI to reporting frequency.
All plots show a positive (although weak) correlation between communication and project performance. Correlation coefficients are as follows:
• Design CPI to Meeting Frequency: 0.259
• Construction CPI to Meeting Frequency: 0.130
• Design CPI to Reporting Frequency: 0.119
• Construction CPI to Reporting Frequency: 0.066
All plots also show negative slope regression lines. Slopes are as follows:
• Design CPI to Meeting Frequency: –0.893
• Construction CPI to Meeting Frequency: –0.589
• Design CPI to Reporting Frequency: –0.429
• Construction CPI to Reporting Frequency: –0.312
Hence, increased communication was shown to improve project performance. Performance was more positively influence by communication in the design phase as compared to the construction phase. Performance was also more positively influenced by meeting frequency than by reporting frequency.
Additional Research and Conclusions
This paper has presented the results of a study of 243 completed engineering design projects. The research consisted of an analysis of the influence between communication (measured as frequency of design team meetings, frequency of design team reports, completion of human relations/communication training by the design squad, and organizational structure of the organization in which the design is done) and performance of the engineering design (measured as cost performance index for designer labor-hours for the design and construction phases). The study found:
• No correlation between increased human understanding communications measures (completion of human relations/communication training by the design squad, and organizational structure of the organization in which the design is done) and engineering design performance for both the design and construction phases of a project.
• Positive correlation between increased information distribution communication measures (frequency of design team meetings, frequency of design team reports) and engineering design performance for both the design and construction phases of a project.
The research hypothesis of the work was therefore found to be partially true.
This work, however, is only a first step in a great need to recognize and quantify the importance eof communication in engineering. Several items need to be further examined by future researchers. Specific needs are:
• Development of stronger engineering design performance measures (especially with respect to schedule)
• Larger sample size of projects from both public and private design firms to confirm the findings of this study
• Development of a method to eliminate other potential positive or negative dependent influences to communication (ensure statistical independence of the study measures).
Nonetheless, the value of this research to our profession is immediate. The research has defined and quantified the importance of communication within project management and presents methods and ideas that can be applied in other fields within the project management research community and industry.
Gushgari, Shakir K., Francis, Peter A., & Saklou, Jamal, H. (1997). Skills critical to long-term profitability of engineering firms. Journal of Management in Engineering, 13 (2), 46 –56.
Kartam, Nabil A. (1996). Making effective use if construction lessons learned in project life cycle. Journal of Construction Engineering and Management, 122 (1), 14–21.
Lui, L.Y., Sumpf, A.L., Kim, S.S., & Zbinden, F.M. (1994). Capturing as-built information far facility management. Proceedings of the First Congress on Computing in Civil Engineering, ASCE.
Luiten, Gijsbertus T., & Tolman, Frits P. (1997). Automating communication in civil engineering. Journal of Construction Engineering and Management, 123 (2), 113–120.
Project Management Institute (PMI) Standards Committee. (1996). A guide to the project management body of knowledge (PMBOK® guide). Upper Darby, PA: Project Management Institute.
Thomas, Stephen R., Tucker, Richard L., & Kelly, William R. (1998). Critical communications variables. Journal of Construction Engineering and Management, 124 (1), 58–66.
Proceedings of PMI Research Conference 2000