ADAPTING THE CAPABILITY MATURITY MODEL (CMM) TO UNRELATED INDUSTRIES AS A PROCESS MATURITY FRAMEWORK

Abstract

No single baseline exists to uniquely address process improvement issues with respect to industrial management from an evolutionary and process maturation perspective. This research illustrates that the software industry's Capability Maturity Model (CMM) can be adapted within unrelated industries as a process maturity framework. This research also shows that existing process improvement paradigms do not address issues of process maturity and that existing industrial process environments do not conform to CMM tenets. [PUBLICATION ABSTRACT]

Headnote

ABSTRACT

No single baseline exists to uniquely address process improvement issues with respect to industrial management from an evolutionary and process maturation perspective. This research illustrates that the software industry's Capability Maturity Model (CMM) can be adapted within unrelated industries as a process maturity framework. This research also shows that existing process improvement paradigms do not address issues of process maturity and that existing industrial process environments do not conform to CMM tenets.

INTRODUCTION

This research provides the data descriptions, analysis, and findings from survey research of the software Capability Maturity Model (CMM) architecture and the potential applications for process improvements. (The survey is available upon request.) In the full paper, the first section provides the discussion of the survey responses and survey instrument. The second section discusses the concerns of non-response bias. A discussion of the findings is contained in the third section. The remaining sections discuss the secondary analytical procedures and conclusions.

SURVEY RESPONSES AND DATA COLLECTION

Fifteen hundred unique candidate firms comprised the superset. A total of 3 OO surveys were distributed to the randomly selected firms. No duplicate recipients were allowed. Sixty-four surveys were completed. The response rate for the survey was 21.3 percent; however, 16 of the respondents declined participation. Forty-seven responses generated the data sets used in this research. The data sets represented a 15.67 percent participating survey response rate.

The survey was divided into three sections: Environmental Characteristics, Improvement Initiatives, and Demographic Data. The first section queried aspects of corporate infrastructure regarding process attributes. The second section queried aspects of process improvement initiatives (both previous and existing) of work environments. The third section of the survey queried demographic data of respondent environments.

Statements contained within section one and two of the survey consisted of five possible scaled responses: 1. Strongly Disagree, 2. Disagree, 3. Unable to Judge, 4. Agree, and 5. Strongly Agree. Section three contained checklist response selections.

Section one of the survey was composed of questions 1-10. The questions investigated various characteristics of working environments. Section two was composed of questions 11-16, which considered improvement initiatives within working environments. Section three consisted of questions 17-23, which collected demographic data that described respondent firms.

ANALYSIS AND EVALUATION OF THE FINDINGS

ANOVA was performed with respect to a portion of the survey data. Survey analysis included ANOVA for managerial and non-managerial personnel for each individual hypothesis statement. ANOVA decomposed variability into primary elements with respect to the collected data. Unlike other tests in which the examination of means, standard deviations, or other values is advocated, ANOVA implements either squared deviations or the variance so that computation of distances of individual data points from their own means or from the overall mean may be summarized (Siegel, 1994). Within this model, individual groups have their own means and values that deviate from that mean. All the data points from all of the groups generate the overall mean. Total deviation is generated via calculation of the sum of the squared differences between the data points and the overall mean.

The Pearson correlation coefficient is a measure of the linear association between two variables and is used when examining whether a significant relationship between these variables exists. The values of the correlation coefficient (r) range in scope from -1.0 to +1.0. The direction of the relationship is represented with the sign of the coefficient. The absolute value of the correlation coefficient indicates the strength of the linear relationship between the given variables. Smaller absolute values indicate a weaker relationship, whereas larger absolute values indicate a stronger relationship. Pearson correlation coefficients were used to investigate relationships between survey questions 14 and 15 and between questions 14 and 16.

PRIMARY DATA ANALYSIS AND HYPOTHESIS TESTING

Primary hypothesis testing concerns the first five questions of the survey. These questions provide the basis for the hypothesis statements of this paper. The primary hypothesis statements are as follows:

H0: μ^sub Mgt^ = μ^sub NMgt^ Production processes may be defined as being ad hoc or chaotic.

H1: μ^sub Mgt^ ≠ μ^sub NMgt^ Production processes may not be defined as being ad hoc or chaotic.

H0: μ^sub Mgt^ = μ^sub NMgt^ Production processes are disciplined and repeatable.

H1: μ^sub Mgt^ ≠ μ^sub NMgt^ Production processes are not disciplined and not repeatable.

H0: μ^sub Mgt^ = μ^sub NMgt^ Production processes are standardized and consistent.

H1: μ^sub Mgt^ ≠ μ^sub NMgt^ Production processes are not standardized and are not consistent.

H0: μ^sub Mgt^ = μ^sub NMgt^ Production processes are predictable.

H1: μ^sub Mgt^ ≠ μ^sub NMgt^ Production processes are not predictable.

H0: μ^sub Mgt^ = μ^sub NMgt^ Production processes are continuously being evaluated for improvement.

H1: μ^sub Mgt^ ≠ μ^sub NMgt^ Production processes are not continuously being evaluated for improvement.

SECONDARY DATA ANALYSIS

Secondary data analysis for this paper considers survey questions 6-16. These questions continued to investigate various characteristics of working environments with respect to the tenets of both the CMM and the proposed IPMM. Questions 6-13 evaluated management and non-management perceptions concerning each given survey statement. Questions 14-16 were examined to determine correlations among the data.

ANOVA was used to examine the data associated with questions 6-13. Questions 14-16 implemented the Pearson correlation coefficient to examine whether a significant relationship between these survey objects exists. The values of the correlation coefficient (r) range from -1.0 to +1.0. The direction of the relationship is represented with the sign of the coefficient. The absolute value of the correlation coefficient indicates the strength of the linear relationship between the given variables. Smaller absolute values indicate a weaker relationship whereas larger absolute values indicate a stronger relationship.

PEARSON CORRELATION COEFFICIENTS

Koenker (1971) and Walpole and Myers (1993) state that the Pearson correlation coefficient is a measure of the linear association between two variables and is used when examining whether a significant relationship between these variables exists. The values of the correlation coefficient (r) range from -1.0 to +1.0, and the direction of the relationship is represented with the sign of the coefficient (Siegel, 1995). The absolute value of the correlation coefficient indicates the strength of the linear relationship between the given variables. Smaller absolute values indicate a weaker relationship whereas larger absolute values indicate a stronger relationship. Pearson correlation coefficients were used to investigate relationships between questions 14 and 15 and between questions 14 and 16. Hinkle, Wiersma and Jurs (1998) present the basic formula used to calculate the Pearson correlation coefficient.

PEARSON EVALUATION OF QUESTIONS 14 &15

The Pearson correlation coefficient formula was used to investigate the relationship between the issues represented with questions 14 and 15. The following values were implemented with Pearson's formula:

The value of the correlation coefficient (r) is within the acceptable range of -1.0 to +1.0. In this instance, the value of r is -0.02177. The negative direction of the relationship is represented with the sign of the coefficient (negative). The absolute value of the correlation coefficient indicates the strength of the linear relationship between the given variables. In this instance, the absolute value is 0.02177. The smaller absolute value indicates a weaker relationship between the data representing the given survey questions. Therefore, little, if any, correlation is indicated from this analysis.

The Pearson correlation coefficient formula was used to investigate the relationship between the issues represented with survey questions 14 and 15. The following survey values were implemented with Pearson's formula:

The value of the correlation coefficient (r) is within the acceptable range of -1.0 to +1.0. In this instance, the value of r is 0.02191. The positive direction of the relationship is represented with the sign of the coefficient (positive). The absolute value of the correlation coefficient indicates the strength of the linear relationship between the given variables. The absolute value is 0.02191. The smaller absolute value indicates a weaker relationship between the data representing the given survey questions. Therefore, little, if any, correlation is indicated from this analysis.

CONCLUSION

This section presents a brief synopsis of the research findings. ANOVA was used to test the primary hypothesis statements using data collected from management and non-management groups regarding their perceptions of process maturity as a component of process improvement initiatives. The outcomes of the primary ANOVA hypothesis testing are presented in Table 3.

ANOVA was implemented for questions 6 through 13. The outcomes of the analysis for each of these statements are given in Table 4:

Pearson correlation coefficients were used to determine relationships between statements 14, 15, and 16 of the survey. Table 5 presents the testing outcomes:

References

REFERENCES

Hinkle, D., W.Wiersma & S. Jurs (1998). Applied statistics for the behavioral sciences (Fourth Edition). Boston: Houghton-Mifflin Company.

Koenker, R. (1971). Simplified statistics for students in education and psychology. Totowa, NJ: Littlefield, Adams, and Co.

Siegel, J. (1994). Practical business statistics (Second Edition). Hillsdale, IL: Irwin Publishing Company.

Walpole, R. & R. Myers (1993). Probability and statistics for engineers and scientists (Fifth Edition). New York: MacMillan Publishing Company.