The longitudinal effects of the ISO 9000 certification process on business performance

European Journal of Operational Research 146 (2003) 580–595 www.elsevier.com/locate/dsw

Production, Manufacturing and Logistics

The longitudinal effects of the ISO 9000 certification process on business performance Mil
e Terziovski b

a,*

, Damien Power

b,1

, Amrik S. Sohal

b,2

a Department of Management, The University of Melbourne, Parkville, Victoria 3010, Australia Department of Management, Monash University, P.O. Box 197, Caulfield East VIC 3145, Australia

Received 24 April 2001; accepted 21 February 2002

Abstract This paper develops and tests several hypotheses relating to ISO 9000 quality system certification process using data collected in a cross-sectional study undertaken in Australia. Multivariate analysis is used to analyze the quantitative data and test the hypotheses. Our findings show that there is a significant and positive relationship between the managerÕs motives for adopting ISO 9000 certification and business performance. Those organizations that pursue certification willingly and positively across a broad spread of objectives are more likely to report improved organizational performance. The individual element found to contribute most to business performance was customer focus. The principal motivation to pursue ISO 9000 certification was found to come from customer pressure. Auditing style was found to have an insignificant (positive or negative) effect on business performance. The ability of the new ISO 9001-2000 standard to capture and meet the conformance and performance requirements of the organization as part of a continuous improvement strategy will be a prime determinant of the extent to which managers embrace or reject ISO 9000 certification in the 21st century.
2002 Elsevier Science B.V. All rights reserved. Keywords: ISO 9000; Certification; Business performance; Australia

1. Introduction During the last several years, there has been a major push by manufacturing firms in over 100 countries to seek certification to the International

*

Corresponding author. Tel.: +61-3-93447868; fax: +61-383443714. E-mail addresses: [email protected] (M. Terziovski), [email protected] (D. Power), [email protected] (A.S. Sohal). 1 Tel.: +61-3-99031030; fax: +61-3-99032718. 2 Tel.: +61-3-99032033; fax: +61-3-99032979.

Quality Standards ISO 9000. These standards require a high level of documentation plus audited evidence that the intended quality is being delivered to the customer. The ISO 9000 standard points out that the quality system of an organization is influenced by the objectives of the particular organization, the product, service and the practices specific to the organization. The quality system, therefore, varies from one organization to another and must suit that organizationÕs operation. Despite the clear definitions and objectives of the ISO 9000 discussed above, there is considerable confusion and frustration surrounding the

0377-2217/03/$ - see front matter
2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 7 - 2 2 1 7 ( 0 2 ) 0 0 2 5 2 - 7

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

applied role and business value of ISO 9000 certification. The business value of certification has varied from unprecedented successes to an increasing workload and Ôcost of doing businessÕ. An extensive literature review revealed major gaps in research in this area of operations/quality management. Juran (1999, p. 30) summed it up very well in Quality Progress by stating that: . . .there is a lack of research that can shed light on whether companies that are certified to the ISO 9000 standards actually produce better products than companies that have not been certified. . .we donÕt really know what benefit we are getting from all that costly certification to the ISO 9000 series of standards. We should establish the researches needed to find out This paper, therefore, presents the findings from a cross-sectional study of manufacturing organizations that had been certified to the ISO 9000 standard, and gauge their motives for certification. Furthermore, third party quality auditors were asked to respond to a separate questionnaire which sought the auditorÕs perceptions of ISO 9000 practice and its effectiveness on business performance. Several hypotheses are developed from an extensive literature review.

2. Literature review The literature is reviewed in order to establish the leading edge research on the role of the quality auditor and the ISO 9000 certification process. The review was used to articulate questions, hypotheses, and design the questionnaires. 2.1. The effectiveness of the ISO 9000 standard on business performance From the literature, there appears to be a relationship between the managerÕs motives for adopting certification and the ensuing experience as identified by the Science and Engineering Policy Studies Unit in the UK (SEPSU, 1994). Those

581

organizations that pursue ISO 9000 certification willingly and positively across a broad spread of objectives are more likely to report improved organizational performance. Customer pressure, however, was the most commonly cited motivating factor for pursuing ISO 9000 certification. These companies were less likely to report improved organizational performance (Brecka, 1994; Allan, 1993; Najmi and Kehoe, 2000). A large study funded by the Australian Government and conducted by the Australian Manufacturing Council (AMC), AMC (1992) study cautioned firms that ISO 9000 certification is not a substitute for delivering high quality products and services as defined by customer needs. The AMC study gathered quantitative data from a large random sample in a mail survey of manufacturing site managers in Australia and New Zealand. The purpose of study was to investigate the extent of Ôbest practiceÕ adoption by Australian and New Zealand manufacturing firms and its impact on organizational performance. Furthermore, the study claimed that ISO 9000 certification could cause managers to become complacent about quality. There is the risk of Ôlocking inÕ or systematizing some poor practices through the certification process. Ideally, organizations should set themselves quality standards well above the minimum prescribed by ISO 9000 standards, and constantly seek ways to improve all facets of the operation. The AMC study concludes that certification is likely to lead to both actual and perceived quality improvements, as well as overall improvements in organizational performance. The other extreme of the mainstream literature is mainly anecdotal. Several large ISO 9000 consulting firms have produced large in-house studies for their clients, but most have released only their results, and not their methodologies. These studies found that the primary business value of ISO 9000 certification was to open doors to markets which were previously closed (Brecka, 1994; Seddon, 1993; Miller, 1993; IQA, 1993). For example, a study commissioned by LloydÕs Register Quality Assurance (LRQA, 1993) found that most of the benefits associated with ISO 9000 certification were external to an organizationÕs processes. Sixty-nine per cent of the managers

582

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

agreed that ISO 9000 certification enabled them to improve their business performance by allowing them to bid for tenders. Batchelor (1992) generally contradicts the literature findings. He studied 647 certified manufacturing and service organizations and found that only 15% of the respondents had derived business value on nine dimensions of organizational performance. These dimensions were: market share, new customers, customer satisfaction, procedural efficiency, staff motivation, staff attitudes, error rates, wastage and costs. Juran in Quality Progress (1999, p. 30), summed up the perceived effectiveness and shortcomings of the existing ISO 9000 standards. He said: . . .companies that are the beginning stages of their quality journeys find that the ISO 9000 series of standards provides them with a guide for implementing a basic quality system. But for companies with good quality systems, the standard often just add costs, delays and burdensome documentation, rather than providing any competitive advantage. . .the focus today should be on quality leadership-to be competitive with the rest. To achieve such leadership demands much more than conformance to the ISO 9000 standards. The design of the new ISO 9001-2000 standard has addressed concerns from commentators such as Juran. The new standard attempts to provide changes on a number of fronts: • ISO 9001-2000 will offer enhanced relevance to the service sector. • Greater implementation flexibility. ISO 9001 is good at identifying ÔwhatÕ needs to be done, and very weak in terms of ÔhowÕ to do it. The ISO 9001-2000 standard addresses this issue by providing an integrated model, which includes the customer, performance management. • ISO 9001-2000 aims to provide a bridge between ISO 9001 and enhanced quality management practices (e.g., continuous improvement). • An improved synergy between ISO 9001-2000 and ISO 14000 (the environmental management system requirements) will set the stage for an organizationÕs Integration of Quality, environ-

mental and other management system goals and objectives. 2.2. The role of the quality auditor in the ISO 9000 certification process Hutchins (1993) claims that quality auditors frequently are unfamiliar with the clientÕs industry, quality system, and process or products/services. This inevitably results in a poor quality audit, which places conformity at risk. He suggests three ways to improve quality auditing standards: • Government regulations for qualifying auditors. • Self-regulation by qualifying organizations. • Public groups self-certify and train auditors to conduct industry, process, and or product/service specific audits. Furthermore, Williamson et al. (1996) claim that there is a poor overall understanding of the current philosophy regarding the proper conduct of quality audits. They believe that auditors give more attention to whether the clientÕs quality system is effectively documented and less to the outcomes of the system. This view is shared by Chan et al. (1993) who argue that the new role of the quality auditor should contribute to the improvement of the organizationÕs overall quality management system. Fiorentino and Perigord (1994, p. 61) go further and define the function of the quality auditor as on the job continuous training. They believe that as companies are forced to change with altered business climates, so quality auditors must take on a new role that emphasizes continuous improvements throughout the organization. The role of the auditor according to Sakofsky (1993) is critical. This author believes that as more organizations have become certified the standard of quality auditing has improved with bettertrained and qualified auditors being able to provide a more facilitative auditing style. As a result many organizations are beginning to see that the quality audit can be used as an improvement tool rather than a feared report card. In a survey of 274 quality auditors in the UK, Williamson et al. (1996) found evidence of

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

implementation of corrective actions as providing the strongest indicator of an effective quality system. Russell and Regel (1996, p. 66) also place great emphasis on this area of the audit process, stating that they believe that the key to effective auditing lies in the phase between audit performance and following through on the results of the audit. This would appear to be in conflict with the standard that clearly states that the quality audit (and the auditorÕs involvement) is complete when the report is submitted (although it does not specifically preclude further involvement of the auditor after this stage). Bettes (1993), on the other hand, takes the view that this is the job of the internal quality auditor to develop a culture of ongoing improvement in part through effective follow up of corrective actions. Najmi and Kehoe (2000) report on their quantitative analysis of 221 firms and 18 qualitative case studies of companies within the UK manufacturing sector. The researchers base their analysis on a framework developed from an extensive review of the literature named Post-ISO 9000 Quality Development. This framework represents and integrated approach in which a TQM strategy is formulated and then implemented within an organization after ISO 9000 certification has taken place in order to improve performance measures. Najmi and Kehoe found as expected that benefits along the bottom line were substantially improved for companies, which had attempted to at least a moderate degree, a Post-ISO 9000 quality development strategy. Those companies, which had stopped short at simply pursuing ISO 9000 certification, were significantly lagging in the performance dimensions of quality, time and financial aspects. Such research concludes that ISO 9000 is an important starting point for the development of a more meaningful result based strategy such as TQM. This conclusion is also supported by Pheng and Alfelor (2000) who insist on the importance of not only meeting the technical requirements of ISO 9000, but also the behavioral approaches within a TQM system. 2.3. Quality culture While there are many definitions of culture, Schein (1985), Saraph (1987), Kotter and Heskett (1992) have indicated that quality culture should

583

focus on having a collective or shared learning of quality-related values as the organization develops its capacity to survive in its external environment. Values are formed in a group over time and can be seen in the ideology, philosophy, charter, or basic credo of an organization. These values, in turn, result in artefacts and creations of the group, which can be easily detected by observing the organizationÕs managerial processes, systems, and human relationships. Quality culture formation occurs through integrated changes in the organizational system; an organizational quality-based vision, mission and goals, consistent formal and informal organizational structures, compatible reward systems, appropriate technology and job design, and attention to important personnel issues. Going beyond previous empirical work on this topic, Kotter and Heskett (1992) provide a comprehensive critical analysis of how the ÔcultureÕ of an organization influences organizational performance. Kotter and Heskett define organizational culture as having two levels, which differ in terms of their visibility and their resistance to change. At the deeper and less visible level, culture refers to values that are shared by the people in a group and that tend to persist over time even when group membership changes. At this level culture can be extremely difficult to change, in part because members are often unaware of many of the values that bind them together. At the more visible level, culture represents the behavior patterns or style of an organization that new employees are automatically encouraged to follow by their fellow employees. Kotter and Heskett (1992, p. 2) stated that: Culture, in this sense, is still tough to change, but not nearly as difficult as at the level of basic values. Based on qualitative research at such firms as Hewlett-Packard, Xerox, ICI, Nissan, as well as a quantitative study of the relationship between culture and performance in more than 200 companies, the authors conclude that shared values and unwritten rules can profoundly enhance economic success, or conversely, lead to failure to adapt to changing markets and environments. The

584

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

study shows that organizations that have participative corporate cultures and well-organized workplaces have better performance records than those that do not. Kotter and Heskett (1992) develop the argument that corporate culture is strongly influenced by the characteristics of the industry in which the company operates. Thus, companies within an industry share certain cultural elements that are required for survival. Collins and Porras (1994), in their popular book ‘‘Built to Last: Successful Habits of Visionary Companies’’ examine 18 exceptional and longlasting companies in the USA, and studied each one in direct comparison with one of its competitors. The average age of these companies is about 100 years. These companies have outperformed the stock market by a factor of 15 since 1926. Collins and Porras (1994) argue that just about anyone can be a key protagonist in building an extraordinary business institution by learning the lessons of best practice companies. The myth that the future of a company depends on whether it is led by people ordained with rare and mysterious qualities that cannot be learned by others, is completely dismissed by the findings of this study. The authors looked for historical consistency through multiple generations of chief executives, instead of relying on the companyÕs most recent values, mission, vision, or purpose statement. According to the authors, the visionary companies do not just declare an ideology; they also take steps to make the ideology pervasive throughout the organization and transcend any individual leader. Several common themes emerged from the study that visionary organizations share. The core ideologies for six of the best companies studied by Collins and Porras (1994) were: • Create strong cultures around the ideology. • Select senior management based on fit with a core ideology. • Attain more consistent alignment with a core ideology in such aspects as goals, strategy, tactics, and organization design. • Focus on continuous improvement. • Serve the customer above all else. • Promote ÔtrustÕ and respect for the individual.

2.4. Employee attitudes and quality culture Perhaps the greatest challenge in creating a quality culture comes with the awareness that managing the change process includes managing employee attitudes towards that change. Unfortunately change is frequently introduced without considering its psychological effect on the employees in the organization – particularly those employees who have not been part of the decision making and planning process to make the change. According to Jick (1993): . . .if the reactions to change are not anticipated- and managed-the change process will be needlessly painful and perhaps be unsuccessful. Hind (1996) argues that a positive attitude towards change is key to ISO 9000 success. For many, conformance to ISO 9000 standards is so strong that this itself creates inflexible attitudes which can often become a barrier to the empowerment associated with a quality culture. Hind argues that the effort and time taken to conform to ISO standards may create negative attitudes towards those whose effort implemented such a system – unfortunately typically top management. Such an argument raises the need for research considering the order of ISO 9000 and quality culture implementation as well as the type of quality culture which characterizes successful ISO 9000 organizations (Najmi and Kehoe, 2000).

2.5. Literature synthesis The literature review has revealed that the benefits attributable to ISO 9000 certification were mainly for procedural efficiency and error rates, and less likely for market share, staff motivation and costs. Our study aims to resolve some of the contradictions in the literature by providing empirical evidence on the relationship between the ISO 9000 certification process and business performance and whether or not the role and style of the quality auditor has any impact on the strength of the relationship.

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

It appears from the literature review that quality auditors are unfamiliar with the clientÕs industry, quality system, and process or products/ services. This inevitably results in a poor quality audit, which places conformity at risk. The role and style of the quality auditor has become critical. Quality auditors are expected to provide a more facilitative auditing style. As a result many researchers in this field and case studies of certified organizations are beginning to see that the quality auditor can play a crucial role in improving the performance of the organization by adopting a more facilitative style, rather than a policeman style which focuses on non-conformance to the ISO 9000 standard. There is consistency in the literature on the characteristics of quality culture formation. Several authors conclude that quality cultures are not created by a single factor but rather by an integrated change in several factors such as the organizational system; an organizational quality-based vision, mission and goals, consistent formal and informal organizational structures, compatible reward systems, appropriate technology and job design, and attention to important personnel issues. Considering the above syntheses of the literature, it appears that the effectiveness of ISO 9000 certification depends on many factors such as the existing quality culture of the organization (H1), the motivation for ISO 9000 certification (H2); the style of the quality auditor (H3). Several previous studies have found H2 to be true. However, these studies are typically descriptive and do not adopt a level of standard methodological rigour such as this study. We test the hypotheses using a large sample size, and then we use qualitative data to further explain the hypothesized relationships. Terziovski et al. (1997) used ANCOVA and MANCOVA to test relationship between ISO 9000 certification and organizational performance covaried for company size. The results showed that company size does have an effect on the hypothesized relationships. However, the study did not gauge perceptions from auditors, and the sample size consisted of only manufacturing companies. This study focuses on all industry sectors of companies certified to ISO 9000 and the

585

auditing fraternity of the largest third party certifier in Australia, Quality Assurance Services. The covariates, company size and number of years an organization has been certified may also have an effect on the hypothesized relationships. For example, larger certified organizations may have greater success with ISO 9000 certification due to the larger level of resources available to them to pursue the implementation process. Hypotheses Based on the synthesis of the literature, the following hypotheses have been articulated for testing in this study: H1. There is a significant and positive relationship between the quality culture of an ISO certified organization and the benefits derived from the certification process. H2. The motivation for ISO 9000 certification is a significant and positive predictor of the benefits derived from certification. H3. A facilitation style of the quality auditor, has a significant and positive relationship with the benefits derived from ISO 9000 certification. H4. The covariates company size, and the number of years an organization has been certified have an impact on the strength of the relationship in H1, H2, and H3.

3. Methodology 3.1. Survey instruments Two survey instruments were developed to gather quantitative data. The first instrument was completed by a sample of certified companies and the second instrument was completed by a sample of practicing quality auditors. The review of the literature provided the theoretical basis for the development of the major constructs contained in the questionnaires. The survey instrument administered to the certified companies was divided into four main areas. These are ‘‘quality culture’’, ‘‘motivation for certification’’, ‘‘auditing style’’

586

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

and ‘‘value of certification’’. With respect to ‘‘quality culture’’, a total of 30 separate items were included in the survey instrument. These items are shown in Tables 1–5. With respect to ‘‘motivation for certification’’ respondents were asked to rate a range of reasons according to the level of relevance to their organizationsÕ decision to seek ISO 9000 certification.

These included two sets of items (total of 10 items) relating to the intrinsic or extrinsic nature of the motivation (i.e., continuous improvement strategy versus reaction to external factors) (see Tables 6 and 7). ‘‘Auditing style’’ was covered by 22 items including the following categories (see Tables 8–10):

Table 1 Quality culture factor 1 Alpha ¼ .9363 Factor score

Factor one – resources provided for quality Item Adequate Adequate Adequate Adequate

resources resources resources resources

are are are are

provided provided provided provided

for for for for

quality quality quality quality

improvement improvement improvement improvement

to to to to

senior management middle management front line management general staff

.841 .848 .842 .812

Table 2 Quality culture factor 2 Factor two – quality awareness Item

Alpha ¼ .8468 Factor score

Attitude of senior management toward the comprehensive use of quality management practices Attitude of middle management toward the comprehensive use of quality management practices Attitude of front line management toward the comprehensive use of quality management practices Attitude of general staff toward the comprehensive use of quality management practices Senior management strongly support the certification audit Responsibility for quality is allocated to all organizational members Quality systems and procedures are clearly understood throughout the company

.714 .717 .685 .640 .589 .452 .524

Table 3 Quality culture factor 3 Alpha ¼ .9156 Factor score

Factor three – bencmarking for quality improvement Item Benchmarking Benchmarking Benchmarking Benchmarking

is is is is

used used used used

extensively extensively extensively extensively

to to to to

improve improve improve improve

quality quality quality quality

of of of of

products/processes products/processes products/processes products/processes

by by by by

senior management middle management front line management general staff

.806 .863 .855 .765

Table 4 Quality culture factor 4 Factor four – customer focus Item

Alpha ¼ .7581 Factor score

There is a recognition throughout the organization of the needs and expectations of external customers There is a recognition throughout the organization of the needs and expectations of internal customers There is a recognition throughout the organization of the needs and expectations of suppliers

.475 .592 .677

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

587

Table 5 Quality culture factor 5 Factor five – quality as a measure of performance Item

Alpha ¼ .7362 Factor score

Quality plan is an integral part of the strategic plan Quality indicators are used as key performance indicators The organization has formal program for employees at all levels to contribute to quality improvement Data relating to quality improvement is shared openly with customers

.561 .710 .448 .507

Table 6 Motivation for certification factor 1 Factor one – continuous improvement strategy Item

Alpha ¼ .8004 Factor score

Part of a TQM based strategy Pursuit of continuous process improvement Desire to provide better products and processes Recognition of the need to formalise systems

.761 .881 .764 .725

Table 7 Motivation for certified factor 2 Factor two – reaction to environmental factors Item

Alpha ¼ .5983 Factor score

Changes in legislation Changes in industry policy/trends Demand by customers for supplier certification Pressure to follow competitors Changes in social trends

.637 .754 .469 .647 .574

Table 8 Auditing style factor 1 Factor one – auditing to facilitate continuous improvement Item

Alpha ¼ .8640 Factor score

The The The The The The The The The

.740 .726 .658 .690 .635 .673 .723 .725 .527

• • • •

role of the auditor is to help companies improve their people, processes, products and services role of the auditor is to help people acquire a common understanding of the system auditor should be flexible and able to adjust his/her style to meet the requirements of the client auditor should work with the client by being a good communicator and an excellent listener auditor should highlight the positive as well as the negative aspects of factors being audited ability to facilitate change is critical for the non-financial auditor ability to facilitate continuous improvement is critical for the non-financial auditor ability to adapt to the needs of the client is critical for the non-financial auditor audit should remove barriers to continuous improvement

The The The The

purpose of the non-financial audit. role of the non-financial auditor. style of the non-financial auditor. skills of the non-financial auditor.

‘‘Value of certification’’ was tested with 13 items covering issues such as productivity, waste reduction, internal systems, sales and customer satisfaction (see Tables 11 and 12).

588

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

Table 9 Auditing style factor 2 Factor two – auditing for compliance Item

Alpha ¼ .8086 Factor score

The audit should solely verify performance against the standard The auditor should solely assess compliance with applicable standards It is the role of the auditor to solely inspect the clientÕs systems The auditor should only be interested in checking the operation of the system

.688 .837 .779 .771

Table 10 Auditing style factor 3 Factor three – auditing style that is detached and inflexible Item

Alpha ¼ .6581 Factor score

The auditor should remain detached and not get involved with the client during the audit The auditor should focus on non-conformances when reporting to the client at the conclusion of the audit The auditorÕs style should not be adjusted for different clients

.814 .665 .713

Table 11 Value of certification factor 1 Factor one – improved business performance Item

Alpha score ¼ .9092 Factor score

Improved customer satisfaction Increased productivity Improved service quality Improved product quality Improved company image Increased flexibility Increased sales Improved organizational systems and documentation Improved supplier quality Improved cycle times

.693 .755 .759 .660 .671 .771 .798 .348 .675 .599

Table 12 Value of certification factor 2 Factor two – reduced waste Item

Alpha ¼ .8705 Factor score

Reduced scrap Reduced obsolete stock

.903 .893

This survey resulted in 400 companies responding (a response rate of 27%). A total of 300 quality auditors were surveyed including 263 auditors employed by Quality Assurance Services, a subsidiary of Standards Australia and 37 quality auditors operating in New Zealand. The quality auditors survey resulted in 126 responses (response rate of 42%).

3.2. Samples 3.3. Demographics A random sample of 1500 companies was selected for the client survey from a database of 7000 companies that had been certified to ISO 9000.

Some notable aspects of the demographic breakdown for the quality auditors include:

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

• High proportion of male auditors; the majority come from manufacturing backgrounds. • Large percentage with tertiary qualifications. For the companies, it was also interesting to note that: • One quarter of managers responsible for quality auditing were at CEO/Managing Director/General Manger level, and • The apparent maturity of the market as indicated by the small percentage certified for less than two years. 3.4. Data analysis Factor Analysis was used to test the validity of constructs, and to establish a group of higher order factor variables reflecting latent factors to be used in subsequent Multiple Regression Analysis. Hierarchical Regression Analysis (HRA) was used with three separate models to test for the contribution to variance at points of entry. The statistical software analysis package SPSS was used for all major statistical analysis. As part of the data screening and preparation process the following actions were taken: Within the data set for each variable missing values were defined using the ‘‘Define Variable’’ function within SPSS so as not to include them in the analysis. In order to meet the requirement of successful multivariate analysis for a complete data set, mean substitution was used for missing values in both the factor analysis and multiple regression. This was deemed appropriate as the percentage of missing data across the full data set was 1.23%. Each individual variable was screened for the presence of values in excess of 3 S.D. Mahalanobis distances were calculated for each case and derived from all of the identified variables of interest to check for the existence of multivariate outliers. A case wise plot of standardized residuals was produced with each regression model in order to further identify and eliminate multivariate outliers; Where appropriate these outlying values were deleted. All the variables selected for analysis used interval scales and as such were also analyzed for excessive skew and kurtosis to ensure normality.

589

Transformations were carried out where appropriate. During regression analysis, a histogram of standardized residuals along with a cumulative probability plot of standardized residuals was generated to check that all linear combinations of variables were normally distributed. Prior to factor analysis scatter plots were produced for variables showing large differences in skewness to indicate a degree of confidence regarding linearity. For the regression models scatter plots of predicted scores versus residuals were also produced to assess linearity of the relationships being measured and to check for homogeneity of variance. During factor analysis correlation matrices were created, and a determinant of correlation matrix statistic generated to test for multicollinearity and singularity. During multiple regression a tolerance statistic was also produced testing the amount of variance in each independent variable attributable to the other independent variables. High tolerance values indicated low multicollinearity. The process followed for factor analysis employed an exploratory model, with a minimum eigenvalue of 1 specified, a principal components extraction employed and a varimax rotation applied. The final factors extracted were also subjected to reliability testing with alpha coefficients produced for each. In order to regulate the development of the regression process, a hierarchical (or sequential) regression model was developed. This was structured in order to test the amount of variance caused by sets of independent variables after variance due to other independent variables had already been accounted for. Factor analysis. Within the groups of variables identified above factor analysis was run and the following combinations extracted that could be used in subsequent analysis as independent and dependent variables. This provided a simplified group of factor variables capturing latent constructs facilitating less complicated analysis and model creation. Quality culture: KMO measure of sampling adequacy BartlettÕs test of sphericity

0.893 0.000

590

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

The KMO test being greater than 0.6 indicated that factor analysis was appropriate for this matrix, while the high degree of significance in the BartlettÕs test indicates that the matrix is not an identity matrix (i.e., the diagonal values are 1 and the off diagonals are 0). Five factors were identified from this analysis: Factor 1: Resources provided for quality construct. The combination of high factor scores and a high alpha coefficient provided a good degree of confidence that this scale had good construct validity and is a reliable measure. Due to the high factor scores on all four items (Table 1) the correlations were checked for multicollinearity. This check revealed some high correlations, but nothing higher than the nominal cut off point of 0.90 (Tabachnick and Fidell, 1996). It was also taken into account that factor analysis is particularly robust to most underlying statistical assumptions (such as normality, linearity, etc.), and that a certain amount of multicollinearity is desirable for good factor analysis (Hair et al., 1998). It was further noted that deletion of any of these individual items reduced the value of the alpha coefficient, thus weakening the reliability of the measure whilst at the same time reducing the strength of the overall construct. Factor 2: Quality awareness construct. The high alpha score for this factor again provided good confidence as to the reliability of the scale. The spread of the factor scores (Table 2) gave a wide range of factor scores, with the lowest being .452. Given that the sample size was greater than 350, a factor score greater than 0.30 could be interpreted as significant in both practical and statistical terms (Hair et al., 1998). Factor 3: Benchmarking for quality improvement construct. Due to the high factor scores (Table 3) correlations between these items were also checked for multicollinearity with no values found to be higher than 0.90. For the same reasons as stated above for factor one the four items were left intact. The high alpha score also gave confidence as to the reliability of the scale. Factor 4: Customer focus construct. The factor scores (Table 4) for this grouping were deemed adequate as per the criteria noted above, and the good alpha score indicated adequate reliability.

Factor 5: Quality as a measure of performance construct. The combination of a good alpha score and significant factor loadings (Table 5) was also evident for these items. Motivation for certification: KMO measure of sampling adequacy BartlettÕs test of sphericity

0.710 0.000

The KMO test being greater than 0.6 indicated that factor analysis was appropriate for this matrix, while the high degree of significance in the BartlettÕs test indicates that the matrix is not an identity matrix. Two factors were identified: Factor 1: Continuous improvement strategy construct. As per the criteria detailed above this group of items provided a good balance of reliability and construct validity (Table 6). These items also represented a group of ‘‘intrinsic’’ motivators for certification that provided a neat contrast with the other variables in this section as they represented a set of ‘‘extrinsic’’ or ‘‘reactive’’ reasons for seeking certification. Factor 2: Reaction to environmental factors. The low alpha coefficient for this factor presented the possibility of deleting an item in order to improve reliability (Table 7). It was decided to leave the factor intact as the best possible result from deletion of an individual item was an alpha value of .6024. This marginal improvement in reliability was not deemed sufficient to offset the reduction in the strength of the construct that would result. Auditing style: KMO measure of sampling adequacy BartlettÕs test of sphericity

0.853 0.000

The KMO test being greater than 0.6 indicated that factor analysis was appropriate for this matrix, while the high degree of significance in the BartlettÕs test indicates that the matrix is not an identity matrix. It should be noted that the questions answered in the survey of clients of nonfinancial auditors used in this analysis related to perceptions of current audit practice. Three factors were identified:

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

Factor 1: Auditing to facilitate continuous improvement construct. This group of variables captured a combination of issues relating to the perception of the purpose of the audit, the appropriate style of the auditor and the skills required to conduct a non-financial audit. The combination of the high alpha coefficient and the factor scores (Table 8) provided a good level of confidence in the reliability and validity of this scale. Factor 2: Auditing for compliance. These variables related to the purpose of the audit and the appropriate role of the auditor. The reliability and validity of the scale are reinforced by the alpha value and factor scores (Table 9). Factor 3: Auditing style construct. These three style related variables represent a detached and non-facilitative style of auditing. The alpha score of .6581 was deemed to be adequate as deletion of any of the items substantially weakened the score (Table 10). Value of ISO 9000 certification: KMO measure of sampling adequacy BartlettÕs test of sphericity

0.911 0.000

The KMO test being greater than 0.6 indicated that factor analysis was appropriate for this matrix, while the high degree of significance in the BartlettÕs test indicates that the matrix is not an identity matrix. The variables contained in this grouping related to the perceived contribution of ISO certification to a range of business performance outcomes. Two factors were identified: Factor 1: Improved business performance. The alpha value and the majority of the factor scores (Table 11) are good for this group of items. The lower factor score for the item relating to the improvement in organizational systems and documentation was deemed adequate based on the sample size being greater than 350 (Hair et al., 1998). Factor 2: Reduced waste construct. Due to the high factor scores these two variables were checked for multicollinearity and found to have a correlation less than 0.90. The high alpha value was noteworthy for a scale with only two com-

591

ponents, and again provided a good level of confidence in scale reliability (Table 12). 3.5. Summated scales For each factor a single new variable was created for the purpose of creating a summated scale. Apart from simplifying further analysis the use of a single composite measure of this type was preferred for the following reasons (Hair et al., 1998). A means of overcoming the inherent measurement error found in all measured variables is provided. The underlying assumption is that the factors capture the particular constructs with an acceptable level of validity and reliability as demonstrated in the five cases above. 3.6. Hierarchical regression models The rationale for using hierarchical regression models derived from a desire to test the nine hypotheses developed with a method that would allow for comparison between the contribution of different independent variables, whilst controlling for specified demographics. The advantage of this approach lies in the fact that: Sequential regression allows the researcher to control the advancement of the regression process. Importance of IVÕs in the prediction equation is determined by the researcher according to logic or theory. Explicit hypotheses are tested about proportion of variance attributable to some IVÕs after variance due to IVÕs already in the equation is accounted for Tabachnick and Fidell (1996, p. 156). In our case the independent variables were finalized from the literature review and from qualitative data derived from the case study analysis. The models developed were based on a theorized logical relationship between the independent variables, with independent variables being entered into the models as groups. A set of control variables were entered first in both models in order to extract the variance due to organization size, industry, number of years certified and the position held by the respondent.

592

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

These were then followed by the quality culture group, motivation for certification and auditing style. The logic in this model being that after organizational demographics have been accounted for, quality culture will precede the motivation for certification, which will precede auditing style. The dependent variables in this model were the two factors relating to value derived from certification and the five individual variables relating to attitudes to the audit process, and the model seeks to test the effect of quality culture, motivation for certification and auditing style on business outcomes and organizational attitudes.

4. Testing of hypotheses The two measures used for comparison of results are change in R2 and the final beta coefficient of each combination for the full model. The change in R2 gives a comparison between subsequent groups of factors of the effect of that group on the dependent variables after the variance attributable to preceding groups has been extracted. Both the strength and significance of the change is noted. The final beta value gives the strength and significance of the relationships between the individual elements of the model when all are included. An adjusted R2 value for the full model is also recorded to evaluate the total per cent contribution of all the independent variables to the change in the dependent variables. Analysis of the results using the change in R2 shows that there are significant changes in contribution at each of Steps 1, 2 and 3. For Step 1 changes of 3% significant at p < :05 for DV1 (improved business performance) and 3.2% significant at p < :05 are recorded. This indicates that the control variables do have some impact on the two business performance variables, and can thus not be ignored as factors affecting performance in the context of this model. For DV1 the final beta values show no individual items in this group having significant or strong relationships, while for DV2 (reduced waste) number of years certified makes the most significant contribution with 0.180 at p < :001. This indicates that over time certified companies

go through a process of organizational learning that has the effect of reducing waste, and perhaps points to certification as a catalyst for this process. For DV1 (improved business performance) quality culture shows a 9.5% increase in variance at p < :001. The final beta scores show that of the five elements of quality culture the strongest and most significant contributor to this variance was customer focus with 0.104 at p < :05. The dominance of customer focus in contribution to improved business performance is not entirely surprising, although it should be noted that the five elements of quality culture are not necessarily mutually exclusive. It is hard to imagine that a high degree of customer focus can exist without resources being provided for quality and there being a poor awareness of quality related issues in an organization. The interdependence and interrelatedness of the elements of quality culture is therefore an underlying assumption of the model. Motivation for certification shows a further 8.3% change at p < :001 for DV1, with the final beta scores indicating that both elements make a significant contribution (being 0.234 at p < :001 for continuous improvement strategy and 0.174 at p < :01 for reaction to environmental factors). The relatively stronger and more significant contribution of the pro-active and intrinsic motivator (continuous improvement strategy) over the reactive and extrinsic element (reaction to environmental factors) supports some anecdotal evidence of links between this type of motivation and ISO related benefits. What is of interest is the fact that both types of motivation are seen to contribute to business performance, suggesting perhaps that both elements are required. It could be argued that awareness of customer requirements and changes in the external environment of an organization can prompt the recognition of the need for continuous improvement strategies. The importance of the combination of these two forms of motivation, and perhaps their interdependence, is being highlighted by the model. For DV2 (reduced waste) quality culture does not show a significant contribution while motivation for certification does with 3.1% at p < :01: The final beta values also show no significant correlations between any of the individual

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

593

Table 13 Hierarchical regression model results Independent variables

Dependent variable: value of certification – factor 1 – improved business performance

Dependent variable: value of certification – factor 2 – reduced waste

All companies

All companies

N ¼ 395

N ¼ 395 2

Change R Step 1

Control variables

.030*

Quality culture Quality culture – factor 1 provided for quality Quality culture – factor 2 Quality culture – factor 3 quality improvement Quality culture – factor 4 Quality culture – factor 5 measure of performance

Step 3

Final beta

)0.041

)0.025

0.016 )0.062 )0.002

)0.083 0.180*** )0.055 0.023

.095*** )0.005

0.031

– quality awareness – benchmarking for

0.071 0.094

0.020 0.032

– customer focus – quality as a

0.104* 0.017

– resources

Motivation for certification

.083***

Reasons for certification – factor 1 – continuous improvement strategy Reasons for certification – factor 2 – reaction to environmental factors Step 4

Change R2 0.032*

No. of people employed by the organization in Australia Organizational Category No. of years certified Position in organization Step 2

Final beta

Auditing style Current audit practice – factor 1 – auditing to facilitate continuous improvement 1 Current audit practice – factor 2 – auditing for compliance Current audit practice – factor 3 – auditing style that is detached and inflexible

elements of quality culture and DV2, a result that on the face of it is surprising given the expected link between reduction in waste and improved business performance. Beta values for motivation for certification with DV2 show that continuous improvement strategy is the dominant factor in contributing to waste reduction with 0.180 at p < :01:

0.028 )0.017

0.031** .234***

0.180**

.174**

0.013

0.006

0.005 0.086

0.075

0.004

)0.005

0.025

0.028

Auditing style shows no significant effect on either dependent variable, and the individual beta scores also show no significant relationships between the elements of auditing style and both dependent variables. All significant correlations were positive in direction. The results from this model for these hypotheses are contained in Table 13. Adjusted R2 values for DV1 (improved business

594

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

performance) was 0.186 and for DV2 (reduced waste) was 0.056.

5. Interpretation of results Based on the analysis several important findings have emerged. There is a relationship between the control variables and business performance, particularly in the case of waste reduction and the number of years an organization has been certified. This relationship does not appear to be overly strong, but is nonetheless significant and therefore a factor that needs to be taken into account when assessing the importance of other relationships. The quality culture variables identified through the factor analysis have a highly significant, if moderately strong, effect on business performance in this model. The individual element found to contribute most to this was customer focus. There is no link recorded between quality culture and reduction in waste in the model. Motivation for certification has a highly significant and moderately strong effect on business performance, and highly significant and weak effect on reduction of waste. Both elements contribute significantly to business performance with continuous improvement strategy being marginally stronger, while this same element is also the key determinant of reduction in waste. Results from this model show there is no significant relationship between auditing style and business performance. The final value for adjusted R2 indicates a higher contribution overall of the independent variables to improved business performance compared to reduction of waste. In terms of the hypotheses proposed these results indicate that: H1. There is a positive relationship between the quality culture of an ISO certified organization and the benefits derived from certification – hypothesis accepted. H2. The motivation for certification can be used as a predictor of the value derived from certification – hypothesis accepted.

H3. The style of the non-financial auditor is positively related to the benefits derived from certification – hypothesis rejected. H4. The relationships in H1, H2, and H3, are independent of the influence of company size and the number of years an organization has been certified – hypothesis rejected.

6. Conclusion Our findings show that the motive for adopting ISO 9000 certification and the maturity of the quality culture are significant predictors of the benefits derived from ISO 9000 certification. The style of the auditor, on the other hand, does not appear to have a significant and positive effect on the benefits derived from ISO 9000 certification. We conclude that organizations that have been audited to the ISO 9000 standards believe that the quality audit process contributes to business performance when the quality culture in the organization is well developed and the managerÕs motivation to gain certification has been to improve business performance and not to conform to an international standard. The individual element found to contribute most to business performance was customer focus. The principal motivation to pursue ISO 9000 certification was found to come from the customers. Auditing style shows no significant and positive effect on business performance. The findings of this study are in agreement with several other studies (Brown, 1995; Hoyle, 1994; Najmi and Kehoe, 2000).

7. Implications for managers The main implication for managers is the emerging question whether the new ISO 90012000 standard will over-time capture and meet the conformance and performance requirements of the organization as part of a continuous improvement strategy. This is likely be the prime determinant of the extent to which managers will

M. Terziovski et al. / European Journal of Operational Research 146 (2003) 580–595

embrace or reject ISO 9000 certification in the 21st century. For example, combining the discipline of the ISO 9001-2000 compliant system with the significant process improvements that are derived from programs such as Six Sigma and cycle-time reduction could facilitate rapid process improvement that becomes the Ôbest practiceÕ on which to build future organizational success.

Acknowledgements The authors wish to thank the Australian Research Council and Quality Assurance Services Pty Ltd., Standards Australia for providing the necessary funding for this research project.

References Allan, M.J., 1993. Implementation of ISO 9001/2 in Large Australian Manufacturers. Unpublished Research Project, The Melbourne Business School, The University of Melbourne. Australian Manufacturing Council Study, AMC Study, 1992. Leading the Way: A study of Best Manufacturing Practices in Australia and New Zealand, 1994, pp. 59–63. Batchelor, C., 1992. Badge of quality, Financial Times, 1 September. Bettes, D.C., 1993. Training of internal quality auditors. Industrial and Commercial Training 25 (7), 18–21. Brecka, J., 1994. Study finds gains with ISO 9000 registration increase over time. Quality Progress May, 18–20. Brown, A., 1995. Industry Experience With ISO 9000, Paper presented at the Second National Research Conference on Quality Management, Monash Mt Eliza Business School, February 1995. Chan, F.Y., Willborn, W., Xiao, H., Li, H., 1993. Research report: An expert system for quality management auditing. Asia Pacific Journal of Quality Management 2 (1), 65– 72. Collins, J., Porras, J., 1994. Built to Last: Successful Habits of Visionary Companies. Century, London. Fiorentino, R., Perigord, M., 1994. Going from an investigative to a formative auditor. Quality Progress 27 (10), 61–65. Hair, J.F., Anderson, R.E., et al., 1998. Multivariate Data Analysis. Prentice-Hall, Englewood Cliffs, NJ. Hind, M., 1996. Are the cultures required to attain ISO 9000 and total quality management mutually exclusive? Training for Quality 4 (2).

595

Hoyle, D., 1994. Quality Systems Handbook, 2nd Ed. Butterworth-Heinemann, Oxford. Hutchins, G., 1993. ISO 9000: A Comprehensive Guide to Registration, Audit Guidelines and Successful Certification. Oliver Wight Publications Inc., Essex Junction, Vermont. Institute of Quality Assurance (IQA), 1993. Survey on the Use and Implementation of BS 5750. Juran, J.M. (1999). Commentary in Quality Progress, June, p. 30. Jick, T.D., 1993. Managing Change: Cases and Concept. Irwin, Boston. Kotter, J.P., Heskett, J.L., 1992. Corporate Culture and Performance. The Free Press, New York. Lloyds Register Quality Assurance Limited, 1993. Setting Standards for Better Business, Report of Survey findings, London, pp. 1–10. Miller, C., 1993. US firms lag in meeting quality standards. In: Marketing News, 15, pp. 1–6. Najmi, M., Kehoe, D.F., 2000. An integrated framework for post-ISO 9000 quality development. International Journal of Quality and Reliability Management 17 (3), 226–258. Pheng, L.S., Alfelor, W.M., 2000. Cross-cultural influences on quality management systems: Two case studies. Work Study 49 (4), 134–144. Russell, J.P., Regel, T., 1996. After the quality audit: Closing the loop on the audit process. Reviewed in Quality Progress 29 (6), 65–67. Sakofsky, S., 1993. Twelve ways to make a quality audit work for you. Journal of Quality and Participation 16 (6), 30–32. Saraph, J.V., 1987. An investigation of quality-management practices in medium to large size companies, Unpublished Ph.D. Dissertation, The Faculty of the Graduate School of the University of Minnesota, April. Schein, E.H., 1985. Organizational Culture and Leadership: A Dynamic View. Jossey-Bass, San Francisco. Science and Engineering Policy Studies Unit, 1994. UK Quality Management-Policy Options, The Royal Society and the Royal Academy of Engineering, SEPSU Policy Study No. 10, June, pp. 1–99. Seddon, J., 1993. BS 5750 Implementation and Value Added, Vanguard Consulting Quality Systems Update: A Global ISO 9000 Forum and Information Service, ISO 9000 Survey: Registered Companies Report Big Savings, Fall, pp. 1–8. Tabachnick, B.G., Fidell, L.S., 1996. Using Multivariate Statistics. Harper Collins College Publishers, New York. Terziovski, M., Samson, D., Dow, D., 1997. The business value of quality management systems certification: Evidence from Australia and New Zealand. Journal of Operations Management 15 (1), 1–18. Williamson, A., Rogerson, J.H., Vella, A.D., 1996. Quality system auditorsÕ attitudes and methods: A survey. International Journal of Quality Reliability and Management. 13 (8), 39–52.