Exploring the relationship between organizational culture and software process improvement deployment

Information & Management 47 (2010) 271–281

Contents lists available at ScienceDirect

Information & Management journal homepage: www.elsevier.com/locate/im

Exploring the relationship between organizational culture and software process improvement deployment Chiao-Ching Shih, Sun-Jen Huang * Department of Information Management, National Taiwan University of Science and Technology, 43, Sec. 4, Keelung Road, Taipei 106, Taiwan

A R T I C L E I N F O

A B S T R A C T

Article history: Received 27 March 2009 Received in revised form 20 September 2009 Accepted 26 May 2010 Available online 4 June 2010

We explored the relationship between organizational culture and deployment of software process improvement (SPI) approaches using a competing values framework. Our results indicated that the organizational culture had an influence on SPI deployment, primarily made possible by a hierarchic culture with its emphasis on procedures, order, and stability. Clan culture, with its emphasis on human development, commitment to others, and participation, appears to be a necessary condition in creating skills development and sharing SPI knowledge in the process of its deployment. Software Engineering Program Group leaders should ensure that internal values are in place to enhance SPI deployment. ß 2010 Elsevier B.V. All rights reserved.

Keywords: Organizational culture Software process improvement (SPI) Competing values framework Clan culture Hierarchic culture

1. Introduction In recent years, software process improvement (SPI) has emerged as the dominant approach for delivering improvements to the software product in software development organizations. Its intent is to enhance software product quality, increase productivity, and reduce the cycle time for product development. A number of advances have been made in the development of SPI approaches such as ISO 9000, the Capability Maturity Model (CMM) and its newer versions: the Capability Maturity Model Integration (CMMI), and Software Process Improvement and Capability dEtermination (SPICE), which focus on defining and measuring processes and practices to achieve quality software. ISO 9000 certification and CMMI are used by software companies all over the world. They guide the process improvement throughout the project in a division, or part or the entire organization. CMMI helps adopters to integrate traditionally separate organizational functions, set process improvement goals and priorities; it also provides guidance for implementing quality processes and a reference model for appraising current processes. CMMI provides a staged representation with five levels of software process maturity ranging from initial (processes poorly controlled and reactive) to optimizing (focused on continuous process improvement) [11]. Despite the widespread adoption of SPI, there is still insufficient quantitative evidence of how software products have been

* Corresponding author. Tel.: +886 2 27376779; fax: +886 2 27376777. E-mail address: [email protected] (S.-J. Huang). 0378-7206/$ – see front matter ß 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.im.2010.06.001

improved by its deployment [3,4,6,16] and there is still a great deal of variability in the success of SPI initiatives [5,23]. A recent review of 322 papers on SPI [7] indicated that the field was dominated by one approach (CMM), and heavily biased towards how SPI practitioners can carry out SPI initiatives. Surveys indicated, however, that the SPI field lacked theoretical frameworks. SPI attempts to change how software professionals think and act in their everyday organizational activities. Therefore, its activities can result in organizational changes. Ravichandran and Rai [21] found that organizations face major hurdles in the implementing SPI and that these are more organizational than technological in nature. Several researchers [1,18] have also indicated that SPI does not deal effectively with the social aspects of organizations. Thus, it needs a managerial focus rather than a technical one. Hofstede regarded organizational culture as the collection of values, beliefs and norms shared by its members and reflected in its practices and goals. This can affect SPI deployment. Results of several studies, e.g. [2,8,10], have also suggested that organizational culture has a significant effect on both the successful implementation and the use of IT. Therefore, we decided to examine SPI approaches, specifically CMMI, to ascertain the influence that organizational culture has on SPI deployment. 2. Background SPI involves understanding existing processes and changing these processes to improve product quality and reduce cost and development time.

272

[(Fig._1)TD$IG]

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

2.1. Deployment of software process improvement approaches The deployment of SPI involves the post-implementation stages of the innovation diffusion process, when the innovation is being incorporated into the organization. It is necessary to focus on the deployment because the SPI may not be used effectively or may not have the intended consequences. The assessment of the deployment adopted by us included three items, as discussed next. 2.1.1. Perceived SPI support for software development This can be linked to user’s understanding of the capabilities and corresponding values to the organization [12]. The production technology in our study was defined as: the impact on the ability of a user in generating planning and design decisions and thus artifacts or products. The co-ordination technology was defined as: functionality that enabled or supported the interactions of multiple agents in the execution of a planning or design task; it consists of control and cooperative functionality. The control technology was defined as: the method of enabling the user to plan for and enforce rules, policies or priorities that will govern or restrict the team members during the process. The cooperative technology was defined as: the way of enabling the user to exchange information with others in order to affect the concept, process, or product of the team. 2.1.2. Perceived SPI impact on the quality of the software product and development process Examining product and process outcomes can reveal each of their impacts on quality. In addition, quality processes are a necessary prerequisite for delivering quality products and satisfying customer needs. Therefore, we concentrated on two dimensions: product quality (the overall evaluation of the product produced by the process) and development process quality [20]. Process quality in our study is defined as the degree to which the process is designed to promote consensus among participants in software development, operate within established resource parameters, and reduce waste and redundancy. The perceived impact of the deployment of SPI can be linked to the probable or actual consequences of its adoption. 2.1.3. The degree of SPI use There has been little research confirming the actual value of using SPI. Our study concentrated on two dimensions of SPI use: horizontal (the degree of penetration of SPI use measured as the percentage of software developers and projects using SPI knowledge) and vertical (the maximum intensity of SPI use within the organization). 2.2. Organizational culture An organization’s culture is its set of shared ideas and values that serve as a means of accomplishing its mission. As such, it can and does play an important role in many facets of the organization. The values define what is important to a group. They answer the question: Why people behave the way they do? Different approaches have been used to study how organizational values affect its culture. The national and organizational cultures represent the most popular approaches. They both define the values that distinguish one group from another. The most popular conceptualization of national culture has been Hofstede’s well-known taxonomy of using the dimensions of power distance, uncertainty avoidance, individualism, masculinity, and Confucian dynamism or long-term Orientation [24]. These allow national-level analyses and allow country or regional comparisons. At the organizational level, the competing values framework (CVF) is most popular. It allows of organizational cultural taxonomies has been to enable the differentiation of comparison of organizations along the dominant values of each organization’s behavior.

Fig. 1. Competing values framework of organizational culture.

The CVF is characterized by a two-dimensional space that reflects different value orientations, as shown in Fig. 1. The first dimension in this framework, the flexibility-control axis, shows the degree to which the organization emphasizes change or stability. A flexibility orientation reflects flexibility and spontaneity, while a control orientation reflects stability, control and order. The second dimension in this framework, the internal–external axis, addresses the organization’s choice to focus on activities occurring internally and those occurring outside the organization. An internal orientation reflects maintaining and improving the organization, while an external orientation reflects a competition, adaptation, and interaction with the outside environment. Thus four types of organizational culture appear: clan (which emphasizes flexibility, change and a focus on the internal organization), adhocracy (which also emphasizes flexibility, but it is externally focused, primarily on growth, resource acquisition, creativity and adaptation), hierarchic (which is externally focused, but is control oriented, dealing with productivity and achievement of well-defined objectives response to external competition.), and market (which emphasizes stability but focuses on the internal organization, its uniformity, co-ordination, internal efficiency and a close adherence to rules and regulations). Though the framework is divided into named quadrants with distinct characteristics, no organization is likely to reflect only one value system. Instead, one would expect to find combinations of values in one company, with some more dominant than others. Good fit between the values embedded in the software development process and the overall organization’s values lead to a more successful implementation. In a content analysis of longitudinal data from three SPI initiatives, Ngwenyama and Nielsen [17] found that cultural assumptions embedded in SPI methodologies could conflict with the cultural assumptions of developers, leading to difficulties in implementing process improvements. We used the competing values framework in our analysis of the relationship between organizational culture and SPI deployment. It focuses on values as its core constituents of organizational culture, successfully reflects the conflicting demands of the organizational context, and has been applied to the study of organizational issues ranging from culture to leadership, being accepted as determining both the type and strength of cultures prevalent in an organization. 2.3. The competing values framework in the context of SPI deployment Little research has specifically examined the framework as it relates to SPI deployment, except for Ngwenyama and Nielsen, who applied it to the content analysis of the cultural assumptions of the CMM and concluded that the design ideal of CMM reflected the market culture, but it becomes more hierarchic at the higher levels of maturity.

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

The significance of organizational culture (OC) as a source of organizational inertia is well known and there has been some interest in its influence on both the successful implementation and the use of IT. Organizations with flexible cultures and having a long-term orientation tend to adopt advanced manufacturing technology. Intranet adoption is likely to succeed in adhocracy culture. In addition, Cooper applied the CVF to understand IT implementation and proposed that different IS may support alternative values, and that if an IS conflicts the values of OC, implementation of the system will be resisted. Therefore, we expected that the competing values framework would play a critical role in the deployment of SPI. 3. Research model and hypotheses

273

An adhocracy culture concentrates on external positioning with a high degree of flexibility supported by an open system that promotes willingness to act. The organization values creativity, experimentation, risk, autonomy and responsiveness and if they have flexible cultures tend to adopt advanced manufacturing technology. Modern software organizations operate in a highly dynamic market, under tight time and cost constraints; therefore they are active in adopting SPI to improve their software quality. Additionally, adhocracy culture should increase employees’ positive attitude toward the organization and equity of its rewards. This led to the hypotheses: H2. There is a positive relationship between the adhocracy culture and SPI deployment. H2a. There is a positive relationship between the adhocracy culture and perceived SPI support.

3.1. Research model We developed our research model to predict the deployment of SPI in terms of the four types of organizational culture (see Fig. 2). In addition, the software process maturity level was expected to moderate the strength of the relationships. 3.2. Hypotheses

H2b. There is a positive relationship between the adhocracy culture and perceived SPI impact. H2c. There is a positive relationship between the adhocracy culture and the degree of SPI use.

H1. There is a positive relationship between the clan culture and SPI deployment.

A market culture concentrates on achieving goals through high productivity and economical operation; it tends to be results orientated and its members value competitiveness, diligence, perfectionism, aggressiveness and personal initiative. Key management activities are designed to maximize profit. It reacts to its environment in a manner that optimizes organizational productivity. According to Jiang et al. [15], organizations that have adopted SPI may not experience much benefit until they reach a higher maturity level. Moreover, strong emphasis on productivity and efficiency will lead to a focus on short-run impact. Iivari and Huisman [13] found that IS developers do not emphasize productivity, efficiency and goal achievement in the deployment of system development methodologies. This led us to the hypotheses:

H1a. There is a positive relationship between the clan culture and perceived SPI support.

H3. There is no relationship between the market culture and SPI deployment.

H1b. There is a positive relationship between the clan culture and perceived SPI impact.

H3a. There is no relationship between the market culture and perceived SPI support.

H1c. There is a positive relationship between the clan culture and [(Fig._2)TD$IG] degree of SPI use. the

H3b. There is no relationship between the market culture and perceived SPI impact.

The clan culture emphasizes on human relations and adopts flexible operation procedures focusing on internal relationships. It is believed to facilitate trust through affiliation and member participation. Managers need to promote employee dialogue, participation, and training to improve cohesive relationships, individual commitment and contribution. Top management commitment has been repeatedly shown to be most important in promoting a project’s success and when employees take ownership of SPI, they are proud of their accomplishments, and promote its use. Employees then are likely to trust their organizations. This led us to the hypotheses:

Fig. 2. Research model.

274

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

H3c. There is no relationship between the market culture and the degree of SPI use. The hierarchic culture concentrates on internal maintenance and strives for stability and control through a clear task setting and enforcement of strict rules. It tends to adopt a formal approach to relationships and leaders need to be good coordinators and organizers. It places a high value on economy, formality, rationality, order and obedience. Emphasis is on the task rather than the individual, who performs it. Ngwenyama and Nielsen found that SPI models reflected the hierarchic culture, especially at higher maturity levels. Iivari and Huisman found that there was a positive relationship between the hierarchic cultural orientation and system development methodology deployment for IS developers. SPI is consistent with a hierarchic culture’s values emphasizing control of activities by specifying methods and performance criteria in their internal focus. Thus following SPI regulations may be a means of supporting control, stability, and efficiency, leading to the hypotheses: H4. There is a positive relationship between the hierarchic culture and SPI deployment. H4a. There is a positive relationship between the hierarchic culture and perceived SPI support. H4b. There is a positive relationship between the hierarchic culture and perceived SPI impact. H4c. There is a positive relationship between the hierarchic culture and the degree of SPI use. Software process maturity shows the extent to which software development process parameters are used to enhance process effectiveness and the extent to which the process is controlled. Its maturity levels influence both software quality and project performance variables, such as cycle time and development effort [9]. However, organizational culture is the context in which SPI takes place. Therefore, SPI activities may result in organizational changes. The design ideal of the CMM reflects the market culture but it becomes more hierarchic at higher levels of maturity, leading us to hypothesize: H5. The level of software process maturity moderates the relationship between the clan culture and SPI deployment (perceived SPI support – H5a; perceived SPI impact – H5b; the degree of SPI use – H5c). H6. The level of software process maturity moderates the relationship between the adhocracy culture and SPI deployment (perceived SPI support – H6a; perceived SPI impact – H6b; the degree of SPI use – H6c). H7. The level of software process maturity moderates the relationship between the market culture and SPI deployment (perceived SPI support – H7a; perceived SPI impact – H7b; the degree of SPI use – H7c). H8. The level of software process maturity moderates the relationship between the hierarchic culture and SPI deployment (perceived SPI support – H8a; perceived SPI impact – H8b; the degree of SPI use – H8c).

4. Research methodology 4.1. Data collection Despite all the attention that SPI approaches have received, there is no solid evidence that they are used across organizations

and how software products have been improved by its deployment. Much criticism is based on case studies and therefore it is not necessarily generalizable. We employed a survey to obtain insight into SPI approaches and investigate the relationship between organizational culture and SPI deployment. The measurement items of variables of our model were drawn from the literature but adapted to the context of SPI. The survey instrument was pilot-tested with four SPI executives, two CMMI consultants, and four researchers working in SPI. Suggestions made by the respondents were incorporated and a new version of the instrument was developed. The population of interest was those organizations that had adopted the CMMI approach in Taiwan. Taiwan is one an Asian country that has organizations aggressively adopting CMMI. According to the results reported by SEI in March 2007 [22]: 46 organizations in Taiwan had obtained CMMI certificates. Taiwan had become number ninth in the world and fifth in Asia in holding certificates. At the time of the survey, 85 Taiwanese organizations had adopted CMMI to improve their software development processes [14]. For purposes of reliability, it was assumed that the Software Engineering Program Group (SEPG) leader should be able to provide reliable and accurate answers to our survey questions; this approach should help to overcome single source limitations. With the help of the Information Service Industry Association of R.O.C. (CISA), questionnaires were addressed directly to the SEPG leader in each organization. In the cover letter, confidentiality was assured and a summary of findings was offered as an incentive for participation. Of the 85 initial questionnaires mailed in the summer of 2007, a total of 62 usable responses were received, representing a response rate of 73%. The demographics of the 62 organizations and respondents are shown in Tables 1 and 2. 4.2. Measures All the questionnaire items had been used in previous empirical research. Principal component analysis was used to determine if all items measuring a construct cluster should be loaded onto a single factor. As a conservative rule, the sample size should be at least four or five times the number of items in the factor analysis. For measuring 11 different variables in our study, 58 questionnaire items were used. However, the sample size was 62, and this is less than the number of organizations required. Therefore, six separate factor analyses for perceived SPI support, perceived SPI impact, and

Table 1 Profiles of responding organizations (N = 62). Organizations characteristics

Number

%

Primary business System Integration Custom project development Software product development Information service Other

23 12 18 4 5

37.1 19.4 29 6.4 8.1

Number of employees Below 50 50–100 100–500 500–1000 Above 1000

19 12 19 6 6

30.6 19.4 30.6 9.7 9.7

CMMI maturity level No appraisal ML2 ML3 ML4 ML5

15 31 15 0 1

24.2 50 24.2 0 1.6

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281 Table 2 Profiles of responding SEPG leaders (N = 62). Respondents characteristics

Table 4 Factor analysis – perceived SPI support provided as control technology. Number

%

Measurement items C1 C2

Education Junior College University degree Master’s degree Title CEO Executive vice president General manager Division chief Project manager System analyst/engineer

7 21 34

11.3 33.9 54.8

5 12 32 2 6 5

8.1 19.4 51.6 3.2 9.6 8.1

Work experience 1–5 years 6–10 years 11–15 years 16–20 years 21–25 years 26–30 years

18 12 9 14 7 2

29 19.4 14.5 22.6 11.3 3.2

Gender Male Female

46 16

74.2 25.8

organizational culture were carried out to obtain a better solution by increasing the ratio of the sample size to the number of items. In addition, the small ratio of subjects to measures may result in instability in the factor loadings due to sampling error. In our study, in order to avoid the bias introduced by sampling, a ‘‘complete’’ survey was conducted because 100% of the population was surveyed to avoid sampling error. 4.2.1. Dependent variables: SPI deployment 4.2.1.1. Perceived SPI support for software development. Adapted from Huisman and Iivari, perceived SPI support as production technology was measured using seven 5-point Likert scale items, ranging from 1 (strongly disagree) to 5 (strongly agree). Principal component analysis followed by varimax rotation resulted in a one-factor solution with loadings greater than 0.7. These items explained 74.1% of the total variance. The results are shown in Table 3. The reliability of measures was tested using Cronbach’s alpha, for which a minimum value of 0.70 is generally recommended. Table 8 shows the Cronbach’s alpha value of 0.94, revealing a high level of internal consistency among measurement items. Seven 5-point Likert scale items were used to assess perceived SPI support as control technology. These items were adapted from Huisman and Iivari. Similarly, a factor analysis with varimax rotation was conducted for the items, resulting in a onefactor solution; each loading was greater than 0.7. These seven items explained 62.9% of the total variance as seen in Table 4. The

Table 3 Factor analysis – perceived SPI support provided as production technology. Measurement items B1 B2 B3 B4 B5 B6 B7

Align the software to be developed with the business Build management commitment in our software development projects Helps in software design Helps in implementing developed software Helps in reviewing developed software Helps in testing developed software Helps to get the software accepted

Eigenvalue Percentage of total variance

275

Factor loadings 0.87 0.92 0.90 0.94 0.83 0.82 0.74 5.19 74.1%

C3 C4 C5 C6 C7

Helps to estimate the size of the software to be developed Helps to estimate the time and effort required for the development of planned software Helps to plan software development projects Helps in defining useful milestones for our software development projects Helps to organize software development projects Helps to keep our software development projects under control Helps to estimate the project risks

Eigenvalue Percentage of total variance

Factor loadings 0.80 0.71 0.84 0.83 0.85 0.71 0.79 4.40 62.9%

Cronbach’s alpha value of 0.90 revealed a high level of internal consistency among measurement items. Also from Huisman and Iivari’s research, perceived SPI support as cooperative technology was measured using seven 5-point Likert scale items. The factor analysis resulted in a one-factor solution with loadings greater than 0.7. These seven items explained 72.9% of the total variance as shown in Table 5. The Cronbach’s alpha value of 0.94 again revealed a high level of internal consistency among measurement items. 4.2.1.2. Perceived SPI impact on product and process quality. Five 5point Likert scale items were used to assess perceived SPI impact on product quality, also adapted from Huisman and Iivari and Rai and Al-Hindi’s work. A factor analysis with varimax rotation was conducted for the product quality items; it showed that the five items loaded on one factor, each with loadings greater than 0.70. These explained 79.3% of the total variance, as can be seen in Table 6. Table 8 shows that the Cronbach’s alpha value of 0.93 shows a high level of internal consistency among measurement items. Adapted from Rai and Al-Hindi, perceived SPI impact on process quality was measured using five 5-point Likert scale items. Factor analysis resulted in a one-factor solution, each with loadings greater than 0.70. These five items explained 71.1% of the total variance, as seen in Table 7; the Cronbach’s alpha value of 0.89 again revealed a high level of internal consistency among measurement items. 4.2.1.3. SPI use. A two-item instrument was used to assess SPI horizontal use. The items reflect the percentage of software developers and projects using SPI knowledge. The Cronbach’s Table 5 Factor analysis – perceived SPI support provided as cooperative technology. Measurement items D1 D2 D3 D4 D5 D6 D7

Describes a sound way of developing software Forms a useful standard for our software development Reminds me about activities/tasks of software development Provides a useful list of possible software development activities Provides useful guidelines for conducting software development Allows us to learn from our software development experience Defines an ideal process of software development that is useful, even though it is not followed in practice

Eigenvalue Percentage of total variance

Factor loadings 0.91 0.89 0.87 0.83 0.80 0.83 0.84

5.11 72.9%

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

276

Table 6 Factor analysis – perceived SPI impact on product quality. Measurement items E1 E2 E3 E4 E5

Factor loadings

Helps to develop more functional software Helps to develop more reliable software Helps to develop more maintainable software Helps to develop better software Helps to make users more satisfied with our software

Eigenvalue Percentage of total variance

0.87 0.91 0.89 0.90 0.89

3.97 79.3%

v1.1, and appraisal method, the Standard CMMI Appraisal Method for Process Improvement (SCAMPI) Class A v1.1. The latter appraisal was conducted by a SEI-certified lead appraiser. As shown in Table 1, 47 organizations had performed a CMMI appraisal, and 15 organizations had adopted a CMMI approach but had not reached the appraisal stage. Due to the extremely small number of maturity levels 4 and 5, two dummy variables were used to represent maturity levels. The MD1 dummy took the value ‘‘1’’ for maturity level 2, and zero otherwise. The MD2 dummy took the value ‘‘1’’ for maturity level 3, and zero otherwise. For those with no appraisal, all dummy variables were set to zero. 5. Data analysis and results

Table 7 Factor analysis – perceived SPI impact on process quality. Measurement items F1 F2 F3 F4 F5

5.1. Analysis of main effects Factor loadings

Helps to improve the degree of agreement among participants in the development process Helps to decrease the duplication of efforts during the development process Helps to complete software within budget Helps to complete software within schedule Improves the quality of the development efforts

Eigenvalue Percentage of total variance

0.75 0.76 0.85 0.93 0.91 3.56 71.1%

alpha value of 0.87 reveals a high level of internal consistency among measurement items for perceived SPI horizontal use, as seen in Table 8. One 5-point Likert scale item was used to measure SPI vertical use, ranging from 1 (nominally) to 5 (intensively). This item reflects the maximum intensity of SPI usage. 4.2.2. Independent variables: organizational culture A 24-item scale was used to assess organizational culture. The items were adopted from the Organizational Culture Assessment Instrument (OCAI), which has been widely used in almost 10,000 organizations worldwide. The scale consisted of four variable groups (each of which consisted of six items) corresponding to the four ideal culture types specified in the competing value framework. Although the organizational culture construct had been previously examined, a factor analysis with varimax rotation was conducted by us. The results showed that four factors were extracted, each of which corresponded to one of the four competing values framework culture orientations. These four factors explained 62.5% of the total variance, as seen in Table 9. Table 10 shows that the Cronbach’s alpha calculated for each variable ranged from 0.71 to 0.80 revealing an acceptable level of reliability. 4.2.3. Moderating variables: software process maturity The CMMI level of organizations in our survey was determined by a CMMI-based appraisal using the benchmark model, CMMI

An examination of mean values of each cultural variable in Table 10 reveals that the hierarchic culture has the highest mean value, followed by clan, market and adhocracy cultures, suggesting that their SPI initiatives were internally oriented. A regression analysis was then conducted, taking SPI deployment as the dependent variable and organizational culture as the independent variable. The results of this analysis are shown in Table 11. The P-value of 0.00 indicated that there was a significant relationship between SPI deployment and organizational culture, suggesting that it had a positive relationship (0.64) with SPI deployment. To further examine the relationships between the individual culture and SPI deployment, another regression analysis was conducted by breaking the overall organizational culture construct into the four cultural orientations; Table 11 shows that the clan and hierarchic cultures were marginally significantly related to SPI deployment. On the other hand, the adhocracy and market cultures were not. To understand the underlying dynamics of these marginally significant or non-significant trends, separate multiple regression analyses was conducted to determine whether individual culture was related to any of the three SPI deployment dimensions. Table 12 shows the relationships between the organizational culture and perceived SPI support. These indicated that there was a significant relationship between hierarchic culture and overall SPI support. The relationships between the different SPI support dimensions and the cultural dimensions were then further examined, showing that the SPI support dimensions of production, control, and cooperation were all positively related to hierarchic culture either significantly (P < 0.01) or marginally (P < 0.10). Control was related to market culture in a marginally negative direction. Therefore, Hypotheses 3a and 4a were supported but 1a and 2a were not. Table 13 shows the relationships between cultural dimensions and perceived overall SPI impact on product quality and development process quality. These indicated that there was a positive significant relationship between hierarchic culture and

Table 8 Reliability of SPI deployment. Composite variable

Variable name

Description

Items

Cronbach’s alpha

Mean

Std. dev.

Perceived SPI support

PRODUCTION CONTROL COOPERATIVE

As production technology As control technology As cooperative technology

B1, B2, B3, B4, B5, B6, B7 C1, C2, C3, C4, C5, C6, C7 D1, D2, D3, D4, D5, D6, D7

0.94 0.90 0.94

3.80 3.40 3.61

0.59 0.61 0.66

Perceived SPI Impact

PRODUCT PROCESS

Product quality Process quality

E1, E2, E3, E4, E5 F1, F2, F3, F4, F5

0.93 0.89

3.73 3.60

0.66 0.66

SPI use

HOR_USE VER_USE

Horizontal use Vertical use

G1, G2 G3

0.87 –

3.98 3.66

0.90 0.79

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

277

Table 9 Factor analysis – organizational culture. Items

Key word

Market culture

A1 A5 A9 A13 A17 A21 A2 A6 A10 A14 A18 A22 A3 A7 A11 A15 A19 A23 A4 A8 A12 A16 A20 A24

Personal place Mentoring, facilitating, nurturing Teamwork and participation Loyalty and mutual trust Human development Commitment and concern for people Dynamic and entrepreneurial place Entrepreneurship, innovating Individual risk-taking, innovation Commitment to innovation Acquiring new resources Having the most unique products Results oriented No-nonsense, aggressive Competitiveness and high demands Emphasis on goal accomplishment Competitive actions and achievements Winning in the marketplace Controlled and structured Coordinating, organizing Security of employment and conformity Formal rules and policies Permanence and stability Efficiency

Percentage of total variance (%) Cumulative percentage of total variance (%)

20.64 20.64

Clan culture

Hierarchic culture

Adhocracy culture

0.77 0.74 0.63 0.52 0.77 0.50 0.69 0.81 0.80 0.62 0.52 0.67 0.48 0.60 0.58 0.75 0.70 0.78 0.55 0.79 0.65 0.53 0.48 0.48 17.60 38.24

14.00 52.24

10.2 62.5

Note: For clarity, only factor loadings > 0.45 are shown.

Table 10 Reliability of organizational culture.

Table 12 Regression coefficients of organizational culture on perceived SPI support.

Variable name

Description

Cronbach’s alpha

Mean

Std. dev.

CLAN ADHOC MARKET HIER

Clan culture Adhocracy culture Market culture Hierarchic culture

0.80 0.79 0.71 0.77

3.96 3.72 3.84 4.01

0.46 0.50 0.41 0.44

overall SPI impact. The relationships between the different SPI impact and cultural dimensions were separately examined, showing that the SPI impact dimensions of product quality and process quality were all significantly and positively related to hierarchic culture. Process quality was related to market culture in a marginally negative direction. Therefore, Hypotheses 3b and 4b were supported, while 1b and 2b were not. Table 14 shows the relationship between the strength of each culture dimension and overall SPI use suggesting that the clan culture had a marginally positive relationship with overall SPI use. The relationships between the different SPI use dimensions and the cultural dimensions were separately examined, showing that horizontal use had a marginally significant relation with clan culture. However, the adhocracy culture did not exhibit any significant association with SPI support, SPI impact, and SPI use.

Independent variable

Dependent variable PRODUCTION

CONTROL

COOPERATIVE

Overall SPI Support

CLAN ADHOC MARKET HIER Model R2 Adjusted R2 F-value

0.04 0.02 0.26 0.62** 0.26 0.20 4.92**

0.02 0.24 0.29# 0.57** 0.32 0.28 6.80**

0.33 0.09 0.09 0.35# 0.42 0.37 10.10**

0.13 0.12 0.23 0.56** 0.38 0.34 8.7**

** #

P < 0.01. P < 0.10.

Therefore, Hypotheses 1c and 3c were supported while 2c and 4c were not. The organizational size (number of employees) was included as a control variable in the revised regression model. However, there were no significant associations with any of the hypotheses.

Table 13 Regression coefficients of organizational culture on perceived SPI impact. Independent variable

Table 11 Regression results. Dependent variable

Independent variable

Coefficient

P-value

(1) SPI deployment

Organizational culture

0.64

0.00**

(2) SPI deployment

CLAN ADHOC MARKET HIER

0.32 0.28 0.21 0.30

0.09# 0.13 0.15 0.08#

** #

P < 0.01. P < 0.10.

CLAN ADHOC MARKET HIER Model R2 Adjusted R2 F-value * ** #

P < 0.05. P < 0.01. P < 0.10.

Dependent variable PRODUCT

PROCESS

0.12 0.28 0.17 0.40* 0.39 0.35 9.16**

0.21 0.07 0.27# 0.56** 0.37 0.33 8.5**

Overall SPI impact 0.18 0.18 0.23 0.51** 0.42 0.38 10.4**

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

278

Table 14 Regression coefficients of organizational culture on SPI use.

5.2. Analysis of moderating effects

Independent variable

Dependent variable HOR_USE

VER_USE

Overall SPI use

CLAN ADHOC MARKET HIER Model R2 Adjusted R2 F-value

0.46# 0.02 0.14 0.04 0.17 0.12 3.0*

0.26 0.35 0.04 0.01 0.32 0.27 6.65**

0.40# 0.19 0.10 0.03 0.27 0.22 5.4

* ** #

P < 0.05. P < 0.01. P < 0.10.

Table 15 Regression coefficients of organizational culture and maturity level on perceived SPI support. Independent variable

CLAN ADHOC MARKET HIER MD1 MD2 CLAN  MD1 CLAN  MD2 ADHOC  MD1 ADHOC  MD2 MARKET  MD1 MARKET  MD2 HIER  MD1 HIER  MD2 Model R2 Adjusted R2 F-value * ** #

PRODUCTION

CONTROL

COOPERATIVE

0.44 0.13 0.75* 0.68* 0.02 0.14 3.47 1.31 0.71 1.78 3.45 4.35 0.68 3.79 0.33 0.13 1.6

0.27 0.54 0.33 0.91** 2.48 2.16 0.01 2.82 0.47 2.97 0.21 0.37 2.91 1.67 0.39 0.21 2.1*

0.57 0.34 0.47 0.61* 2.80# 1.06 3.10 0.20 0.74 2.31 1.66 2.48 2.19 1.14 0.50 0.35 3.3**

Table 16 Regression coefficients of organizational culture and maturity level on perceived SPI impact and use.

CLAN ADHOC MARKET HIER MD1 MD2 CLAN  MD1 CLAN  MD2 ADHOC  MD1 ADHOC  MD2 MARKET  MD1 MARKET  MD2 HIER  MD1 HIER  MD2 Model R2 Adjusted R2 F-value * ** #

P < 0.05. P < 0.01. P < 0.10.

6. Discussion and implications

Dependent variable

P < 0.05. P < 0.01. P < 0.10.

Independent variable

Regression analysis was conducted to test the roles of software process maturity in moderating the association between organizational culture variables and SPI deployment. Table 15 shows that this did not find any significant interactions between maturity dummies and the four types of organizational culture on perceived SPI support, indicating that maturity level did not moderate the influence of organizational culture on perceived SPI support. Therefore, Hypotheses 5a, 6a, 7a, and 8a were not supported. Table 16 shows that only the interaction of clan culture with the dummy for maturity level 3 (MD2) had a significantly negative effect on vertical use. Therefore, Hypothesis 5c was weakly supported while 5b was not. However, the effect of the interaction of adhocracy, market, and hierarchic culture with maturity level did not exhibit any significant association with SPI impact and SPI use. Therefore, Hypotheses 6b, 6c, 7b, 7c, 8b and 8c were not supported.

Dependent variable PRODUCT

PROCESS

HOR_USE

VER_USE

0.02 0.50 0.38 0.82** 1.59 1.87 1.52 3.53 0.08 2.31 3.26 0.39 3.46 3.42 0.52 0.38 3.67**

0.44 0.07 0.51 0.71* 1.46 1.57 2.12 0.06 1.13 0.07 0.83 1.29 1.24 2.88 0.40 0.21 2.2*

0.80# 0.08 0.30 0.08 2.42 1.10 3.42 5.16 0.75 1.66 0.45 0.03 0.01 2.59 0.35 0.15 1.7#

0.90* 0.42 0.52 0.32 0.31 0.10 3.18 6.96* 0.65 0.57 0.89 2.49 2.25 5.22 0.49 0.33 3.1**

One important finding was that the organizational culture does indeed have an influence on SPI deployment, particularly for the relationships between hierarchic culture and SPI deployment. However, the deployment of SPI is most likely to occur in organizations where the organizational culture is hierarchic, which apparently acts as a facilitator for further SPI implementation. A summary of the results is given in Table 17. Our results also indicated that SPI support and its impact were most related to hierarchic culture. More specifically, an organization so characterized could increase understanding of the capabilities of the SPI approach, its probable value to the organization, and the actual consequences of adopting an SPI approach. Over 70% of the responding companies in our study were, however, at the lower maturity levels (below 3). The environment facing software organizations today is marked by extreme competition and uncertainty. According to Panayotopoulou et al. [19], when employees feel uncertain and insecure about their future, they show greater tolerance towards factors that could negatively influence their work, such as bureaucratic procedures and tight control. Thus an hierarchic culture is a necessary condition for successful deployment of SPI. The study also suggests that SPI use could be improved by clan culture. It could increase the proportion of software developers and projects using the SPI knowledge, as it emphasizes teamwork and employee commitment through the development of a strong value system that promotes corporate identity. It can thus be surmised that clan culture is a necessary condition in creating skills development and sharing SPI knowledge in the process of SPI deployment. The market culture, as expected, did not exhibit any significant association with SPI deployment. It requires a great amount of time and money before benefits can be realized. However, to our surprise, the adhocracy culture was not related to SPI deployment. Possibly its focus on growth and innovation was not met through SPI deployment. In the competing values framework, both clan and hierarchic cultures reflected internal values. Our overall results showed that the deployment of SPI was associated with an internal orientation, reflecting an emphasis on the maintenance and improvement of the existing organization. The moderating effect of maturity levels on the relationship between organizational culture and vertical use was significant. However, the significantly negative interaction only existed between clan culture and maturity level 3. This result implied that the negative effect of clan culture on the maximum intensity

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

279

Table 17 Summary of results. Hypothesis number

Results from study

Is hypothesis supported?

H1 H1a H1b H1c H2 H2a H2b H2c H3 H3a H3b H3c H4 H4a H4b H4c H5 H5a H5b H5c H6 H6a H6b H6c H7 H7a H7b H7c H8 H8a H8b H8c

Clan culture was marginal significantly related to SPI deployment Clan culture was not related to SPI support Clan culture was not related to SPI impact Clan culture was marginal significantly related to SPI use Adhocracy culture was not related to SPI deployment Adhocracy culture was not related to SPI support Adhocracy culture was not related to SPI impact Adhocracy culture was not related to SPI use Market culture was not related to SPI deployment Market culture was not related to SPI support Market culture was not related to SPI impact Market culture was not related to SPI use Hierarchic culture was marginal significantly related to SPI deployment Hierarchic culture was positively related to SPI support Hierarchic culture was positively related to SPI impact Hierarchic culture was not related to SPI use The interaction of clan culture with maturity level was positively related to SPI deployment The interaction of clan culture with maturity level was not related to SPI support The interaction of clan culture with maturity level was not related to SPI impact The interaction of clan culture with maturity level was related to SPI use The interaction of adhocracy culture with maturity level was not related to SPI deployment The interaction of adhocracy culture with maturity level was not related to SPI support The interaction of adhocracy culture with maturity level was not related to SPI impact The interaction of adhocracy culture with maturity level was not related to SPI use The interaction of market culture with maturity level was not related to SPI deployment The interaction of market culture with maturity level was not related to SPI support The interaction of market culture with maturity level was not related to SPI impact The interaction of market culture with maturity level was not related to SPI use The interaction of hierarchic culture with maturity level was not related to SPI deployment The interaction of hierarchic culture with maturity level was not related to SPI support The interaction of hierarchic culture with maturity level was not related to SPI impact The interaction of hierarchic culture with maturity level was not related to SPI use

Yes No No Yes No No No No Yes Yes Yes Yes Yes Yes Yes No Yes No No Yes No No No No No No No No No No No No

of SPI usage was stronger in organizations with higher maturity levels. None of the interaction terms were significant, indicating that maturity levels do not moderate the influence of adhocracy, market, and hierarchic cultures on SPI deployment. 7. Conclusions Our study applied a competing values framework to analyze the relationship between the organizational culture and SPI deployment. Results suggested that the organizational culture did indeed have an influence on the deployment, which was primarily associated with hierarchic culture. On the other hand, clan culture was a necessary condition for creating skills development and sharing SPI knowledge in the process of SPI deployment. Therefore, SEPG leaders should recognize all the steps involved in software process improvement, learn how to be good coordinators and organizers, and encourage the

(weak support)

(full support)

(weak support)

(weak support)

(weak support)

development of internally oriented cultures. This study points out the need to consider culture when a new SPI approach is implemented; it may be incompatible with the existing culture. One limitation of our study was that it focused only on CMMI as the software process improvement approach. Another limitation of this study is that its findings were based on investigations in Taiwan. Obviously, generalizing the results to different cultural or economic contexts should be made with caution. Acknowledgements This research was supported by the National Science Council (NSC) of Taiwan under the contract 98-2410-H-011-004. The authors also wish to thank anonymous reviewers for their constructive comments and the chief editor Prof. Edgar H Sibley for his editorial effort on the manuscript of this paper.

Appendix A. Questionnaire items This appendix describes the questionnaire items that pertained to the constructs used in the study. A.1. Organizational culture For each item listed below, to what extent do you agree with the following statements? (1 = strongly disagree, 3 = neither disagree nor agree, 5 = strongly agree): A1. A2. A3. A4. A5. A6. A7. A8. A9. A10.

The The The The The The The The The The

organization is a very personal place. It is like an extended family. People seem to share a lot of themselves. organization is a very dynamic and entrepreneurial place. People are willing to stick their necks out and take risks. organization is a very results orientated. A major concern is getting on with the job. People are very competitive and achievement orientated* organization is a very controlled and structured place. Formal procedures generally govern what people do. leadership of the organization is generally considered to exemplify mentoring, facilitating or nurturing. leadership of the organization is generally considered to exemplify entrepreneurship, innovation or risk taking. leadership of the organization is generally considered to exemplify a no-nonsense, aggressive, results-orientated focus. leadership of the organization is generally considered to exemplify coordinating, organizing, or smooth-running efficiency. management style in the organization is characterized by teamwork, consensus and participation management style in the organization is characterized by individual risk-taking, innovation, freedom and uniqueness.

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

280 A11. A12. A13. A14. A15. A16. A17. A18. A19. A20. A21. A22. A23. A24.

The The The The The The The The The The The The The The

management style in the organization is characterized by hard-driving competitiveness, high demands and achievement management style in the organization is characterized by security of employment, conformity, predictability and stability in relationships glue the holds the organization together is loyalty and mutual trust. Commitment to the organization runs high. glue the holds the organization together is commitment to innovation and development. There is an emphasis on being on the cutting edge. glue the holds the organization together is the emphasis on achievement and goal accomplishment. Aggressiveness and winning are common themes. glue the holds the organization together is formal rules and policies. Maintaining a smooth-running organization is important. organization emphasizes human development. High trust, openness and participation persist. organization emphasizes acquiring new resources and creating new challenges. Trying new things and prospecting for opportunities are valued. organization emphasizes competitive actions and achievement. Hitting stretch targets and winning in the marketplace are dominant. organization emphasizes permanence and stability. Efficiency, control and smooth operations are important. organization defines success on the basis of the development of human resources, teamwork, employee commitment and concern for people. organization defines success on the basis of having the most unique or the newest products. It is a product leader and innovator. organization defines success on the basis of winning in the marketplace and outpacing the competition. Competitive market leadership is the key. organization defines success on the basis of efficiency. Dependable delivery, smooth scheduling and low cost production are critical.

A.2. Deployment of SPI approaches To what extent do you agree with the following statements? (1 = strongly disagree, 3 = neither disagree nor agree, 5 = strongly agree): Perceived SPI support provided as production technology B1. Our software process improvement approach B2. Our software process improvement approach B3. Our software process improvement approach B4. Our software process improvement approach B5. Our software process improvement approach B6. Our software process improvement approach B7. Our software process improvement approach

helps helps helps helps helps helps helps

to to in in in in to

align the software to be developed with the business. build management commitment in our software development projects. software design. implementing developed software. reviewing developed software. testing developed software. get the software accepted

Perceived SPI support provided as control technology C1. Our software process improvement C2. Our software process improvement C3. Our software process improvement C4. Our software process improvement C5. Our software process improvement C6. Our software process improvement C7. Our software process improvement

helps helps helps helps helps helps helps

to to to in to to to

estimate the size of the software to be developed. estimate the time and effort required for the development of planned software. plan software development projects. defining useful milestones for our software development projects. organize software development projects. keep our software development projects under control. estimate the project risks.

approach approach approach approach approach approach approach

Perceived SPI support provided as cooperative technology D1. Our software process improvement approach D2. Our software process improvement approach D3. Our software process improvement approach D4. Our software process improvement approach D5. Our software process improvement approach D6. Our software process improvement approach D7. Our software process improvement approach

describes a sound way of developing software. forms a useful standard for our software development. reminds me about activities/tasks of software development. provides a useful list of possible software development activities. provides useful guidelines for conducting software development. allows us to learn from our software development experience. defines an ideal process of software development that is useful, even though it is not followed in practice.

Perceived SPI impact on product quality E1. Our software process improvement approach helps to develop more functional software. E2. Our software process improvement approach helps to develop more reliable software. E3. Our software process improvement approach helps to develop more maintainable software. E4. Overall, our software process improvement approach helps to develop better software. E5. Overall, our software process improvement approach helps to make users more satisfied with our software. Perceived SPI impact on process quality F1. Our software process improvement approach helps to improve the degree of agreement among participants in the development process. F2. Our software process improvement approach helps to decrease the duplication of efforts during the development process. F3. Our software process improvement approach helps to complete software within budget. F4. Our software process improvement approach helps to complete software within schedule. F5. Overall, our software process improvement approach improves the quality of the development efforts. SPI use Horizontal use G1. What is the proportion of projects in your organization that are developed by applying software process improvement approach knowledge?

None 1–25% 26–50% 51–75% Over 75%

G2.

1 2 3 4 5

What is the proportion of people in your organization who apply software process improvement approach knowledge?

None 1–25% 26–50% 51–75% Over 75%

1 2 3 4 5

Vertical use G3.To what extent is your organization using CMMI at present? (1 = nominally, 5 = intensively)

C.-C. Shih, S.-J. Huang / Information & Management 47 (2010) 271–281

References [1] I. Aaen, J. Arendt, L. Mathiassen, O. Ngwenyama, A conceptual MAP of software process improvement, Scandinavian Journal of Information Systems 13, 2001, pp. 79–99. [2] S.S. Al-Gahtani, G.S. Hubona, J. Wang, Information technology (IT) in Saudi Arabia:Culture and the acceptance and use of IT, Information & Management 44 (8), 2007, pp. 681–691. [3] N. Ashrafi, The impact of software process improvement on quality in theory and practice, Information & Management 40 (7), 2003, pp. 677–690. [4] N. Baddoo, T. Hall, De-motivators for software process improvement: an analysis of practitioners’ views, The Journal of Systems and Software 66 (1), 2003, pp. 23– 33. [5] F. Cattaneo, A. Fuggetta, D. Sciuto, Pursuing coherence in software process assessment and improvement, Software Process Improvement and Practice 6 (1), 2001, pp. 3–22. [6] D.L. Gibson, D.R. Goldenson, Performance results of CMMI-based process improvement, CMU/SEI-2006-TR-004, 2006. [7] B. Hansen, J. Rose, G. Tjørnehøj, Prescription, description, reflection: the shape of the software process improvement field, International Journal of Information Management 24, 2004, pp. 457–472. [8] G.R. Harper, D.R. Utiey, Organizational culture and successful information technology implementation, Engineering Management Journal 13 (2), 2001, pp. 11– 15. [9] D.E. Harter, M.S. Krishnan, S.A. Slaughter, Effects of process maturity on quality, cycle time, and effort in software projects, Management Science 46 (4), 2000, pp. 451–466. [10] N. Hoffman, R. Klepper, Assimilating new technologies: the role of organizational culture, Information Systems Management 17 (3), 2000, pp. 36–42. [11] S.-J. Huang, W.-M. Han, Selection priority of process areas based on CMMI continuous representation, Information & Management 43 (3), 2006, pp. 297– 307. [12] M. Huisman, J. Iivari, Deployment of systems development methodologies: perceptual congruence between IS managers and systems developers, Information & Management 43 (1), 2006, pp. 29–49. [13] J. Iivari, M. Huisman, The relationship between organizational culture and the deployment of systems development methodologies, MIS Quarterly 31 (1), 2007, pp. 35–58. [14] Information Service Industry Association of R.O.C. (CISA), http://www.cmmi-taiwan. org.tw/content/index.aspx, 2007. [15] J.J. Jiang, K. Gary, H.G. Hwang, J. Huang, S.Y. Hung, An exploration of the relationship between software development process maturity and project performance, Information & Management 41 (2), 2004, pp. 279–288. [16] J.P. Kuilboer, N. Ashrafi, Software process and product improvement: an empirical assessment, Information and Software Technology 42 (1), 2000, pp. 27–34. [17] O. Ngwenyama, P.A. Nielsen, Competing values in software process improvement: an assumption analysis of CMM from an organizational culture perspective, IEEE Transactions on Engineering Management 50 (1), 2003, pp. 101–111. [18] P.A. Nielsen, J. Nørbjerg, Software process maturity and organizational politics, in: Proceedings of the IFIP TC8/WG8.2 Working Conference on Realigning Research and Practice in Information Systems Development: The Social and Organizational Perspective, 2001, pp. 221–240. [19] L. Panayotopoulou, D. Bourantas, N. Papalexandris, Strategic human resource management and its effects on firm performance: an implementation of the

[20]

[21]

[22] [23] [24]

281

competing values framework, International Journal of Human Resource Management 14 (4), 2003, pp. 680–699. A. Rai, H. Al-Hindi, The effects of development process modeling and task uncertainty on development quality performance, Information & Management 37 (6), 2000, pp. 335–346. T. Ravichandran, A. Rai, Total Quality management in information systems development: key constructs and relationships, Journal of Management Information Systems 24 (3), Winter 1999–2000, pp. 81–415. SEI, Process maturity profile of the software community, Software Engineering Institute, Carnegie Mellon University, 2007. M. Staples, M. Niazi, Systematic review of organizational motivations for adopting CMM-based SPI, Information and Software Technology 50 (7), 2008, pp. 605–620. C. Yoon, The effects of national culture values on consumer acceptance of ecommerce: online shoppers in China, Information & Management 46 (5), 2009, pp. 294–301.

Chiao-Ching Shih is a lecturer in the Department of Information Management at St. John’s University, Taiwan. She received her Master degree in Information Management from National Chung Cheng University in 1997, and is currently a PhD candidate in the Department of Information Management, National Taiwan University of Science and Technology (NTUST). She is also a member of the Software Engineering and Management Laboratory at NTUST, Taiwan. Her research interests include software process improvement, system development, and project management.

Sun-Jen Huang received his B.A. in Industrial Management in 1988, and his M.S. in Engineering and Technology in 1991, both from the National Taiwan University of Science and Technology, and his PhD degree from the School of Computer Science and Computer Engineering, La Trobe University, Melbourne, Australia, in 1999. He is currently a professor in the Department of Information Management, National Taiwan University of Science and Technology, Taipei, Taiwan. He also servers as the head of the Software Engineering and Management Laboratory at NTUST, which hosts research projects every year from National Science Council and software industry in Taiwan. Dr. Huang is also a chairman of Software Quality Promotion Committee at the Chinese Society for Quality. His research interests include software engineering and project management, software process improvement, software measurement and analysis, and software quality management. Dr. Huang has published more than 20 papers in journals including Information & Management, IEEE Transactions on Software Engineering, Software Practice & Experience, Journal of Systems and Software, Informal and Software Technology, The Service Industries Journal, European Journal of Operational Research, Applied Intelligence, Journal of Information Science and Engineering, and Journal of Software Engineering Studies.