7thInternational Software Quality Management Conference SQM 99, Southampton, UK, 29-31 April 1999
Software Quality Management from a Cross-Cultural Viewpoint Kerstin V. Siakas,2,1 Elli Georgiadou,1 Chris Sadler1 1
University of North London School of Informatics and Multimedia Technology 2-16 Eden Grove, London, N7 8DB, UK Tel: + 44 171 753 3142, Fax: + 44 171 753 7009 E-mail:
[email protected],
[email protected],
[email protected] 2,
Technological Educational Institution of Thessaloniki, Department of Informatics, P.O. Box 14561 GR-54101 Thessaloniki, Greece Tel: +30 31 791 296, Fax: +30 31 799 152 E-mail:
[email protected]
Abstract This paper describes our attempts to look at how different cultures influence Software Quality Management [1,2]. Quality of the software product cannot be determined until late in the development cycle. It is widely accepted that reliance has to be placed on process quality as a means of achieving product quality. In recent years software organisations have been making efforts concentrating on improving the process in order to improve software quality. Hofstede [3] had defined four dimensions of work related value differences for classifying cultural influences in different countries. These are: power distance, collectivism/individualism, femininity/masculinity and uncertainty avoidance. In this paper software development organisations are being assessed for their awareness and acceptance of Software Quality. These results will be correlated with the work-related values identified by Hofstede. The intention is to identify cultural factors, which may have a bearing on successful adoption and implementation of Software Quality Management.
1.
Introduction
The techniques described as ”software process assessment” have their origin in the Total Quality Management (TQM) movement, which derives from the basic premise, introduced by Deming, that the quality of manufactured products is mainly determined by the quality of the processes which produce them. TQM is a business strategy that emphasises continuous improvement of products and services within organisations. The focus in TQM is on customer satisfaction and continuous commitment to quality
by the entire workforce. TQM aims to create a cultural change by changing the way people think and work. Culture is mentioned as a corporate culture, which has to be changed in the light of TQM [4,5,6,1]. We pose the question: “How can such a new corporate culture be implemented without taking into consideration the underlying cultural values of the society in which the organisation exists, and cultural differences in organisational behaviour and attitudes”?
2.
Software Quality Assessment
Different Maturity Models, including the Capability Maturity Model (CMM) [7,8,9,10,11,12], Bootstrap [13,14,9] and SPICE [15,16,17] have been developed to provide the framework for process and capability determination by measuring the maturity of an organisation. These models are technically considered as hard i.e. not taking into consideration the soft view [18,2]. The People Capability Maturity Model (P-CMM) [19,5] is an attempt to consider people-issues. P-CMM focuses on three interrelated components namely people, process and technology. The motivation for P-CMM is to improve the ability of software organisations to attract, develop, motivate, organise and retain the talent needed to continuously improve software development capability. The focus in all these models is on the assessment of the overall technical capability of an organisation. In 1998 SPICE became the ISO software process assessment standard called ISO15504. The ISO15504-standard combines different methodologies and it is linked with the ISO12207 standard, which provides a framework for software processes [17].
3.
The Socio-Technical Perspective
Boström et.al. [20] assert that many of the problems and failures of Management Information Systems (MIS) are the result of organisational behavioural problems and especially of inadequate design. They believe that there is a need to reframe MIS design methodology within the Socio-Technical System (STS) [20] design approach in order to change systems designers’ perspectives. By designers they include all people who influence MIS design decisions, i.e. systems analysts, users, senior management and computer systems designers. The STS approach is introduced as a realistic view of organisations and as a way to change them. An organisation or organisational work system is described as two jointly independent, but correlative interacting systems – the social and the technical, the socio-technical system. Any design or redesign of a work system must deal with both of these systems in an integrated form. Two potential improvements are possible, the improvement of task accomplishment and the improvement of the quality of working life. Figure 1 shows the interaction between variables in a socio-technical system. Direct changes in MIS tasks or technology variables result in changes within variables like structure and people [20]. There is a strong association between work relationship
changes like attitudes, motivation and interpersonal behaviour within the work system. Secondary changes in structure and people variables also cause changes in variables like technology and task. Changes in any of these variables result in changes in the direct MIS and thus they should be designed to complement and reinforce each other. Social System
Technical System
Structure
Technology
People
Task
Fig. 1.
The Interacting Variable Classes within a Work System (adapted from Boström et.al.)
4.
Cultural Studies
MIS
The concept of culture is understood in different ways. Confucius once said that all people are the same, it is only their habits that are different. Hall [21] defines culture as: “the pattern of taken-for-granted assumptions about how a given collection of people should think, act and feel as they go about their daily affairs”. Hofstede’s [3] definition of culture is: “Culture is the collective programming of the mind, which distinguishes the members of one human group from another”. By this definition, Hofstede emphasises that culture is not a property of the individuals, but of groups. It is a collection of characteristics possessed by people who have been conditioned by similar socialisation practices, educational procedures, and life experiences. There are cultures of a family, a tribe, a region, a national minority, or a nation. Hofstede [3] states that everyone belongs to a number of different groups and categories of people at the same time. He calls this mental programming within people, corresponding to different levels of culture, for example at: · a national level according to one’s country; · a regional and/or ethnic and/or religious and/or linguistic level ; · a social class level, associated with educational opportunities and with a person’s profession; · An organisation or corporate level according to which the employees have been accustomed by their work organisation. He proposes that the next step is to identify what problems are common to all societies, through conceptual reasoning and reflection on field experiences, as well as through statistical studies. Following his comprehensive study [3]
Hofstede identified four dimensions of culture, which represent common problems with different solutions from country to country: · · · ·
Social inequality, including relationship with authority; The relationship between the individual and the group; Concepts of masculinity and femininity; Ways of dealing with uncertainty, relating to control of aggression and the expression of emotions.
Culture seems to be a source of competitive advantage. In order to be able to use culture as a competitive advantage we have to learn how to deal with different cultures.
5.
Comparative studies
5.1.
Cross-national studies
Thinking in comparative terms is inherent in sociology. Recent globalising trends have generated more cross-national studies. All the eternal and unsolved problems inherent in sociological research are unfolded when engaging in cross-national studies [22]. Oyen [23] found in her literature review that although some researchers disagree about goals and the theoretical framework for cross-national studies they agree on basic rules of scientific analysis. These rules are concerned with constructing concepts and typologies, and securing ties between data and theory, as well as making use of judgement. According to her, country comparisons underpin much of the empirical foundation of macro social, economic and political theory. These country comparisons dominate how we think about political and social systems - groups and organisations and our effort to discipline ideas. In this sense, she considers that we talk about comparative methods, referring to research about any impact of macro or more encompassing systems on micro ones, and vice versa. Social science disciplines compare countries, anthropologists compare culture and institutional change. She also believes that one scientific objective is to reduce variance or variety by putting countries into general categories or groups or typologies on dimensions asserted to be theoretically significant, as for example wealth, democracy, size, culture, socialism etc. She also points out that different years and time intervals can have different meanings for different countries and activities. Strauss et al. [24] also emphasise that any classification system should include both spatial and temporal dimensions. The conclusion one can draw is that it is important that selecting countries and time points has to be theoretically justified.
5.2
National Cultures
Hofstede [3], who is a native of the Netherlands, undertook in 1966 extensive research in social psychology. 116.000 questionnaires were completed by IBM-employees at all levels, from unskilled workers to top managers, located in 50 developed and less developed nations. The questionnaires were administered in the language of each
country; a total of 20 languages were used. The principal difference among the respondents was culture; all of them were otherwise similar because they were carefully matched for other characteristics such as age, sex, and job category and they all worked for the same corporation, IBM. Using this unprecedented quantity of data, Hofstede was able to distinguish four key elements, or "dimensions", of culture as described below: Hofstede consider that the basic problem areas correspond to the following dimensions: ·
·
Power Distance : Power Distance Index (PDI) indicates the extent to which a society accepts the fact that power in institutions and organisations is distributed unequally among individuals. In small PDI countries subordinates and superiors consider each other as existentially equal and decentralisation is popular, while large PDI countries subscribe to authority of bosses and centralisation. Collectivism / Individualism:Individualism indicates the extent to which a society is a loose social framework in which people are supposed to take care only of themselves and their immediate families. Collectivism is a tight social framework in which people distinguish between in-groups and out-groups and expect their ingroup to look after them. In individualist countries people are supposed to take care of themselves and remain emotionally independent from the group. The dominant motivation is selfinterest. In collective societies the concern is for the group. Individuals define their identity by relationships to others and group belonging.
·
Femininity / Masculinity: Masculinity indicates the extent to which the dominant values in a society tend toward assertiveness and the acquisition of things. In masculine cultures importance is placed on assertiveness, competitiveness and materialism in the form of earnings and advancement, promotions and big bonuses. Femininity indicates the concern for people and the quality of life. In feminine cultures the concern is for quality of relationships and the work of life, nurturing and social well being.
·
Uncertainty Avoidance: Uncertainty Avoidance indicates the extent to which a society feels threatened by ambiguous situations and tries to avoid them by providing rules, believing in absolute truths, and refusing to tolerate deviance. In weak uncertainty avoidance countries anxiety levels are relatively low. Aggression and emotions are not supposed to be shown and people seem to be quiet, easy-going, indolent, controlled and lazy while in high uncertainty countries people seems to be busy, fidgety, emotional, aggressive and active.
All the four dimensions are a continuum between two extremes and only very few national cultures, if any, are wholly at one or the other extreme.
5.3
Organisation Culture
Organisation culture is often reflected as the results of management activity [25] or is looked at through manifestations like ‘rites’ and ‘ceremonies’ [26, 27]. Others have used the structure of the organisation to explain the corporate culture. The Bureaucratic Organisation and the Collectivist-Democratic Organisation analysed by Lofland [28] is shown in figure 2. The Bureaucratic Organisation has many similarities with Hofstede’s high Power Distance Index and Individualist Index, while the Collectivist-Democratic Organisation has many similarities with Hofstede’s low Power Distance Index and Collectivist Index. Dimensions
Bureaucratic Organisation
Collectivist-Democratic Organisation
Authority
Hierarchical Organisation Incumbency (trust) in office and expertise. Compliance to fixed rules Fixed rules, Calculability and appeal of decisions Social control through direct supervision, standardized rules and sanctions Ideal of impersonality Relations: role-based, segmental, instrumental -Employment based on specialized training and formal certification - Advancement base on seniority or achievement Remunerative (reward) incentives are primary
Collectivity as a whole Authority delegated temporarily Compliance open to negotiations
Rules Social Control
Social Relations
Recruitment and Advancement
Incentive (motivating) Structure Social Stratification Differentiation
Minimal stipulated rules Individual decisions Social control personalistic or moralistic Appeals Ideal of community Relations: holistic, personal of value in themselves -Employment based on friends, social-politic values, personality attributes, informally assessed knowledge and skills -No hierarchy of promotion Normative and solidarity incentives are primary. Material incentives are secondary. Reward differentials are limited by the collectively Minimal division of labour Administration combined with performance task Generalization of job functions: holistic roles
Isomorphic distribution of prestige, privilege and power Dichotomy between intellectual and manual work and between administrative task and performance task Maximal specialization of jobs and functions Figure 2. Comparison of two Ideal Types of Organisations adapted from Lofland [28]
6.
Research strategy in this study
6.1
The Research Methodology
In this research the general aims are to find out how culture influences the implementation of Software Quality Management Systems, and to see if it is possible to develop a model which will take culture into consideration when implementing Software Quality Management Systems. The Research Method in this research will be a contemporary comparative multimethod using both quantitative and qualitative investigation Contemporary research is according to Orlikowski [29] a great range of research perspectives that operates concurrently. Hirschheim [30] express the view that there is no correct method of research. Multiple methods, often called multimethod, can be used concurrently to achieve a valid research result. The quantitative investigation will be a survey collecting hard data by using a postal questionnaire. The results from the questionnaire will be analysed statistically according to Oppenheim [31]. Subsequently, a qualitative method in form of case studies will be performed in order to address different aspects of the research problem, to confirm the findings from the questionnaire and to prove the hypothesis.
6.2.
The Research Hypothesis
The main hypothesis in this research is: Cultural factors intervene in the successful application of Software Quality Management Systems. Two cultural variables are identified, namely national and organisational culture Two variables of successful application of Quality Management Systems are identified: · ·
The existence of quality oriented management procedures, similar to the procedures identified in Capability Models; The awareness of quality issues amongst the workforce.
The main hypothesis can thus be broken down into four sub-hypothesis: · · ·
·
National Culture affects take up of Software Quality Management; National Culture influences awareness of quality issues; Organisational Culture affects take up of Software Quality Management; Organisational Culture influences awareness of quality issues.
Analysis of Research Variables 6.3.1
National culture
National culture will be established via Hofstede’s original categories, although these will be subject to some validation within the investigation.
6.3.2
Organisation Culture
Organisation Culture will be measured by placing each organisation on a scale in the Bureaucratic/Collectivist-Democratic spectrum. Within each nation some crosscomparisons between multinational and purely local organisations is expected to yield some interesting variations.
6.3.3
Existence of quality oriented management procedures
The extent of quality oriented management procedures will be determined by using the Process Indicators and the Process Management Indicators adopted by SPICE, although it is not intended to adapt any formal capability assessment. In SPICE software processes are grouped into five process categories (Customer, Engineering, Management, Organisation and Support) according to the type of activity they address. Every process category contains different processes and in every process there are primary activities called base practices. According to the SPICE process evaluation the process is broken down in the following variables: · Process performance; · Adequate resources including people for performing the process; · Responsibilities; · Documentation of the process. To provide guidance to the assessors on what to probe for in the organisation to determine whether the generic base practices have been adequately implemented there are indicators called Process Indicators and Process Management Indicators. The Process Indicators provides guidance to the assessors on how to judge a base practice to determine its existence or adequacy rating. Process indicators define work products, which are the results of the execution of the practices within an organisation, and their characteristics and relate them to each base practice. Process Management Indicators are associated which each generic base practice in capability levels 2-5, and are used to provide assistance in rating generic practice adequacy and help to identify the ability of the organisation to manage a process effectively. The Process Management Indicators also provide a structured way for recording their findings. During an assessment, process indicators in combination with process management indicators are used to give the assessor a view of process capability. In SPICE the Process Management Indicators are classified by levels, while the author has both grouped and summarised them according to the type of activity they address. The Process Indicators and the Process Management Indicators have been adjusted depending on if they can be considered to address Software Quality awareness.
The following figure shows the basis for the design of questions. The double-hollow arrows denote direct questions whilst the single line denotes indirect questions. Basis SPICE Base Practices
Questions Quality Policy
Process Indicators Process Management Indicators
Awareness of Quality
Figure 3. Software Quality Questions
6.3.3. Awareness of software quality amongst workforce The determination of quality awareness uses a more intricate instrument. Indirect expressed questions based on the Process Indicators and the Process Management Indicators will be posed to different people with different responsibilities and specialities at different levels in the organisation. This is expected to measure both the breadth and the depths of quality conscious in the work force in comparison to the extent of quality indicators established above.
6.4
The Questionnaire
6.4.1
Questionnaire Objectives
A questionnaire will be used as a tool to explore the hypotheses and correlate the variables described above.
6.4.2
Research Population
The questionnaire will be sent to multinational organisations in 4-5 countries depending on in which countries the organisation operates, and also to some purely local organisations in the same countries. Although the extent of quality oriented management procedures will be determined at an organisation level; the determination of awareness of software quality will be sought from several sources within the organisation. People who should answer the questionnaire should be: · Working in software development projects · Managing software development projects · Dealing with software quality issues.
7.
Conclusions and further work
In this paper cultural factors, which may have a bearing on successful adoption and implementation of Software Quality Management, have been discussed. Software quality variables and the cultural variables have been analysed and a number of hypotheses have been derived. Using Hofstede’s four dimensions of national culture as an underlying discriminator, a questionnaire has been designed in order to determine the adoption of quality-oriented software processes and the awareness of them within the workforce. Further publications will describe the fieldwork in different countries and the analysis of the results.
References 1
2
3 4 5 6 7 8 9 10 11 12 13 14
Siakas Kerstin V. The Effect of Cultural Factors on the Implementation of Software Quality Management Systems, 5th Software Quality Conference, Dundee, Scotland, 9-10 July, 1996 Mohamed Walaa-Eldeen A., Siakas Kerstin V. Assessing Software Quality Management Maturity (SQMM): A new model incorporating technical as well as cultural factors, 3rd international conference on Software Quality Management SQM95, April 1995, Seville, 1995 Hofstede Geert. Cultures and Organisations, Intercultural co-operation and its importance for survival, Software of the mind, McGraw-Hill International, UK,1994 Fisher Ger, Verveers Frans, Drenth Yvonne. A Software Process and Product Improvement Model, 5th Software Quality Conference, 9-11 July, Dundee, 1996 Jack, Rickard. Personal Issues in SoftwareCost Estimation, 5th Software Quality Conference, 9-11 July, Dundee, 1997 Kondo Yoshio. Importance of Employee Motivation in TQM: 5th World Congress on Total Quality, New Delhi, Feb.1995, pp.46 – 52 Humphrey W. A Discipline for Software Engineering, Addison Wesley, USA, 1995 Humphrey Watts S. Managing the Software Process, Addison Wesley, 1989 Järvinen J. On comparing process assessment results: BOOTSTRAP , and CMM Software Quality Management, SQM94 in Edinburgh, pp 247-61,-94 Paulk Mark C., Curtis Bill, Chrissis Mary Beth. Capability Maturity Model, Version 1.1 IEEE Software July 1993, pp 19-27, 1993 Paulk Mark C. Comparing ISO 9001 and Capability Maturity Model for Software Software Quality Journal 2, 1993, pp. 245 – 256, 1993 Ross M., Staples G. Maintaining quality awareness, 3rd int. conf. on Software Quality Management SQM95, Apr. 1995, Seville, pp. 369-75 Haase Volkmar and Mesnarz Richard. Bootstrap. Fine-Tuning Process Assessment IEEE Software, July 1994, pp. 25-35, 1994 Haase Volkmar H.Bootstrap - Measuring Software Management Capabilities, First Findings in Europe, the fourth IFAC/IFIP Workshop, Austria, May, 1992
15 Rout Terence P. SPICE: A Framework for Software Process Assessment in Software Process-Improvement and Practice, Pilot Issue 1995, pp. 57-66, 16 Rout T. P., Nielson M. P., Gasston J. L. Experiences with the use of different approaches to software process assessment, Software Quality Management, SQM94 in Edinburgh, 1994 pp.767-779, 17 Sanders Marty (Editor). The SPIRE Handbook, Better, faster, Cheaper Software, Development in Small Organisations, SPIRE project, The European Community, ESPIT/ESSI 23873, 1998 18 Siakas Kerstin V., Berki Eleni, Georgiadou Elli, Sadler Chris. The Complete Alphabet of Quality Software Systems: Conflicts and Compromises, 7th World Congress on Total Quality & Qualex 97, New Delhi, India, 17-19 Februar 1997, 19 Curtis Bill, Heflleey William E, Miller Sally. Overview of the People Capability Maturity Model, CMU/SEI_95-MM-01, 1995 20 Boström Robert P. MIS Problems and Failures: A Socio-technical Perspective Part I: The causes, MIS Quarterly Sept. 1977 pp. 17-32, 21 Hall E.T. The Silent Language, Doubleday, New York, 1959 22 Hoecklin Lisa. Managing Cultural Differences, Strategies for Competitive Advantage, Addison-Wesley Publishers Ltd, 1995 23 Oyen Else. Comparative Methodology, Theory and Practice in International Social Research, Sage Publications Ltd, 1992 24 Strauss Anselm, Corbin Juliet. Grounded Theory in Practice, Sage Publications Ltd., 1997 25 Lessem R. Global Management Principles, Prentice-Hall, London, -89 26 Hodge B.J., Anthony W.P. Organisational Theory, Allyn and Bacon, Boston, 1988 28 Lofland John, Lofland Lyn H. Analytic Social Settings, A guide o qualitative observation and analysis,Wadsworth Publishing Company,1995 27 Joynt Pat, Warner Malcolm . Introduction: Cross-cultural perspectives in Managing across Cultures: Issues and Perspectives, by Joynt Pat, Warner Malcolm (Ed.) International Thomson Business Press, 1996 29 Orlikowski Wand J., Baroudi Jack J. Studying Information Technology in Organisations: Research Approaches and Assumptions, Information Systems Research 1991,2:1, 1-28 30 Hirschheim R.A. Information Systems Epistemology: An Historical Perspective In Information Systems Research, Issues, Methods and Practical Guidelines Edited by Galliers, Blackwell Scientific Publications 31 Oppenheim A.N. Questionnaire Design, Interviewing and Attitude Measurement, Pinter Publishers Ltd, 1996
