This book is about a systems design methodology called ETHICS - standing for the Effective Technical and Human Implementation of Computer-based Systems. The first part of the book will present the arguments and beliefs which lie behind the development of ETHICS. The second part will provide a detailed step-by- step account of the method and how to use it.
ETHICS has three objectives related to the management of change. First, it seeks to legitimate a value position in which the future users of computer systems at all organizational levels play a major part in the design of these systems. The argument here is that people should be able to influence the design of their own work situations and that if this kind of intervention is encouraged then there are likely to be both job satisfaction and efficiency gains. Job satisfaction gains happen because the members of the user group whose job satisfaction is going to be affected by the system's change are better able to diagnose their own job satisfaction needs than any outside group of specialists; efficiency gains because these same people are likely to have an excellent knowledge of day-to-day information needs and work problems and can make useful contributions to the specification of the first and the solution of the second. Also, they will be committed to operating efficiently a system which they have themselves partly designed.
The second objective is to enable groups concerned with the design of computer systems to set specific job satisfaction objectives in addition to the usual technical and operational objectives. Here it is argued that unless job satisfaction and quality of working life objectives are made explicit, and the computer system and associated organization of work designed to achieve these, the human impact of a new computer system will be unpredictable because it has not been consciously planned for. The result can be that the new system will have undesirable human consequences such as a routinization or de-skilling of work, or other features that are not welcomed by the user. Staff in the user department may then respond in a negative way, refusing to operate the system or ensuring that it runs at low efficiency; and, in addition, absenteeism and labour turnover may increase. All of these responses can impose high financial costs on management.
The third objective is to ensure that any new technical system is surrounded by a compatible, well functioning organizational system. Today, technical design is only one part of a large, complex, design process which includes everything that both interacts with and surrounds the technical system in the total design task. This covers direct man-machine interaction; the design of work procedures, individual jobs and workgroup activities; the specification of roles and relationships within the department or functional area where the new technical system is being introduced, and the creation of good boundary management techniques.
Experience has shown that user involvement, clear job satisfaction objectives and a recognition of organizational factors will greatly assist the successful planning, design and implementation of computer-based work systems. ETHICS is a way of enabling everyone concerned with the introduction of new technology to consider human as well as technical factors when embarking on the design of a new system. In other words, to take what is called a "socio-technical" approach to the systems design process. A socio-technical approach is one which recognises the interaction of technology and people and produces work systems which are both technically efficient and have social characteristics which lead to high job satisfaction. Such an approach takes account of the fact that different individuals and groups have their own needs, interests and values and that these must be met if employees are to willingly and enthusiastically accept major change.
Managers often argue that resistance to change is a natural human condition and one which they have to live with. An alternative proposition is that people will actively welcome change if they believe that it brings with it personal benefits. If they see it as a threat they will oppose it for very logical reasons, some related to their previous experience with computer systems. Bad experience in the past will have been contributed to by the fact that technical specialists have been keen to optimise the use of a technology which they knew and understood and this led them to design systems which had a high technical competence but were poor at catering for human needs, such as a desire for job satisfaction. User groups which had no active role in the design process were unable to challenge the technical knowledge of the specialists and this often forced them into a dependency relationship and the subsequent acceptance of a technical system which did not adequately meet their needs. This produced low commitment to the system, together with increased resistance to any future change.
Systems that were received with reluctance and antagonism often placed, and can still place, a company in a high risk situation. It could incur financial risks through introducing expensive computer systems which were then operated at a low efficiency, or reduced job satisfaction and increased labour turnover. It could also incur organizational risks, for a poorly functioning department could spread dissatisfaction and inefficiency through departments which interacted with it. In addition there were human relations risks, for a new technical system that is introduced against the wishes of a user department has the potential to produce serious industrial relations problems.
The strong technical bias of conventional systems design
has not been without criticism. Professor Lou Davis, an
engineer who pioneered the socio-technical approach in
the United States, has written:
"No clear objectives concerning roles for men as men are visible, although objectives are clearly defined for men as machines. When man is considered only as a link in a system, design rules do not exist for allocating appropriate tasks to man. Nor are there design rules for designing task configurations to make complete and meaningful jobs ....."
Computer systems design, although it has been with us for only a comparatively short period of time, appeared to become structured and formalised very quickly. It was associated with a problem solving philosophy which was accepted uncritically by most systems designers. This philosophy saw the design of computer systems as a technical process directed at solving problems which were defined in technical terms.
Computer systems design when applied to office functions usually took the form of breaking down the manual system into its component parts, identifying which of these were essential to the problem-solving function the department was there to perform, and reassembling these in a form appropriate to the logic of computer usage. The new system would usually include a more rigorous set of controls than the old to ensure that it worked with the required speed and accuracy. In this way a new discipline followed an older engineering discipline which had always taken this technicalrational approach, expecting the human being to conform to the demands of the machine.
In the past, and also today, the values of technologists and managers have influenced them to look primarily for efficiency gains when they were introducing technical systems and they have frequently designed these systems to achieve such goals as a reduction in staff numbers and tighter operational controls. They have therefore focussed on structural elements in the change situation and used machines to replace people and to monitor more strictly the work of those people who remained. Yet because people are an integral part of all work systems, and people have values, attitudes and psychological needs, a neglect of their interests can throw a change situation into a state of disequilibrium so that instead of the change producing the desired increase in efficiency the new system operates more inefficiently than was anticipated and gains may not outweigh the costs of innovation.
It is interesting to speculate how this particular design philosophy has arisen and become so widely accepted. It seems to have its roots in the analytical method of work study and in the early scientific management attempts to reduce human error by breaking work down into small bits and increasing work controls; logical at a time when skilled labour was in short supply. It also seems to owe part of its approach to the engineering philosophy of designing machines and assuming that the human being will adapt to their requirements, whatever these may be. This systems design philosophy has been reinforced by the training given to systems analysts which has ignored the fact that unless a work system is completely automated, it is a man-machine system. The human being cannot operate at a high level of efficiency without the machine and the machine cannot operate at all without the human being. Such a system is unlikely to function effectively if this mutual dependency goes unrecognised and only the machine part of the system is consciously designed.
Successful change requires a number of things. Objectives have to be set and attained; an existing, complex, organizational and technical system has to be helped to adapt to a new and perhaps more complex system, equilibrium has to be restored so that the new system can operate effectively and this state of equilibrium has to be maintained into the future until there is a need for another major change. These change processes involve "the setting and attainment of system objectives", "adaptation", "integration" and "stabilisation" - stabilisation requiring that once new patterns of behaviour have been successfully initiated, they must be established and reinforced.
Traditionally, system objectives have been set by top management and computer technologists. But increasingly other groups are asking or demanding to be involved in this process, in particular, line management, white and blue collar workers and the trade unions. Each of these groups tend to have different interests, so that when decision taking is shared a great deal of conflict is brought into the open and has to be resolved as part of the systems design process. This means that systems design today and increasingly in the future will involve negotiation and the reconciliation of different interests. This is a major deviation from the situation of ten years ago when only the technical group understood computers and, if they were persuasive and good salesmen, they were allowed to design computer systems more or less as they wished. Today we have two new factors in the change situation. First, many non-technical groups are becoming familiar with, and knowledgeable about, the advantages and disadvantages of technical systems. Second, the distribution of power has changed in many Western European firms and powerful trade unions will resist any change that they do not see as directly in the interests of their members. As managers are increasingly joining unions this resistance no longer applies only to the shop floor or to clerical groups, it is moving up the organization hierarchy.
The kinds of objectives that are set for technical systems are usually subject to a number of constraints. These may be the amount of money that is available, the level of development of current hardware or software, legislation, and in Britain this has had a major impact on civil service computer applications, and what the various interested groups in the organization want and are prepared to accept. Another problem is that the objectives set at one moment in time may become irrelevant or appear misjudged by the time the system is installed. Let us consider, for example, a computer system designed in the late sixties when there was a tight employer's market for labour and skilled clerical staff were very difficult to obtain. The system was therefore deliberately designed to routinise work so that lower calibre clerks could be recruited. This system is still in operation and is unlikely to be changed for some considerable time. Today the labour market has dramatically altered and many highly qualified young people are clamouring for these jobs; the result is a system which will cause frustration and disillusionment with work to intelligent young people who no longer have the opportunity to find more challenging jobs. The moral here is that both system objectives and the final system need to be adaptive. Objectives that, once set, are never revised may merely enable an organization to tackle today's problems on the basis of yesterday's obsolete needs.
Rigid technical systems which cannot be adapted to changing human needs will produce an alienated workforce.
The process of adaptation is concerned with moving from one kind of organizational structure and state to another and the means by which this change is assisted to take place smoothly and successfully. It is what normally happens during the implementation phase of a new system. Rapid adaptation does not easily come about of its own accord and there is a need for philosophies, facilities and strategies to assist the process. These will include values, attitudes, incentives and mechanisms for conflict resolution for, in the same way as objective setting, adaptation requires the reconciliation of different interests. Adaptation will be easier for some groups than for others as they may feel less threatened by the change, they may be more change minded and accustomed to coping with new situations, or they may have power and so be able to mould the change situation to their own interests. Adaptation is therefore a political and negotiating process in which a compromise rather than an optimal solution is likely to be achieved, although the aim should be the easy and speedy adoption and acceptance of change by all the groups involved.
Integration is the action taken, once the system has been designed and is being implemented, to ensure a new situation reaches a state of equilibrium. This involves bringing the different components of task, technology, people and company objectives together into a viable and stable relationship as shown in the diagram:
Technology(with requirements and constraints)
People(with values and needs)
Tasks(which require motivation and competence
Organizational environment(reflecting company objectives)
The relationship between these four variables needs to be stable but it should not be static. Organizations must be able to respond to new pressures from their environments while at the same time either maintaining a state of equilibrium or being able to make adjustments which restore equilibrium if internal relationships are disturbed. The introduction of a new technical system is likely to disturb each of the variables in the diagram. A new level of technology will bring with it a new man-machine relationship incorporating both opportunities and constraints. Because tasks are influenced by technology, the task structure of departments using the system will be altered. New tasks mean that new demands are made of people and this will affect job satisfaction positively or negatively depending upon whether the new situation meets their values on what is desirable work.
Finally the technology, tasks and people variables interact with an internal organizational environment which provides a structural context for the achievement of the company's objectives and this interaction may start the looping process again by making new demands of technology.
Integration requires that adaptation produces a harmonious relationship between the technology, task, people and organizational variables.
Once integration has been achieved it has to be maintained, and this means that the new patterns of relationships must be stabilized so as to meet the values and interests of the groups involved over a period of time. But, a successful relationship on the four variables for one group may make it difficult for another group to achieve a similar relationship. To avoid "one man's job enrichment becoming another man's job impoverishment" imaginative solutions or even compromises may be required. The maintenance of a state of equilibrium into the future requires processes for socializing and educating new group members; it also requires mechanisms for controlling tension when the fit between the variables slips and for resolving conflict when major divergences are experienced.
When technical systems are designed it is usually not difficult to establish a successful relationship between technology and the attainment of company objectives. But this concentration on two of the four variables leads to a neglect of the relationship between task structure and employee job satisfaction needs. Because jobs are not consciously designed to meet these needs, the systems are not well integrated in human terms. In recent studies of a number of very large computer systems it was found that objectives were set solely in terms of technical and efficiency variables. The computer system was introduced with the dual objectives of reducing costs through eliminating staff, while maintaining the same level of service to the client group. The values behind these objectives were efficiency values, and objectives relating to employee job satisfaction or improving the quality of working life were not seen as valid. The result was that when top management and the technical specialists evaluated the system after implementation, both groups were extremely pleased with its success. The objectives set at the start of the design process had been attained.
Interviews with the clerks who were operating the system produced a very different picture, however. There were complaints that work had been routinised and jobs split into two halves by the computer. Clerks who were previously able to progress a job through from start to finish were now left with the boring task of inputting data to the computer and the slightly more interesting task of handling computer rejections. The system objectives that had been set had been successfully attained but were they the right objectives to produce a stable well functioning system? As absenteeism and labour turnover increased and it became more difficult to recruit staff, management began to recognise that a neglect of human objectives had led to high human costs. Adaptive efforts to secure a smooth transition from the old situation to the new seem often to be ill-thought out and patchy. The systems designer assumes compliance in the user and appears to believe that providing communication is effective and the user understands what will happen, the system will be accepted. Unfortunately compliance as a user characteristic seems to be a part of history. Today's users may listen to what they are told but, once the system is in and they can actually see its results, they are likely to say that they do not like it and will not use it until it is changed. The systems designer then has to adapt to the users, and the cost of redesigning the system and changing it into a form which the user will accept may be high.
When efforts are made to assist adaptation they are often partial and involve some groups but not others. In another large organization senior management were determined to gain the knowledge to design their own system as they mistrusted the ability of outside technical experts to understand the complexities of their business. Three managers played a major role in all the design processes with the result that they produced a computer system that fitted extremely well with the needs of the business and their needs as managers. In addition, because of their involvement in the design process, they understood in detail how the system worked. And so at managerial level an excellent adaptation process had taken place. But when the clerks who operated the system were interviewed, they said that their work had been degraded by the computer and had lost its interest. The result was a low morale group at a level where the organization was most vulnerable, for these clerks interacted directly with customers and provided the customer service. In this situation the adaptation of management was facilitated through their involvement in system design. It was not thought necessary to help the clerks to adapt but assumed that adaptation would occur spontaneously once the system was installed.
In another organization a central computer services department designed a very large system for about one thousand local offices. Strategies to assist adaptation are always difficult in this kind of decentralised situation but the only efforts made were the creation of an implementation team which visited the local offices immediately prior to their conversion to EDP, and the training of local office staff in their new tasks. In these local offices, instead of integration having taken place and a good fit achieved between the technology, the needs of staff and the needs of the local unit, a situation in turmoil was found. These clerks were responsible for customer liaison and derived a great deal of job satisfaction from giving a helpful and efficient service to the public. However, incorrect assumptions made by central computer services and incorporated into the computer system meant that this level of service could no longer be provided. Information which had previously gone direct to customers from the local offices, now went by post from the central computer centre and frequently did not arrive on time because of delays in the mail. Similarly, records previously held in the local offices were now in the computer and clerks could not answer customer queries until up-to- date information was sent via the mail to the local office. The result was frustrated clerks and irate customers and a system which was a long way from settling down and moving into a state of equilibrium.
Current practice in relation to stabilization, or ensuring that once integration has been achieved this is maintained, can also be unsatisfactory for a number of reasons. First, unless the organization into which the new system has been introduced is entirely free from outside pressure, stabilization will require a constant, regular, adaptation to changes in the external situation. Many technical systems are rigid and very difficult or expensive to modify and as time passes they become constraints instead of facilitators. Second, it is unusual for organisations to have a clear understanding of what contributes to their stability, or to check, once the system is operational, that a stable state has been reached. Many computer systems are installed and then neglected as the systems designers move on to the next project, leaving the user desperately trying to cope with a system that is a hindrance rather than a help. The local office application described above was a case in point. User managers repeatedly complained to the systems designers about their problems, but were regarded as either foolish or incompetent. The technical view was that there was nothing wrong with the system, the fault lay in the user's inability to handle it.
Many systems designers do not appear to realise that with their present approach they are designing only partial systems and that to do their job effectively they must be competent designers in both technical and human areas. This book aims to help them with the human design task. If a more humanistic approach to work is not adopted by systems designers, it can be predicted that some important new technology based work systems will prove unacceptable to employees and their trade union officials; and also to managers, who will not welcome work systems with built-in industrial relations problems. A system that is designed to achieve objectives defined solely in technical terms is likely to have unpredictable human consequences. The reason for this is that technical decisions taken at an early stage of the design process will impose constraints on the organization of the human part of the system. Because the human part of the system is either not included in the design process, or not considered until a very late stage, the human consequences of these decisions may not be recognised until the system is implemented. This can lead to the technical system influencing the human system in a way which was never envisaged by the systems designers. Well functioning work systems require the successful integration of a number of different sets of values, interests and needs and reconciling these is not easy to accomplish. A new system will incorporate the values of the technologists who understand the hardware and how it can be used efficiently; but it has to fit with the values of top management who are interested in receiving the kind of information that will help them to make policy decisions that ensure the long term viability and prosperity of the company; it has to fit with the needs of line managers who are responsible for producing an efficient product or service; and it has to fit with the values of white and blue collar workers who would like to work in a satisfying and stimulating environment that meets both their psychological and economic needs. A well integrated system requires that all these needs are identified and understood and, even more difficult, are catered for in the system design process.
A new systems design approach is badly needed— one which not only attempts to avoid human ill effects but has the positive aim of encouraging the setting and achieving of human objectives as an integral part of the work-design process. Such an approach implies that those associated with the introduction of new work systems - systems designers, engineers, managers, trade union officials and clerical or production workers—are prepared to take up a value position which accepts that technology can and must be used to achieve human as well as technical objectives. One important means for achieving this result is the sharing of the design task and the participation of users at all levels in the design of the systems which they will eventually operate.
As an important feature of ETHICS is the
participation of users in the systems design
process, it is necessary to be clear about what
participation means and what it involves.
The one thing that can be said with
certainty about participation is that it is not
a new concept, although it may sometimes have
been given other names such as democracy,
involvement, sharing, co- operation, etc. The
Greeks used it to describe a certain kind of
decision taking. For them a decision was
participatively taken if the answer to the
question "who takes it?" was "more or less
everybody". This kind of democratic
decision taking contrasted with
meritocratic in which decisions were taken by an
elite, and with autocratic m which they were
taken by one person. Later writers on participation have been of a
very different kind from the early philosophers.
Many were contributors to the human relations"
movement in industry and they include famous
names from the United States as Mayo, McGregor,
and Likert. The Tavistock Institute in
England had a major influence on
organizational participation from the 1950's
onwards as did Norwegian social scientists such
Thorsrud and Herbst. These groups were
concerned with questions about authority,
motivation and job satisfaction. All of these
were seen as being increased through
greater employee participation in decisions
which concerned the work environment.
Mayo's very broad definition of participation is
that it is polyphasic. It cannot be
represented in an organizational chart or
engineering blueprint. A democratic system has
no fixed concentration of authority at the top.
On the contrary the focus of authority moves
from place to place according to the demands of
the situation. McGregor's more conservative model is that
participation "creates opportunities under
suitable conditions for people to influence
decisions which affect them", while Likert
describes his model of participation as "all
subordinates in a work group who are affected
by a decision are involved in it" (a work
group is defined as a superior and all
subordinates who report to him).
What is Participation?
The one thing that can be said with certainty about participation is that it is not a new concept, although it may sometimes have been given other names such as democracy, involvement, sharing, co- operation, etc. The Greeks used it to describe a certain kind of decision taking. For them a decision was participatively taken if the answer to the question "who takes it?" was "more or less everybody". This kind of democratic decision taking contrasted with meritocratic in which decisions were taken by an elite, and with autocratic m which they were taken by one person.
Later writers on participation have been of a very different kind from the early philosophers. Many were contributors to the human relations" movement in industry and they include famous names from the United States as Mayo, McGregor, and Likert. The Tavistock Institute in England had a major influence on organizational participation from the 1950's onwards as did Norwegian social scientists such Thorsrud and Herbst. These groups were concerned with questions about authority, motivation and job satisfaction. All of these were seen as being increased through greater employee participation in decisions which concerned the work environment. Mayo's very broad definition of participation is that it is polyphasic. It cannot be represented in an organizational chart or engineering blueprint. A democratic system has no fixed concentration of authority at the top. On the contrary the focus of authority moves from place to place according to the demands of the situation.
McGregor's more conservative model is that participation "creates opportunities under suitable conditions for people to influence decisions which affect them", while Likert describes his model of participation as "all subordinates in a work group who are affected by a decision are involved in it" (a work group is defined as a superior and all subordinates who report to him).A definition that fits user involvement in systems design defines participation as: "a process in which two or more parties influence each other in making plans, policies or decisions. It is restricted to decisions that have future effects on all those making the decisions or on those represented by them".
All definitions recognise that participation involves more than one set of interests and that it is concerned with decision taking. There is, however, little discussion of the nature of the different interests involved or of the subjects with which the decision taking is or should be concerned. If we wish to use a participative approach in systems design it is necessary to consider the functions, structure and processes of participation that can be associated with the introduction and management of technical change in industry.
The arguments of the classical theorists were based on a vision of a "common good" and on how universal involvement in decision taking could help secure this. The nation state was seen as committed to ideas of justice and freedom, its problem was how to realise these. Parallel notions were concerned with the rights of the individual and the importance of personal autonomy, which could only be secured if men could decide for themselves what they wanted to do.
The arguments for participation today tend to be rather different. In industry participation is supported for many different reasons. Management may decide to try a participative approach because they agree with the early philosophers that participation is morally right - people should be able to determine their own destinies. Or they may see participation as helping to produce an interested and committed workforce and therefore assisting in the avoidance of industrial relations problems. They may believe that participation is a valuable educational experience and provides understanding and knowledge that can assist an organization to more effectively realise its objectives or, alternatively, that such knowledge will enable any bargaining that takes place to do so with all interest groups operating from an informed position. These are all positive reasons for participation. It is clear, however, that some managements favour participation for what could be regarded as negative reasons. They see it as a way of persuading their employees to accept changes that otherwise might be rejected. Participation then becomes a manipulative tool and is carried out without any serious attempt to make it truly democratic.
Trade union arguments for participation can be equally varied. Unions may encourage participation because they see it as a lever for increasing shop floor control over the work situation. They may also support it because they believe it will contribute to industrial democracy and the more equitable sharing of power. But some unions reject participation, stating that it is their responsibility and right to represent the interests of their members and that they can do this effectively without involving these members. Employees are usually interested in participation for different reasons. They see it as a protective device that will enable them to prevent things that they believe to be undesirable from happening. They want to avoid being made redundant, or deskilled. They also hope to have more interesting jobs, better opportunities for promotion and an improved quality of working life. Participation may assist them to achieve these.
The outside observer may have another set of objectives for participation. He or she may believe that active participation leads people to develop responsible characters; that it enhances group harmony. develops a sense of co-operation and of community and produces a willingness to accept group decisions. These are the kinds of arguments that might be of importance to community leaders or politicians. Other arguments for participation are related less to the achievement of specific objectives and more to subjective individual feelings. For example that it enables people to feel free, to be their own masters and in control of their own destinies. Again educationalists might be interested in the arguments for the character building aspects of participation. That it assists people to develop active, non- servile characters and democratic personality structures, and also enables them to broaden their horizons and appreciate the viewpoints and perspectives of others.
Participation then can have different meanings for every individual or group concerned with it. Each interest group is likely to define the word in a different way and to direct the participation process at very different objectives. The challenge for any organization that decides to try participation as a strategy is to create the, kinds of structures and processes that will assist all the different stakeholders to obtain some gains from using this approach. These gains will not necessarily be all of the same kind but they should enable each group to say with conviction "participation has clear benefits for us".
The discussion so far highlights the difficulty of trying to arrive to at precise definitions of participation. Equal difficulty is likely to be experienced when considering how participation should be applied. What should it he used for? The traditional answer is that participation is concerned with decision taking and with the representation of different interests and points of view in this process. The theory here is that a decision publicly arrived at is better understood and therefore likely to be better carried out.' But this says nothing about who implements these decisions once they are taken. Participation is seen as an expression of consent to something being done. It says nothing about how it is done, or who does it.
Any comprehensive discussion of participation must go further than this. It needs to discuss the structure of participation, the content of participation and the process of participation. The structure of participation is concerned with the mechanisms for enabling participation to take place. These can take many different forms. Participation may be a product of an institutionalised political system with democratically elected representatives, or it may be a formally organised vote such as a referendum. It can take the form of pressure groups which make their views known through legitimate channels such as the press or parliamentary representatives.
Alternatively it can be a pressure group that is prepared to use illegitimate means such as sabotage, high-jacking or kidnapping to make its views known. Participation can also be spontaneous and unorganized and take legitimate forms such as letters to the papers, or illegitimate forms such as Luddism, verbal abuse etc.
An employing organization creating a participative structure will be less concerned with the varieties described above and more with decisions such as whether to choose direct or indirect forms of participation. Direct participation is when everyone with an interest can exert influence. Indirect participation is when influence is exerted through intermediaries. If an indirect approach such as a representative group is chosen as the vehicle for decision making, then attention has to be paid to such matters as ensuring that all interests are represented, to deciding how the members of the participative forum are selected or elected and whether a number of groups at different organizational levels are required. When a structure for participation is being created decisions have also to be taken about the extent to which certain groups of participants will be allowed to influence decisions. There can be "consultative" decision taking in which the participants give evidence to the decision makers, this evidence being taken into account when decisions are taken. There is also "democratic" decision taking in which all participants have an equal voice in the decision making process, and "responsible" decision taking in which the participants are given full authority for making decisions in any way they see fit and are not constrained by any laid down procedures. The author has used a number of different participative structures in the firms in which she has contributed to systems design. She has used a "consultative" decision process in which employees contribute ideas to the design process, although the final decision is prepared by the specialist group. She has also used a "representative" approach in which design groups formed from selected or elected representatives of the various interest groups take design decisions and a "consensus" approach in which members of a design group constantly discuss ideas and solutions with their constituents. They then prepare a number of alternative solutions, the final choice being made by the constituents and not the design group. These options of "consultative", "representative" and "consensus" participation can also be used at different organizational levels in the company.
The content of participation is the nature of the issues about which decisions are taken. But it is more than this. It involves a consideration of decision boundaries; that is, what subjects can be considered participatively and what subjects are outside the jurisdiction of the participative group and are seen as executive decisions which are not taken democratically. The process of participation involves the acquisition of knowledge so that decisions are taken from an informed position; it involves learning, the development of effective working relationships over time, the setting and achieving of goals, and the implementation of solutions. It is also concerned with power and in true participation all the parties will have equal power. If power within a participative group is not equal then the interests of a faction can unduly influence decisions. Where power is equal but there are conflicts of interest, the outcome will be negotiated and is likely to be a compromise. Occasionally a meta-solution may be striven for and achieved. By meta-solution is meant a solution that is so effective that it succeeds in meeting the objectives of all the stakeholders. Occasionally a group may find it is impossible to arrive at a solution and the outcome then is a product of default, not intention.
If one ceases to consider participation as a blanket activity applied to all decision taking and considers it in relation to a specific set of needs and problems, then it becomes easier to discuss it with clarity. The author has been particularly concerned with using participation to assist the design of new work systems involving the use of new technology. This task is very appropriate for a participative approach for the following reasons:
It involves a number of interest groups (management, systems designers, clerical or shop-floor users, trade unions, customers, etc.)
It requires knowledge and this knowledge is likely to be spread throughout the different interest groups and not be located solely in one.
The design task is complex and takes time. Participation enables a shared learning process to take place in which each of the interest groups can contribute to the problem solving process.
The various interest groups are likely to have different values, needs and objectives and these can be brought out into the open, discussed and attempts made to reconcile them as part of the participative process.Participation used for design tasks therefore involves a process over time and not merely the giving of an opinion at one moment in time. This kind of participation requires a high commitment from those involved, who may be associated with all aspects of systems development including analysis, design. construction, implementation and evaluation.
Once participation is seen as a process that involves shared problem solving over what may be a prolonged period of time, then a wide variety of options become available. Frank Land has provided some examples of the available options. For example, it has to be decided:
It must be emphasised that with this kind of participation the analysis and design processes of the development of a new system cannot be separated from the decision processes. Small but important decisions are taken at every stage of systems development. For example, decisions relating to the scope and depth of analysis, to the relevance of received information, to the importance of different design options, etc. Participation in these kinds of decisions is an essential prerequisite to acquiring the knowledge which enables a design group to make informed judgements on the major issues.
Comprehensive participation in the total design task for a new system would involve the following:
Consultative, Representative and Consensus Participation Systems design can occur at three different organizational levels. The top where it is concerned with strategic planning, the middle where it covers system definition for a number of divisions, functions or departments and the bottom where it relates to the detailed design of an organizational subsystem such as single department or function. The participative approach can be used at all of these levels although it may take a different form depending on whether it is concerned with higher or lower level systems.
When a new computer system is seen as part of a company's plan for the future, then getting discussion and agreement at all organizational levels is not easy. Even the democratic Scandinavian legislation requires only negotiation between top management and senior trade union officials and does little to ensure employee involvement at the bottom. If present values, or existing management problems, mean that consensus participation throughout the organization is seen as to unwieldy and slow, or undesirable for other reasons, then good consultation becomes even more important when strategic plans are being made. Consultative committees can be used as sounding boards to check that management's vision of the future is not in conflict with that of other employees, and to ensure that particular strategies will receive majority support. This is consultative participation.
The consultative approach is seen as most appropriate for securing agreement on strategic planning objectives. Here the major planning decisions are taken by senior management, probably at Board level, whose hierarchical position enables them to take a broad view of the enterprise's future needs. However, they will only take these decisions after extensive consultation with interested groups lower down the organizational hierarchy and a consultative structure must exist or be created so that this sounding out of opinion can be thorough and accurate.
Once the computer plan has been decided and the broad strategy of design begins it is important to recognise that many different groups will have an interest in the proposed change, and will want to influence this. Here is an opportunity for representative participation. Design groups need to be set up which enable representatives from different user areas to meet together to set objectives and devise tools and tactics for achieving these.
Representative participation is seen as appropriate at the system definition stage when powerful interest groups at middle management level will wish to express an opinion on where system boundaries are to be drawn and on the broad form any future system should take. Although, in most companies, decisions at this level will be mainly taken by management, representative design teams will include some selected or elected representatives of other grades of user staff, the computer professionals and trade union interests.
When the new system has a shape and detailed design for specific local user areas can take place, then consensus design becomes possible, particularly for the design of the organizational context which will surround the new system. A representative design group will still be required, but it now requires representatives from all functions and levels in a particular user department. This design group will see an important part of its task as involving its departmental colleagues in the design activities and in the decision taking on how work is to be reorganized around the technical system.
Consensus participation attempts to enable all the staff in a department to play a part in the design of a new work system. They are involved when efficiency and job satisfaction needs are being diagnosed through feedback and discussion in small groups. As the design team formulates alternative design strategies these will be discussed at staff meetings and the choice of work organization and task structure to be associated with the technical part of the system will be greatly influenced by the views of the staff. Experience has shown that a consensus on a system solution does not always emerge easily and conflicts which result from different interests within a department may have to be resolved first. However, sometimes consensus participation is not possible even at local level and then a move back to representative or consultative participation becomes necessary.
One condition for a participative approach to systems design is that no one must suffer and as many employees as possible must gain from the change. When a computer system is being introduced there is a potential for reducing the numbers of staff and usually guarantees have to be given that there will be no redundancy and that any staff saving will be achieved by not replacing staff when they leave voluntarily. This approach has also to be acceptable to top and departmental management who must be convinced that participation will lead to both efficiency gains and the creation of a more desirable work environment which will help attract and keep staff.
The process of forward planning, including systems design, implementation and evaluation, should be carried out by two types of teams. The first team is responsible for steering the project, it is the direct link between the senior management of the organization and the team responsible for the detailed design. This steering group sets the basic organizational objectives and constraints under which the new system is to be developed. The steering group will generally include the managers of affected departments, official representatives of trade unions and other major interests affected by the new system. The second team is responsible for the detailed systems design and consists of representatives of the department where the new computer system is to be introduced. This team will define the scope of the local problem, scan the environment for new opportunities or changing constraints, analyse deficiencies in the current system, define development goals in consultation with other groups likely to be affected and provide decision makers with assessments of the likely impacts of alternative strategies on these goals, and finally design and test the selected system and the work organization and task structures associated with it. Such a team will include members from the management services department, but their function will be to transfer the required skills to other members rather than to carry out the design themselves. Their role is that of technical consultants to the design team.
Both representative and consensus participation need to have trade union support and involvement in unionised situations, if they are to succeed, and it is becoming apparent that in some companies local union participation is not enough and regional union officials wish to be informed before changes in work organization are carried out.
All strategies for change have their problems and participative design is no exception. Those discussed below are derived from the author's experience.
It must not be assumed that because a group of office or shop floor workers agree to be involved in participative design that this means that they are convinced of management's good intentions. In an engineering firm, suspicion of management's motives in permitting this unusual degree of democracy lasted for some time. It required many statements from management that it had no evil intent for trust to be established.
The author strongly believes that if consensus or representative design is used then the design group should be democratically elected. In the engineering firm managers were worried that an election would lead to the recruitment of the most militant clerks in the department to the design group. They therefore preferred to select the group themselves. Although this produced an admirable representation of the different interests in the department, for a long time the design group were seen as management "blue eyes" and this affected their relationships with their clerical colleagues.
It cannot be assumed that a group of clerks, although they work on similar jobs in the same department, will all have the same interests or favour the same design solution. With a consensus design approach these conflicts of interest will quickly surface and will have to be discussed and negotiated. This is not an easy task for a design group.
A design group formed from clerical or shop floor workers has "no place to hide". Whereas a professional group of technical systems analysts can return to their own department when relationships become difficult, a participative design group must continue to work side by side with its colleagues during the design process. As this inevitably has its conflict periods they are likely to experience stress from time to time and this is something they will not be used to.
In order to avoid conflicts of interest and problems of stress design groups need to have some skills in communication and consultation. In the author's experience it is more difficult to transfer such skills to a design group than it is to give them design expertise.
With a participative design approach the role of the professional systems designers is greatly altered. Instead of being "designers" in the traditional sense they become consultants to the design group. Some systems designers may regard this change of role as a threat to their authority and status.
Ideally departmental managers should have close contact with the work of design groups and an effective strategy is for them to join the group for part of each meeting or for a full meeting at regular intervals. But if departmental managers decide not to become permanent members of the design group because they believe that their presence will have an inhibiting effect on discussion, then they must always be kept in touch with what is taking place. They must also approve the solution that is eventually chosen by their department staff.
An important role in participative design is that of group "facilitator". The facilitator can be an external consultant or an internal employee of the company, although if this last is the case then they should not be from the user department or the group of systems analysts immediately responsible for technical design. They need to have a neutral status and preferably to come from a department such as personnel or management services which crosses many departmental boundaries in its activities. The role of the facilitator is to help the design group to manage the project. This help involves teaching them the ETHICS methodology, keeping them interested and motivated towards the design task, helping them to resolve any conflicts and making sure that important design factors are not forgotten or overlooked. The facilitator must in no circumstances take decisions for the design group or persuade them that certain things should be done or not done. The facilitator's role is to help the design group systematically to analyse their own problems and needs and arrive at an organizational solution that solves the problems and meets the needs.
At the start of a project new design groups are usually confused about the nature of the task upon which they are embarking and uncertain of their ability to tackle it. The facilitator, helped by ETHICS, must assist them to overcome these doubts. One way of doing this is to show how another, similar, group approached a design problem and the author now has a number of video-films made with design groups who were using the ETHICS method. These can give a new design group some indication of how they will be handling the design task and demonstrate that other groups made up of people very like themselves have successfully accomplished it. The facilitator can also act as a resource person for the design group. When design alternatives are being considered for the new system it is often useful for the group to visit other companies and see how they have organized their work around similar technical applications. The facilitator can arrange these visits. The facilitator can also invite experts to talk to the group about different aspects of the design task—possible organizational and technical options that they might wish to consider, for example.
The facilitator has a very important role in keeping the design group's morale high. Participative design groups seem to go through four distinct psychological stages during a project. At the start there is anxiety and uncertainty. Then, as data collecting proves interesting, enjoyable and relatively easy, the group gains confidence and morale is high. However, about half way through most projects there is a slump in morale. This demoralisation can be caused by a number of different factors. It may be stimulated by the reaction of other groups who become aware of the project after it has started and question its legitimacy. It may be due to the fact that some of the information the design group needs is hard to obtain or there may be relationship problems within the design group or between the design group and the steering group. Once these are overcome and the group moves on into the design phase of the project high morale usually returns and lasts until the end of the project.
Whereas there are no serious difficulties in assisting a design group to acquire the skills necessary to design their own work situation in association with an established technology, difficulties increase when this technology is developing very rapidly. This problem was experienced with a word processing project in a chemical company. Word processing was a transitory technology soon to be replaced by general purpose terminals as part of a move towards the electronic office. It was very difficult to visualise the consequences of the electronic office for the clerks and therefore to design a satisfying work situation .
Today the reality of most change situations is one of negotiation. There is a recognition of different interests, and a more equal distribution of knowledge and power within organizations means that groups which previously would have passively submitted to a major alteration in their work situation may now question and even fight a change which they think is against their interests. Some managements may resent protracted negotiations which they see as due to the new "cussedness" of the labour force, and which they would have preferred to avoid. But other, perhaps more enlightened managements will recognise that successful change requires the identification and resolution of conflicts of interest and may also perceive that such conflicts of interest are not solely between management and subordinates but can occur between members of different grades or functions who perceive a change in relationship as jeopardising their interests versus those of another group. It can therefore be argued that participative methods for introducing change which assist the identification of different interests and the resolution of these through joint discussion and negotiation will assist the production of a form of change which most people in the situation will either welcome or not wish to oppose actively.
The participative approach described in this book is one way of bringing about such an acceptable change process, although it is not free from problems and the very fact of bringing differences and conflicts of interest into the open may produce a series of stressful relationships which management may believe a more traditional approach would have avoided. The counter argument is that the stress will be both temporary and functional as its existence implies that conflicts of interest are out in the open and, given goodwill by all parties, can be resolved.
The arguments for a participative approach therefore run as follows. All change involves some conflicts of interest. To be resolved, these conflicts need to be recognised, brought out into the open, negotiated and a solution arrived at which largely meets the interests of all parties in the situation. Differences of interest will not be confined to management and subordinates but will occur between employees at different hierarchical levels as shown by grading systems, and in different functions. Therefore successful change strategies require institutional mechanisms which enable all these interests to be represented and participation provides these.
Arguments against participation put forward by managements who have not tried it are that it will be slow and expensive. In the author's experience it will require longer elapsed time at the design stage than conventional approaches, but the number of man days required are likely to be the same and implementation will be faster and easier. There is no reason why higher costs should be incurred.
Today, participation is viewed both pragmatically and ideologically by companies which use it, as something that helps efficiency, satisfaction and progress but which is also morally right. Participation can take many forms but at the lower levels of a company it is increasingly concerned with the relationship between individuals and their work environment. It must be stressed, however, that if employees are to be able to exercise some control over this then they need opportunity, confidence and competence and methods for achieving the two last need careful thought. One prerequisite for a participatory community is that it shall be generally desired and accepted and this requires the development of some shared values and objectives concerning participation. Political parties have these in their manifestos, it is important for industry to adopt them also.Top of Page
An attribute of most of today's computer systems is their flexibility in terms of work organization. Unlike some shop floor technology such as the moving assembly line, they do not as a rule force a particular pattern of work activities on those employees who have to interact with them. Providing systems analysts recognise the advantages of flexibility and do not design in technical constraints, most computer systems can accept a variety of task structures.
To help systems designers, managers and other interested groups take advantage of this flexibility and achieve good organizational as well as good technical design, the author and her research colleagues developed the ETHICS method. This method is still being improved through work in this country and the United States, but it is sufficiently tested to be of practical assistance to organizations introducing new systems of work, who wish to try and improve the job satisfaction of their employees.
The ETHICS method consists of a set of logical, sequential analytical steps which should be taken when a new computerbased work system is being designed. At each stage business objectives and human needs are taken into account, so that the system is designed specifically to meet these objectives and needs at one and the same time. The new technology is seen as one means of achieving these goals and one part of the total design task.
One of the main aims of the ETHICS method is to achieve a better balance between technology and people in the design of working systems. The problem in achieving well designed effective systems is not simply that of adjusting people to technology or technology to people. It consists of organizing the two so that the best match can be obtained between them. This means that the potential businessefficiency benefit of the technology can be maximised, while at the same time the working lives of people using the technology are improved. Traditionally. economic and technical objectives have dominated the thinking of those designing new systems. The effects of the system on job design, work relationships and job satisfaction have been viewed as random and outside the system's designers brief, rather than as factors which need to be consciously considered and planned. The specifications for new systems have tended to be almost entirely concerned with technical/efficiency objectives and very rarely with human needs and interests. Today, it is increasingly being recognised that human consequences cannot be left to change or to ad hoc adjustments after implementation. The analysis and specification of the social system, the design of jobs and of the organizational unit as a whole, has now become as important as the specification of the technical system.
For this point to be generally recognised a new design philosophy is required. There is a need to abandon the traditional Tayloristic design philosophy which has regarded man simply as an extension of the machine. This approach has ignored, or failed to see, that new systems can act as catalysts and provide opportunities for an improvement in the quality of working life and in job satisfaction. When these opportunities are neglected, the potential benefits of a system may not be attained, and the human situation can deteriorate.
By giving greater weight to a consideration of social factors, the ETHICS method seeks to achieve greater realisation of the advantages of new systems. One result of designing systems in this way will be to create jobs which are meaningful and fulfilling. At the same time, such systems are likely to achieve a higher level of human efficiency than systems which people feel have been imposed on them, and to which they have little personal commitment.
The ETHICS method includes the following systematic steps:
The design group will normally enlist the help of all its constituents in identifying needs through small group discussions and a request for written comments. This is done using a technique called variance analysis which has as its aim the identification of potential problem areas in a system. This knowledge enables the variance to be avoided altogether or to be easily and effectively controlled and corrected if it does occur. A variance is defined as a tendency for a system or subsystem to deviate from some desired or expected norm or standard.
The notion of variance emerged from the early sociotechnical experiments in Norway. Trist describes how an "action group" of workers, technicians and supervisors diagnosed the reasons for the malfunctioning of the work system with which they were concerned. In this Norwegian study a variance was defined as "key" if it significantly affected l) either the quantity or quality of production or 2) the operating or social costs of production. In the ETHICS method variances are categorised as either "key" or "operational" but they have a somewhat different meaning from the early socio- technical use. In ETHICS key variances are defined as: variances which are systemic (built into the system) and occur because of the essential goals and functions of the user area. For example, the tendency for a sales department to experience communication difficulties with customers or with production units because its principal function "selling" requires it to interact with both of these groups.
Operational variances stem from the organizational inadequacies of the old system and the technical and procedural problems which have inadvertently been built into it. For example, paperwork bottlenecks or lost documents. These kinds of variances should be eliminated through the way the administrative procedures and technology of the new system are designed. Efficiency objectives for a new system are therefore related to gaining more effective control over key variances and to eliminating the old system's operational variances. Variance analysis in its original form is used by socio-technical practitioners in Britain and the United States. The author became familiar with it through working with Professor Lou Davis and his team at UCLA
If we are to make the improvement of job satisfaction a major goal of the systems design process, then it is important to know precisely what is meant by the term. A universally acceptable definition of job satisfaction is extremely hard to find and the author has therefore developed her own.
Job satisfaction is defined in ETHICS as the attainment of a good "fit" between what employees are seeking from their work - their job needs, expectations and aspirations - and what they are required to do in their work - the organizational job requirements which mould their experience.
|The employee's job expectations|
|Job satisfaction = a good "fit" between:|
|Job requirements as defined by the organization|
Examination of the literature on job satisfaction shows that it is split into a number of different schools of thought, each with its own particular focus. First, there is what can be called the psychological needs school. This includes well known psychologists such as Maslow, Herzberg, Likert, etc., who see the development of motivation as the central factor in job satisfaction and concentrate their attention on stimuli which are believed to lead to motivation - for example, the needs of individuals for achievement, recognition, responsibility, status. A second school devotes its attention to leadership as a factor in job satisfaction. Psychologists such as Blake and Mouton see the behaviour of supervision as an important influence on employee attitudes and they therefore direct their observations at leadership style and the response of subordinates to this.
A third school, represented by a number of British experts in wage payment systems, who work, or have worked at the Manchester Business School, approach job satisfaction from a quite different angle and examine the effort- reward bargain as an important variable. This leads to a consideration of how the wages and salaries of particular groups are constructed, and the influence on earnings, and attitudes to these, of factors such as overtime pay and the state of the labour market. Some psychologists maintain that people have a subjective perception of what is a fair day's work. They believe that if this is not obtained then job satisfaction will not be high.
A fourth school of thought consists of behavioural scientists who say that the factors described above are extrinsic to the tasks an employee is required to carry out, and therefore less important in job satisfaction than the work itself and the way it is structured. This group concentrates on the content of work and on job design factors. Some of the most influential exponents in Britain are members of the Tavistock Institute and of the Work Research Unit.
The fifth school of thought approaches job satisfaction from an entirely different angle and sees management ideology and values as an important influence. An American behavioural scientist Alvin Gouldner has described some of the different value systems to be found in companies. For example, he categorises certain forms of management behaviour as "punishment-centred", "representative" or "mock" bureaucracy. Punishment-centred bureaucracy is the type of management behaviour which responds to deviations from rules and procedures with punishments. Representative bureaucracy is the kind of management practice which today would be called "democratic". Here rules and procedures are jointly developed by management and workers to meet a group of shared and mutually agreed objectives. Mock bureaucracy is when an organization has rules and procedures but neither management nor workers identify with these or accept them as legitimate. Consequently they are generally ignored. Although a discussion of values as such does not appear often in the job satisfaction literature, it is clear that the ethics and moral philosophy of a company, together with the kind of legislation that management formulates, and employee perceptions of the legitimacy of this, must have an influence on job satisfaction.
An analytical framework for describing and measuring job satisfaction was developed some years ago by the author's research unit at the Manchester Business School. This is based on the job satisfaction theories set out above together with some ideas derived from an American sociologist, Talcott Parsons. Parsons sees organizations and individuals as constantly making choices when they are presented with different situations. The choices are the following (the comments in brackets are the author's).
A set of categories which brings together Parsons' analytical theory and the ideas of the different schools of thought on job satisfaction is set out in the diagram below. Job satisfaction is again defined as the achievement of a good fit between job needs and expectations and job experience.
|The Employee's Job Needs
||The Employee's Job Experience|
||A good 'fit' exists when the employee:
|The KNOWLEDGE 'fit'
||Wants personal skills and knowledge to be used and developed
||Believes personal skills are being well used and developed
|The PSYCHOLOGICAL 'fit'
||Seeks to further personal interests, e.g. to have a sense of achievement, recognition, responsibility, status.||Believes that personal interests are being successfully catered for.|
|The EFFICIENCY 'fit'
||Seeks an equitable effort-reward bargain, and controls,
including supervisory ones, which are acceptable. Seeks efficient
support services such as information, technical aids, supervisory help.
||Believes that financial rewards are fair and other control systems
are acceptable. Believes that support services are efficient
|The TASK STRUCTURE 'fit'
||Seeks a set of tasks which meets requirements, e.g. which
incorporate variety, interest targets, feedback, task identity
||Has a set of tasks and duties which meet needs for task
|The ETHICAL (social value) 'fit'
||Seeks to work for an employer whose values do not contravene personal values.||Believes that philosophy and values of the employer do not contravence personal values.|
If an employee's needs in these five areas are met then that individual should have high job satisfaction. The "fit" between employee needs and expectations and employee job experience may be good on all five variables or it may be good on some and poor on others. If the "fit" is not entirely good then the question "Why is this?" must be asked and answered by the group responsible for designing the new system. The cause may lie in poor personnel policies or it may lie in an unsatisfactory task structure. If the latter is the case then remedial measures should be possible through better job design.
If there is a bad 'fit' on any of these variables then the "fit" can be improved by the opportunities for change presented by the new computer-based work system. The psychological, efficiency and ethical "fits" can be improved through alterations in personnel policies as well as through organizational design. The knowledge and task/structure "fits" can be improved primarily through a socio-technical approach to work design which creates forms of work organization and job structure related to people's needs.
These then are the areas which are measured before the introduction of the new work system in order to establish if there is a good "fit" between job requirements and employee job needs and expectations. They are measured again once the new system has been implemented and has settled down, to establish if the use of socio-technical design principles, together with any necessary alteration in personnel policies, has improved the aspects of work which were previously unsatisfactory and causing poor job satisfaction.
The knowledge and psychological "fits" are strongly influenced by an individual's personality, education and background. We shall call these sets of needs, needs associated with personality.
There appear to be considerable differences in the extent to which people recognise their own skill and knowledge potential and wish these to be fully utilised, and the design group must diagnose carefully the needs of the change situation for which they are responsible. For example, research into the attitudes of clerks showed that some older clerks were looking for an easy life in work and did not want to be mentally stretched. In contrast, another younger clerical group, bank clerks, who entered employment with good education qualifications, complained that their jobs were too easy and that they were not able to use fully the skills and knowledge which they possessed.
There will be a good "fit" on this variable when employees believe that their skills are being adequately used and that they are being assisted to develop these skills to a higher level.
There will be a poor "fit" if employees believe that their skills and knowledge are being under-utilised, and their opportunities for personal development restricted.
Herzberg has provided a great deal of important research data which shows that if employees are to be motivated by and satisfied with work, then employers must meet their needs for recognition, achievement, responsibility, status, advancement and work interest (the Herzberg motivators). The theory behind this research of Herzberg is that we all have powerful psychological needs, many of which we seek to gratify within the work situation. If the employing organization can ascertain and meet these needs then it will develop motivated and satisfied employees.
The redesign of work which is a result of a new or modified technical and social system can help the achievement of a better "fit" in this area. Job enlargement and enrichment can provide greater opportunities for feelings of achievement, responsibility and status. The redesign of work cannot meet all these psycho- logical needs, however, and attention will also have to be given to personnel policies related to career development and to other aspects of the work situation.
There will be a good "fit" on this variable when the employees believe that personal aspirations for recognition, achievement and other psychological need factors are being adequately met within the work situation.
There will be a poor "fit" if employees have psychological needs related to work which the work situation cannot provide for.
The next two "fits", efficiency and task structure, are related to needs associated with competence, control and efficiency in the work role and the successful performance of work activities.
Important factors in the efficiency "fit" will be adequate support services, a level of earnings considered to be fair and acceptable work controls. Efficiency is assisted by a set of support services which help individuals to work in a well organised way with all the necessary back-up facilities which they require. These will include information, materials, technical aids, specialist knowledge and supervisory help. Employees who do not receive the support services which they regard as essential to the efficient performance of their jobs are likely to become frustrated and dissatisfied.
The organization also seeks employees who will meet its productivity and quality standards and who will accept its administrative procedures and controls. The wage payment system is one element in the control process, for through this a financial agreement is struck with employees that buys their conformity. In return employees require a fair financial reward and for these procedures and controls to be arranged in such a way that they retain a degree of personal influence over their activities.
The effort-reward bargain is the amount that a firm is prepared to pay to get the skill and competence it requires, set against the evaluation of individuals of how much their skills are worth, and their expectations of what they are likely to receive. This contractual area has traditionally been seen. by management as the most important and the one with greatest influence on employer- employee relationships. Yet studies of many white and some blue collar groups have shown that in certain circumstances employees will place financial rewards low down on their list of needs. But if there is dissatisfaction with pay structures and awards, then the introduction of a new work system will provide the opportunity for re-formulating these. Also, because new systems change job requirements it is of the utmost importance for management to ensure that wages and salaries are appropriate for the new task structure.
Firms vary greatly in the controls which they use. Some favour tightly structured rules and procedures which they believe reduce the margin of misunderstanding and error. Others leave their staff wide limits within which to set their own targets and monitor their own performance. McGregor, in his analysis of Theory X and Theory Y management styles suggests that Theory Y, with it emphasis on autonomy and self-control is more acceptable and effective than Theory X, but other evidence suggests that people adjust to the kinds of control which are in operation and it is possible that this is not always an important factor in a job satisfaction.2X The critical factor will be the relevance of selected controls to the needs of a particular work group or work situation. If a group is used to self-imposed, flexible controls and approves of these then it is important that the new work system does not tighten these. If, in contrast, a group has tight external controls and does not like these, then the design group must give some thought to how the work system can be formulated so as to permit a greater degree of self-control by the work groups.
There will be a good "fit" on this variable when employees achieve the support services, rewards and controls which they require.
There will be a bad "fit" if employees believe that they are unfairly paid, undesirably constrained by procedures and controls, or unable to work efficiently because of poor support services.
The task structure "fit" is the organization of work activities in such a way that, from a negative point of view, employees are not required to undertake anything that they regard as too onerous, too demanding, too dull, or too simple; and, from a positive point of view, provides them with a set of challenging tasks and decisions. The "fit" on this variable will be a good one if the level and kind of work provided by the employer meets employees' needs of stimulus and variety.
This element of job satisfaction is strongly influenced by technology, for many of the jobs a firm requires its employees to perform are directly related to the technical processes which it uses. A particular structure of tasks will lead to simple, routine work, another to work which is complex and challenging. The organization, in the past, has sought employees who would adapt unquestioningly to its technology and contingent task structure. There was little understanding that employees might work more effectively in a situation where they were provided with a level of work variety and challenge which met their needs. Technology has for some time been recognised as exerting a powerful effect on behaviour and attitudes at shop floor level but its direct influence on the jobs of white collar staff has been less potent until the advent of the computer.
The nature of the tasks and responsibilities given to an employee is clearly an important element in job satisfaction although one that until recently has only received limited attention. Cooper, a psychologist, who has studied job content and design, suggests that work can be analysed in terms of:
Jobs differ greatly in the blend of these four characteristics that they provide. If a particular job mix does not meet an employee's expectations then there will be a bad "fit" between individual needs and job requirements on this task structure variable and job satisfaction will be reduced.
This is the aspect of job satisfaction most open to improvement through a socio- technical approach to the design of a new work system, and most vulnerable to the technically dominated approach. The human part of the system can be designed so as to produce individual jobs which are high in the work characteristics set out above, and it can be designed to produce a group structure in which autonomous work groups allocate work activities among their members, set their own targets, take their own decisions, etc.
There will be a good "fit" on this variable if technology and task structure produce a work situation wherein employees have the amount of work variety and opportunities for the use of discretion which broadly fit their personal requirements.
There will be a bad "fit" on this variable if employees who require variety and challenge are expected to work on narrow, specialised jobs.
Groups which cannot obtain variety and challenges from their work may obtain this through waging a "war game" with management on the industrial relations front - an expensive and dysfunctional process.
The last "fit" area is concerned with needs related to employee values.
In work, people are evaluated both for their performance and for their behaviour when relating with others. Some organizations place a great deal of importance on performance and rate personal qualities such as sympathy, trust, integrity, much lower. Other organizations value highly employees who are successful as human beings as well as workers. Organizations seek employees who match their ideologies and cultures. In turn employees seek employers who hold similar sets of values to their own.
The "fit" between what employees want and what they receive on the ethical variable is an interesting one about which a considerable amount has been written but little related to the subject of job satisfaction. Yet the author has found it one of the most important factors in job satisfaction and it is likely to be increasingly important as employees demand better communications and more involvement in decision taking.
The ethical/social value "fit" is likely to be as much concerned with the planning and implementation as with the operation of a new work system. It will be important for the employer to meet the employee's needs for communication, consultation and participation in the design of the system. Failure to do this may lead to distrust and apprehension and a feeling on the part of the employees that they are not being treated as they should be.
There will be a good "fit" on this variable when the organization is able to meet those employee values which concern communication, consultation, participation and other aspects of human relationships. The use of the ETHICS method is one way of doing this.
There will be a bad "fit" if employees are not human relations conscious while their employer is.
This job satisfaction framework provides a useful tool which can be used both to measure the overall job satisfaction of a particular group of employees or their satisfaction with different aspects of work. In any systems design process there needs to be a good understanding of why certain "fit" areas are successful; in other words why there is a good match between what employees are seeking from the work situation on this variable, and what they are receiving. Where job satisfaction "fits" are poor then the reason for this must be ascertained and the improvement of this "fit" made a design objective.
In ETHICS the job satisfaction theory described above is used as the basis of a self-completion questionnaire which is given to all employees who will be affected by the new work system. The results are analysed by section, job, grade and age and graphs printed out for each department or section to be affected by the proposed system, showing on which of the five job satisfaction measures there is a good or bad "fit". The results are given to the members of the design group and to all those who completed the questionnaire. The design group then holds small group discussions with the respondents on both efficiency and job satisfaction problems and needs. These small group discussions have two advantages. First, they provide information on the reasons for problems and bad "fits". While questionnaires are useful in identifying problems, they. are not usually as good at providing explanations as to why these problems exist. These explanations are more readily arrived at through discussion. Second, the fact of discussing problems often causes employees to think out possible solutions to these, and this can be a very useful input to the design of the new system. It also gets all potential users involved in the design process.
Job satisfaction questionnaires are based on the framework below which is a summary of the theory discussed earlier. The wording of individual questions is discussed and decided on by the design group who may wish to add additional questions of specific relevance to their own situation. An example of the questionnaire will be found in the ETHICS method in Chapter 5.
Knowledge needs. How, ideally, would each individual or group forming part of the system like their existing skills and knowledge to be used? What opportunities would they like for these to be developed further?
Psychological needs. What are their needs for responsibility, status, esteem, security and advancement and how do they define these needs?
Support/control needs. What kind of support services do different users believe will enable them to carry out their work responsibilities more efficiently? These support services will include the information and materials necessary to work at a high level of competence as well as supervisory support and good working conditions. What kind of control systems do users believe will assist their motivation and efficiency? The level and structure of wages and salaries is an important part of any control system.
Task needs. What kind of task structure do
different groups of users find motivating,
interesting and challenging?
For example, to what extent do users want jobs that include elements of the following - opportunities for self management; for developing new methods, services or products; for co-ordinating their own activities and taking organizational decisions for solving their own problems and monitoring their own progress.
Ethical needs. How do users at every level want to be treated by management? Do the organization's policies on communication, consultation and participation meet employee expectations? Do other kinds of policy also meet these expectations?
If the "fit" between what employees are seeking from the organization which employs them and what they are receiving is found to be a good one before the introduction of the new system, then it is likely that job satisfaction will be already high, and there should be an absence of serious management-employee conflict. This may assist the acceptance of change as there will be fewer attempts by either side to use the introduction of a new work system to further existing conflicts of interest.
However, a situation where there is a very good job satisfaction "fit" may have little internal dynamic for change. The problem for the design group which is introducing a new technical system into this kind of environment will then be to present the change in such a way that recipients are assured that they will gain more job satisfaction from it and improve on an already good "fit". If the "fit" is poor, then providing that the design group is aware of the reasons for this they can design the new work system so as to achieve a "human" improvement.
The efficiency and job satisfaction diagnoses incorporate forecasts of future as well as present needs and this provides some guidance on how responsive the new system must be to future change. An analysis of likely and possible changes in the organization's internal and external environment is made and used to identify those parts of the new system that must be kept flexible so that they can adapt easily.
The next, and most important, step is to set specific efficiency and job satisfaction (social) objectives for the new system. Alternative design strategies can then be measured against these and the strategy that best fits both efficiency and job satisfaction objectives is the one chosen for implementation. The objective setting process is complicated and usually involves negotiation between the different members of the design group.
The analysis of variances, the job satisfaction questionnaire, and the small group discussions associated with both of these, will have identified the problems and needs of different departments sections, roles, grades and age groups. The design group should represent all these different interests. If it does not then representatives of the missing groups should now be co-opted either as permanent members, or to attend those meetings at which their particular interests will be discussed. Design group members should list the efficiency and job satisfaction objectives of most importance to the group whose interests they represent. They should also allocate priorities to these objectives. When all objectives are set out, those which are of high priority and common to all or most groups should be noted. These are the key objectives to be achieved. Conflicts of objectives between different interest groups should also be identified and attempts made to resolve the conflict before alternative design strategies are put forward. Ideally, design solutions should be clever enough to enable all groups to achieve their most important efficiency and job satisfaction objectives. These are sometimes called integrated solutions. If an integrated solution is not possible then trade-offs between groups will have to take place and some objectives will have to be modified or abandoned.
It is most important that the design of the new system shall not begin until the design group has agreed a list of unambiguous, specific efficiency and job satisfaction objectives. It can greatly assist subsequent evaluation of system success if, at this stage, a note can be made of how the successful attainment of an objective can be measured or checked. Set out below are some of the key objectives associated with the design of a personnel information system.
To ensure information from departmental line management is legible, complete and punctual.
To improve procedures for new starters, transfers and leavers.
To create an agreed timetable for all input to the system.
To provide a comprehensive employee file for:
To provide a means of accessing, analysing, formating and displaying information from the employee file in response to both regular and ad hoc requests.
Job satisfaction objectives for personnel clerks
To increase job satisfaction generally.
To reduce routine tasks and provide more work variety.
To improve work co-ordination. This should increase job satisfaction, provide more interesting work and lead to a better social atmosphere.
To create a base of data capable of being extended to answer new questions as and when they arise, in this way reducing frustration and improving morale.
To improve the work environment.
The next step is to identify the technical, administrative and organizational options that can best achieve the efficiency and job satisfaction objectives and to choose the best mix for implementation. The socio-technical methodology has already been referred to in this chapter but it is appropriate here to give more detail on its origin and development.
From the beginning of this century until the present day the design of work has been strongly influenced by what is known as the Tayloristic philosophy of scientific management. Frederick Winslow Taylor was an engineer who attempted to combat the inefficiencies of early manufacturing industry by applying work study principles to the performance of labouring tasks. What Taylor did was to split each job into its component operations and take the time of each with a stop watch. This led to the systematic analysis and breakdown of work into the smallest components and the rearrangement of these elements into the most efficient combination. Taylor's fame reached its height in 1899 when he taught a Dutchman named Schmidt to shovel forty-seven instead of twelve-and-a-half tons of pig iron. Every detail of the man's job was specified. Taylor recognised that a man who could do this tightly regimented work needed certain mental characteristics. He wrote, "one of the very first requirements for a man who is fit to handle pig iron as a regular occupation is that he shall be so stupid and so phlegmatic that he more nearly resembles an ox than any other type". The unfortunate Schmidt did not appear to quite fit this description for we are told that while working for the Bethlehem Steel Company he was also building his own house.
Taylor's principles reached their peak in 1914 when Henry Ford established a car assembly line at Highlands Park, Michigan. The principle of machine-paced work which this embodied led to the creation of jobs which were routine, often stressful and with the speed of work controlled by the machine.
One of the most powerful counter forces to the scientific management approach has been provided by the socio-technical school of systems design. This has concentrated on the design of the human work system that must fit with any technical system. Its principal objective has been to make work more satisfying for the person doing it while at the same time enabling this person to contribute to a high level of technical efficiency.
Eric Trist, one of the pioneers of socio- technical design, has described its origins. The socio-technical concept developed from a project in the British coal industry in 1949. The research came about because the National Coal Board was disturbed at the low morale and low productivity in British coal mines. During this project the research team came across a colliery in South Yorkshire which was using a different form of work organization from other collieries. Instead of having the miners organized in shift groups, with each shift responsible for a different task and no communication or shared responsibility between groups, they found that the men had introduced a work system similar to that found in the pre-mechanized days of mining. The men were organized into small groups, with each group taking responsibility for the entire work cycle of a part of the face, and working autonomously with little supervision. These observations and the implementation of this form of work organization in another colliery led to a set of work design principles very different from the Tayloristic philosophy of small tasks, tight controls and little responsibility.
These principles included the following:
These principles are still being used by practitioners of socio-technical design and there is considerable evidence that they however, have two limitations. First, they have been used to improve the social system associated with a given technology and have not contributed to the design of this technology. Second, they have generally been used at the shop-floor level of an industry and not in the office or for the design of higher levels of organization.
The ETHICS method uses many of the socio-technical principles for the design of jobs, for example, the notions of unit operation and autonomous group, but also considers the technical design options that are available and evaluates the efficiency and human advantages and disadvantages of each of these as part of the design task. Alternative technical and social solutions are first considered separately and later merged.
Experienced design groups may prefer not to separate technical and human alternatives in this way but to think through a series of alternative socio-technical solutions. This approach is quite acceptable but in our experience, more difficult unless the design group has a very good knowledge of the technology they are concerned with and of work design.
|Set social objectives
(improved job satisfaction and quality of working life)
|Set efficiency objectives
(improved business efficiency)
|Specify social alternatives
(work organisation, job design)
|Specify technical and administrative alternatives
(hardware, software, work procedures, information flow
|Match as socio-technical alternatives
|Rank in terms of ability of each alternative
to meet social and efficiency objectives
|Select best socio-technical solution
If the technical system is already installed or specified so that it cannot be altered in any way, then the ETHICS method can still be used, although the consideration of technical alternatives will not form a part of the design procedures. Instead analysis will focus on identifying the impact of the technical system on the jobs of people who interact with it and will set out its social disadvantages and advantages. Technical constraints or imperatives that reduce the number of social alternatives available for consideration will be noted, together with any technical features that can be exploited to design a work structure which provides job satisfaction.
In order to specify human alternatives a design group needs to have a good knowledge of the different ways in which work can be organized. Many of these are concerned with improving the employee's "task variety" and this is true of job enlargement and job rotation. Here variety is introduced into the job by giving an individual more tasks to carry out or by allowing movement around a number of tasks, with a period of time being spent on each. None of these tasks may require much skill for their performance. This kind of approach can be useful in that it reduces work monotony and will be appropriate if a particular employee group is generally satisfied with work and merely wants more variety in order to have a higher level of job satisfaction. It is unlikely to be adequate for groups whose job satisfaction needs are more complex and related to a desire for opportunities to use skill, meet challenge, and exercise control.
An approach that is popular at the present time is job enrichment. This also tends to focus on giving work greater variety, although the approach is more sophisticated. Work is now designed so that the employee is able to use a number of different skills, some of which are quite complex and require judgement to be exercised, choices made and decisions taken. Many offices where clerks deal directly with customers are able to offer this kind of work. The author has worked in a sales and distribution department where, when a new computer system was introduced, the work was redesigned so that all clerks had the same set of responsibilities. These were carrying out a number of routine activities associated with inputting order information into a computer via visual display terminals but also undertaking more complex activities associated with dealing with customers and handling customer queries; a set of activities which were intricate and demanding and required the use of excellent judgement. In this department the ability to use judgement and take decisions was a product of knowledge, not job grade, so that even the newest clerk could start handling difficult customer problems as soon as he or she felt that they had the knowledge to do so. The clerks in this department also had additional responsibilities in that they were able to requisition their own materials and personally obtain any customer information they required through interrogating the computer via the terminals. They were able to handle their own problems and organize their own work, and supervisory assistance was only requested when the problem was very difficult indeed.
In this situation work was enriched by putting together two sets of tasks which had previously been handled by different groups and by giving the clerks responsibility for activities, such as material requisition and the obtaining of information, which previously had been carried out by supervision. The role of supervision was now primarily that of long-term planning and co-ordination with other departments in the firm. This kind of job enrichment seems to improve both job satisfaction and efficiency in many situations, providing that employees can be trained to the necessary level of competence. It may be difficult to introduce into departments where there are few experienced employees and where there is a very high level of labour turnover. Also, because this approach does not give employees any responsibility for the development of new methods, it does not lead to any new thinking on better ways of carrying out the work.
It is not easy to incorporate development aspects of work into non-managerial jobs, although intelligent staff are likely to find that an opportunity to develop new ideas and try out new methods is one of the most satisfying aspects of their work. If jobs have this component it is also of great advantage to management as work methods will be constantly reviewed and improved and suggestions will be coming from staff for innovations that assist the prosperity of the business. A way of incorporating development activities into work becomes possible if we stop thinking about individual jobs and turn our attention to the socio-technical concept of "autonomous" groups or, as the author prefers to call them, self-managing groups. The focus of attention is now the group rather than the individual and this switch provides an opportunity to consider how more complex responsibilities can be incorporated into the work of a group of non-supervisory staff who see themselves as a team.
If we concentrate our attention on the kind of self-managing group that is multi-skilled in the sense that each member is competent to carry out all the operational activities for which the group is responsible, then many things become possible. The group is now co-ordinated and organizes its own work so that individual task responsibilities integrate well together and the group works as an efficient team . Such a group is more easily able to initiate and try out new ideas and methods than an individual employee. Similarly, if management has confidence in the ability of its self- managing groups then it can hand over a great many control activities. It can, for example, let the group organize its own work activities and set its own performance targets and monitor these. It can give it responsibility for identifying and correcting its own mistakes. It can give it a budget and allow it to buy its own materials and even organize the selling of its own products to customers. This kind of group may require little supervisory intervention in its activities and management's responsibility will become one of long-term planning and boundary management. By boundary management is meant ensuring that the work of all the selfmanaging groups in a department is co-ordinated and the work of the department as a whole integrates well with that of other contingent departments. The self-managing group can be excellent in the right situation and it provides a stimulating work environment in which staff can readily develop their talents. However, for it to succeed certain things are necessary. First of all the work of the department into which the self-managing groups are introduced must provide scope for multi-skilled work that provides challenge and responsibility. In many situations the work has been so strictly allocated between departments that no rearrangement of tasks or creation of self- managing groups can make it much more interesting. In this kind of situation the challenging, problem solving aspect of the work has been separated off and handed over to a specialist group in a separate department. Any real improvement in work interest can now only be achieved if several departments are merged together, thus providing the required work variety. Secondly, the creation of multiskilled, self-managing groups with the competence to control many of their activities requires intelligent, responsible employees together with excellent long-term training. It may take several years to make all members of a group multiskilled and if the group suffers from a high labour turnover then management will find the training process an expensive one. Thirdly, the creation of self-managing groups has implications for salary levels and grading schemes. Grading can no longer be related to length of service, it must be related to knowledge and skill. Therefore a new employee who is adept at learning all the jobs for which a group is responsible will reach the highest grade quickly and this may be resented by long serving members of the firm who have worked their way slowly up a hierarchy of grades over many years. But this is a problem associated with a change from one philosophy of work to another and it should not cause permanent difficulty.
The design group's task is not ended once the new system is designed. There is then need for a carefully thought out implementation strategy and when the system is being implemented it is essential to monitor closely what is happening. Although the ETHICS diagnostic and design tools can provide useful starting points for systems change, the implementation process requires careful planning and monitoring to ensure that design strategies are staying in line with job satisfaction and efficiency objectives and that objectives set an early stage of the design process continue to be valid. If the change process deviates from the intended course of giving considerable weight to human needs then mechanisms must be available to bring it back on course.
The success of a new work system in either human or efficiency terms is rarely evaluated after implementation is completed. Yet this is essential if planning and systems design are to incorporate a learning process in which the mistakes of the past are avoided in the future.
The diagnostic tools in ETHICS can be used once again to evaluate how well the new system is achieving the objectives set for it. The design group will examine first the extent to which "key" variances—those problems which are endemic to the objectives and tasks of the department or function - are being better controlled. "Better" being defined as more easily, quickly, cheaply. The evaluation of variances will also cover the extent to which pre-change operational variances—those due to work organization or technology—have been eliminated without new ones being introduced. The next step is to examine the nature of the "fit" between employee needs and organizational job requirements now that the new system is in. If the post-change "fit" is better than the pre- change "fit", then the new work structure has led to an improvement in job satisfaction. If the "fit" is still unsatisfactory then some remedial action needs to be taken. Finally, it is necessary to check how responsive the new system is proving in its ability to adapt to change.. Has change been introduced without expensive and time consuming large scale redesign?
A final point that must be made is that the design process is continuous. Major effort and thought is required when a new system is being introduced; after this stage is completed monitoring and evaluation lead to system adjustments which continue until the time arrives for a new system and a new innovation effort.Top of Page
Ethics is a problem solving methodology that has been developed to assist the introduction of organizational systems incorporating new technology. It has as its principal objective the successful integration of company objectives with the needs of employees and customers. In the author's view it is an ethical approach to systems design because it assists designers to maximise human gains while achieving business and technical excellence.
ETHICS has been built on the proposition that good systems design is far more than good technical design. It requires a design approach that covers technology and the organizational context in which the technology is placed. This implies the total design of departments, functions or areas using new technology including roles, relationships, activities and jobs. This is the design area.
Involvement requires the creation of participative groups, and decisions on the amount and nature of their contribution to the total design process must therefore be made.
The Steering Committee will set the guide- lines for the Design Group. It will have as its members senior managers from affected user areas, senior managers from Management Services and Personnel and in unionised companies, senior trade union officials. It should meet with the Design Group once a month throughout the project.
The Design Group will design the new system, including in its brief the choice of hardware and software, man-machine interaction, the overall re-organization of the design area and the allocation of responsibilities and tasks to groups and individuals.
The Design Group will consist of representatives of all major interests in the design area. There should be a representative from each major section and function, each grade, age group and sex. It will also include the professional systems analysts engaged on the project.
Design Groups should not have more than eight to ten members. All, except the systems analysts, should be selected or elected in a way that is acceptable to their constituents and seen as democratic.
Design Groups should meet for a whole or half a day once a week or once a fortnight. A reasonably straightforward system will require around twelve to fourteen meetings for systems design. This is an elapsed time of four to six months.
The following questions will need to be answered before the project begins:
The systems analysts chosen to work as members of the Design Group should be willing to assume a role very different from the traditional one. Instead of being "designers" they will be teachers, advisers and learners. An important advantage of participation is that users and technical specialists learn from each other.
User groups new to the design task will need help. They have to be assisted in their use of ETHICS, motivated to work enthusiastically and helped to resolve any interpersonal or inter- group conflicts that arise. An extrnal or internal consultant who is in a neutral position in relation to the new system and who has training and human relations skills is the best choice for the facilitator role.
Before the project begins the manager of the user department in the design area must decide how he or she is going to relate to the Design Group. If the manager is not a member of the Steering Committee, because the senior manager is there, then it is essential that close contact is maintained with the Design Group. At the same time, as a believer in participation, the manager will wish to avoid unduly influencing the ideas of Design Group members.
A solution to this problem that works well in practice is for the departmental manager to leave the Design Group on its own for its first three of four meetings. This enables group members to gain confidence and knowledge.
Once the Group feels confident and has an understanding of its task the manager may either join in for part of each meeting or at regular intervals.
Systems analysts, facilitators and departmental managers should not chair the Design Group. The Chairman should be selected from one of the other members, preferably after the Group has worked together for some meetings.
ETHICS has four important objectives:
New technology and user participation are means to assist the achievement of these objectives.
Once the Steering Committee and the Design Group exist they should be familiarised with the ETHICS approach. This is best done through a two day training course attended by all the members of the two groups plus any other interested individuals from management, management services, the user area or the trade unions.
The first meeting therefore considers the question "Why do we need to change?" This will enable the Design Group to discuss in broad terms existing problems, future demands, the opportunities for improvement that new organization plus a new technology could bring. Documentation and recording of meetings As the discussion takes place points made by individual members of the Design Group are written on flip charts.
It can be helpful to video each meeting of the Design Group. This enables a visual history of the design process to be built-up and this can be turned into a training film at a later stage.
After each meeting minutes of the discussion and proceedings are written up and circulated to the Steering Committee and all staff in the user area. If the Facilitator writes the minutes then these should be approved by members of the Design Group before circulation.
The Design Group must be absolutely clear where its design responsibilities begin and end and so it must identify the boundary around the system which it is designing:
Groups with a direct relationship to the new system, who will use it as part of their daily activities, will fall within the boundary.
Groups with an indirect relationship, who will use it infrequently for only part of their work, will be outside.
The Design Group has a responsibility to design in detail the system within the boundary which it has identified and to ensure that the system links easily and effectively into functions which adjoin it and into the external environment. The design of the boundary interface is likely to be a difficult task. After system boundary location has been discussed the Design Group should meet with the Steering Committee to obtain its view on where the boundary should be placed and to discuss any guide lines which have been formulated for the project.
Setting system boundaries is not easy and any created at the start of a project may have to be revised at a later stage when design opportunities become clearer.
Similarly guide-lines produced by a Steering Committee at its first meeting may have to be later revised if they turn out to be too broad, too restrictive or otherwise inappropriate.
If the system boundary embraces a number of departments then ETHICS will be used first of all to make a broad overall design for the total system. A second stage will be the detailed design of the various sub-systems. This detailed design is best carried out by lower level design groups representing the users in the sub-systems.
Before starting to design a new system it is essential that all the members of the Design Group understand how the present system works. Many will have only a partial view, knowing the detail of their own jobs and those of their immediate neighbours but not having a comprehensive knowledge of the working of the whole department.
Step three requires the Design Group to make two kinds of description. A horizontal input/output analysis and a vertical analysis showing different levels of work complexity and importance. The latter is based on a cybernetic model of a viable system developed by Professor Stafford Beer. These two forms of analysis are shown in the diagrams below.
In addition to ensuring that the Design Group understands the existing system before replacing it with a new one, this descriptive activity helps the group to begin working together as a team. It is relatively easy to carry out, the Design Group members are all experts in the requirements of their jobs and there is interest in what each member does, and why and how.
At this early stage in the system design process the Facilitator must emphasise to the Design Group members that their task is not merely to polish up the existing system. They are required to produce a new system based on a careful analysis of company, department and individual needs and problems. Their new system may resemble the present system in certain aspects, but it may not.
The vertical description of the department requires an analysis of activities at five different levels.
The first and lowest level consists of operating activities. These are the day-to-day or regular tasks that enable the principal functions of the department to be carried out. These should have been described in the input/output analysis.
The second and more complex level consists of activities directed at preventing work problems occurring and correcting these when they do occur. These are called problem prevention/solution activities.
The third level consists of activities and tasks that have to be co-ordinated within the department and between the department and other departments and sections. These are co- ordination activities.
The fourth level is the important, but often neglected, aspect of development. What activities, products, services etc. handled by the department need to be developed and improved. These are development activities.
Lastly, comes the fifth and most complex level of control. How is the total department controlled so that it works efficiently, meets its targets and achieves its objectives. These are control activities.
The first question is why do these exist? What is their primary role and purpose?The second question is given this role and purpose what should be their responsibilities and functions? The third question is, how far do their present activities match what they should be doing?
The answers to the first and second questions will provide the Design Group with the information to make a list of KEY OBJECTIVES.
These are the important objectives which the design areas should be striving to achieve. From now on the design task will be directed at creating a new system which will achieve these key objectives.
Often these key objectives can be readily identified but occasionally this is not a simple process and attention has to be paid to the reasons a unit was set up, how its mission was defined at that time and how applicable this definition is to its mission in the future.
Once key objectives have been specified then it is not difficult to ask and answer the following question.
If these are the key objectives then what are the key tasks which must be carried out if the objectives are to be achieved?
Some of these key tasks will already be carried out effectively; some may be badly done, and some not done at all.
At this stage of the design process it is important not to go into too much detail. This will come later.
As a principal contribution of a new computer-based work system will be more accurate and timely information, the key information requirements associated with the key tasks should now be specified.
The five level model used to describe work in Step 3 can be used again here.
Efficiency needs can be identified by looking for variances. A variance is a tendency for a system or part of a system to deviate from some expected or desired standard or norm. In other words it is a weak link, a part of the system where problems tend to occur. Variances can be of two kinds. Key or systemic variances and operating variances.
Key variances are potential problem areas which cannot be eliminated although they may be effectively controlled. They are built into a system and arise from the key objectives and key tasks which the system has been designed to meet. They frequently arise at the interface between the system and other systems. For example, a production department will always have potential problems in its relationship with the Sales department. The same will be true of Sales and its customers.
Operating variances are not so deeply embedded in the system and are weak links inadvertently designed in by earlier systems designers through the way they have organized procedures, machines and activities. When a new system is designed many of these variances can be eliminated altogether.
All staff are usually aware of key variances but operating variances may only be known to those groups and individuals which have to cope with them in their work.
The diagnosis and listing of key and operating variances can be done by all the staff in the area within the system boundary. This step provides the Design Group with an opportunity for actively involving its constituents in the design process.
Staff are asked to describe the variances which they encounter in their own jobs; to say how and why these occur; the action which is taken to correct them, and how they could be prevented altogether or more easily and quickly corrected.
Although this information is best written down and handed to the individual member of the Design Group who represents a particular group of staff, this identification of variances provides an opportunity for Design Group members to meet with their constituents in small groups. Informal discussions can then take place on the nature of variances, how they manifest themselves and why they occur. Such meetings do not only generate useful information, they also provide an opportunity to get all staff interested and involved in the design of the new system.
At this point the Facilitator often has to warn a design group that it must not rush into design. There is a great deal of diagnostic information to be collected first.
The ETHICS method gives efficiency and job satisfaction equal weight. The argument here is that the two reinforce each other. An efficient, well run department reduces frustration and increases job satisfaction. Similarly high job satisfaction generally requires feelings of competence, responsibility and pride in work. These are associated with an ability to work in a well organised environment. If job satisfaction is to be a specific system objective then it needs to be defined and measured with problem areas clearly identified. In chapter three the definition and framework used in ETHICS were described. Job satisfaction is the FIT between an individual or group's job needs and expectations and the requirements of the job which they presently occupy. The ETHICS job satisfaction framework covers three broad areas. Needs associated with personality, including knowledge needs and psychological needs. Needs associated with competence and efficiency in the work role, including efficiency factors such as support services and systems of work control, and task needs — the kinds of task structure that different users find motivating, interesting and challenging. Lastly, needs associated with employee values — in particular ethical needs or the manner in which employees at every level wish to be treated by management and to relate to each other.
The ETHICS questionnaire is based on this framework although the Design Group is encouraged to alter it in any way it feels appropriate. In particular, by adding questions of relevance to its own company and department.
The job satisfaction diagnosis provides a second opportunity for involving all potential users of the new system in its design. The questionnaire is handed to all individuals working within the system boundary. They are asked not to put their names on it, but to indicate some or all of the following: age, sex, grade, job and section, as the diagnosis of good and bad FITS will be made in terms of these characteristics. (Participation in the survey is, of course, entirely voluntary.)
They are also told that they will receive a copy of the results and be given an opportunity to discuss these with the Design Group.
In the interests of confidentiality the Facilitator should hand out and collect the questionnaires and should be the only person connected with the Design Group and the user area to see these. This means that the Facilitator must assume responsibility for questionnaire analysis.
After the questionnaires are analysed the Design Group is given a very comprehensive breakdown of the results and each individual who completed a questionnaire is given a less detailed breakdown. The Steering Committee also receives a copy of the results.
Questionnaires are a good means for identifying problems but they are less effective at explaining the reasons for the problems. It is in this respect that the small group discussions are important.
This is a study of what people like and dislike about their jobs. Its aim is to obtain information which will enable jobs to be designed in such a way that people really enjoy doing them.
There are no right or wrong answers. The best answer is your personal opinion.
Each questionnaire is completely anonymous, the only information you are asked to give is your section and grade. This helps the researcher to find out if particular jobs have special problems and if people of different grades have different points of view.
To begin, please fill in the following:Section .....................
These job satisfaction questionnaire results are discussed with users in small groups before the Design Group considers ways of improving job satisfaction.
Each member of the Design Group will meet with his or her constituents to discuss the questionnaire results. They will be asked to check the accuracy of the data, to discuss the reasons for job satisfaction problems and to make suggestions as to how these might be avoided or more easily solved.
These discussions again give all users an opportunity for contributing to the design of the new system and influencing the thinking of their representatives on the Design Group. The future users of the system have now received the Design Group's minutes after each meeting. They have had discussions with their representatives, in small groups, on efficiency problems identified by the variance analysis and on job satisfaction needs and problems identified by the job satisfaction analysis.
The Design Group will also have been having regular monthly meetings with the Steering Committee to ensure that the higher level group is aware of what is taking place and is feeding its views into the discussions.
The Design Group now has to make use of this job satisfaction data to influence its thinking about the new system. It does this by completing the Analysis of Social Needs form shown below.
The purpose of this form is to start the Design Group thinking about how job satisfaction can be improved when the new system is designed. A form should be completed for the user population as a whole but also for each major subgroup. These will be employees in different jobs, sections and grades.
The form helps the Design Group to note those aspects of job satisfaction which are satisfactory and those which are unsatisfactory. Job aspects where there is a very good fit between employee job needs and job requirements should be recorded for incorporation into the new system, if this is possible. Similarly, job aspects where there is a bad fit should be recorded as areas that must be improved when the new system is designed.
There is a column for preliminary suggestions on how this improvement could be achieved, although it must be stressed that the Design Group has considerably more work to complete before it begins designing the new system in earnest.
Most new work systems today have limited lives because of major changes in their environment. These changes can be a result of many factors -changes in legislation, in markets, in labour availability and in technology.
A new system must therefore be able to adapt to change and must be designed to do so. This means that it must have enough built-in flexibility to cope with future change.
A broad assessment of the required amount and focus of system flexibility is derived from an identification and analysis of future changes likely to affect the system within the next five years.
A Design Group may need to talk with internal and external experts in order to assess the nature and impact of such change.Changes likely to affect a company are:
The agreement of objectives can be straightforward but it can also be a demanding process in which the interests of individual members of the Design Group and their constituents have to be reconciled with the interests of the Design Group as a whole and with those of other groups in the company. The Facilitator may have to help the Group resolve these internal and external conflicts of interest and this is not always an easy task.
It is essential, at this stage of the design task, for the Design Group to be aware of the wishes and priorities of all groups who will be affected, either directly or indirectly, by the new system. Any who do not have representatives as members of the Design Group must now be asked to join the group for one or two meetings so that their views can be heard.
External groups such as customers or suppliers must not be forgotten and if it is impossible to invite a genuine customer or supplier then an existing member of the Design Group should be asked to think and speak for them.
The procedure now is for all the minutes and records of earlier meetings to be carefully studied and for the Design Group to make separate lists, on flip charts, of the efficiency, job satisfaction and future change problems and needs which they have already identified. These should be stated positively as objectives to be striven for through the way the new system is designed.
Each member of the Design Group should rank these objectives from 1 to 5 to show their importance to his or her constituents (5 indicates highest importance).
The Design Group now identifies and lists those objectives which are most important to all staff and to different interest groups.
The Design Group next needs to examine these objectives closely and note if any are in conflict, so that the achievement of one would inhibit the achievement of the other. If this proves to be the case an attempt must be made to resolve the difficulty by replacing that objective with another, less divisive one.
Sometimes conflict of this kind cannot be resolved and the Design Group has to live with it. But ETHICS has the advantage of bringing the conflict out into the open where it can be discussed and a solution sought.
The Design Group should now make a list of priority objectives which all members are able to agree with and work to achieve through the design of the new system. These should be ranked 1, 2 or 3 to show their relative importance with I this time indicating the most important.
This does not mean that other objectives should be forgotten or ignored. The aim of the design task is to create a work system which achieves all the priority objectives and as many of the others as possible.These objectives are broad and their achievement is difficult to measure. The next stage is to translate as many as possible into specific, measurable, targets.
It must again be emphasised that the Design Group is not cleaning up the existing system but making fundamental improvements. It would now be wise to check the Group's work on steps 4 and 5 to ensure that measures relate to key objectives and key tasks.
A problem at the systems evaluation stage is that measurement of improvement is often subjective because the firm has not kept detailed records of the pre-change situation. It does not know how slow and inaccurate information was before the new system was designed and introduced.
ETHICS incorporates two measures of improvement that work well in practice. The first is derived from a second analysis of variances at the evaluation stage of the project. The other is from a second job satisfaction survey, also at that stage. These will be discussed later.
It is extremely important for the Design Group to check its selected objectives with other groups before proceeding any further. Meetings should be held with the Steering Committee, with the Design Group members' constituents, and with staff indirectly affected by the new system. All interest groups will then be able to discuss the objectives which the new system is going to strive to achieve and register their approval or objections to these. In this way rumour and misconceptions are avoided. Employees are able to understand and appreciate the efforts which are being made on their behalf to develop a system which will both improve their efficiency and their job satisfaction.
Readers will now have noticed that ETHICS closely resembles the Japanese approach to problem solving. It spends considerable time on a careful, systematic diagnosis of problems and needs before moving to solutions. It also involves employees at every level in the problem and need identification task.
Ideally, organizational and technical design should take place simultaneously and interactively. Organizational options, by which is meant different ways of organizing the department so as to achieve the efficiency and job satisfaction objectives, should affect the technical options that are considered, and vice versa.
Again, it must be stressed that technology is always a "means" to greater efficiency and higher job satisfaction, and never an end in itself.We have stressed previously that ETHICS provides a learning experience for users and systems analysts, particularly those who are members of the Design Group. So far they have been acquiring diagnostic skills but they now commence to learn design skills. They also begin to learn about the various technical options that are available to meet their needs and help solve their problems.
A great deal of this information will come from the systems analysts who are in the Design Group but a useful, additional, aid to knowledge at this time is for the Design Group to visit other companies who have introduced systems similar to the one which they are designing. Invitations to technical experts from inside or outside the company to come and talk to the Group provides another valuable source of information.
Before a number of organizational options can be identified the Design Group must re-examine its conclusions for steps 4 and 5. These answered the questions (l) What should be the key objectives of the area within the system boundary, and how do these differ from what is actually happening? (2) Given these key objectives what should be the key tasks carried out within this area? (Do not confuse these key objectives with the objectives for the new system in step 10.)
The key tasks associated with the key objectives now need spelling out in greater detail as these will form the work activities of the design area.
The vertical analysis set out in step 3 can be used again here.
The Design Group now documents its answers to the following questions:
These categories 1-7 provide the basic data for the organizational design of the system. Organizational options will be created by arranging these seven kinds of activities in different ways.
Each organizational option should specify the following:
Each organizational option should contribute to the achievement of efficiency and job satisfaction objectives set out in step 10.
It is advisable for a Design Group to develop between three and six organizational options. Although the analysis of activities within each option should be comprehensive, it should not be done in too great detail until step 13.
Each organizational option is now checked against the efficiency, job satisfaction and future change objectives listed in step 10 and their advantages and disadvantages in terms of ability to meet these objectives are noted.
The Design Group must now establish the following. Does the option achieve the priority efficiency and job satisfaction objectives listed on page 89 in step 10. Does it also achieve all or some of the other less important efficiency and job satisfaction objectives listed on pages 86 and 87. These relate to the needs of particular groups.
The identification of organizational and technical options should be done in parallel, not sequentially, and a range of technical options must be specified.
Technical options will include hardware, software and the design of the man-machine interface.
Whenever possible experimental examples of different options should be set up in the design area so that the Design Group members and their constituents can try these out for themselves. Technical options should be evaluated against efficiency, job satisfaction and future change objectives in the same way as organizational options. Examples from the Purchase Invoice Department are set out on the next page.
It is good practice to have a change in organizational structure but no change in technology as one option. Sometimes objectives can be achieved without investment in expensive hardware.
Once the organizational and technical options have been evaluated in terms of their ability to meet priority objectives and the objectives of particular interest groups, a short list of those organizational and technical options that best meet objectives can be drawn up.
This list should consist of two or three options in each category.
It is now necessary for the Design Group to merge the organizational and technical options which have been short listed. This requires a check of which organizational options fit best with which technical options and vice versa. It is essential that organizational and technical options are completely compatible when merged together, otherwise the new system will not function at optimum smoothness.
These merged organizational/technical options should once again be evaluated against the priority objectives and the objectives of particular interest groups. The one which best fits with these objectives is selected for detailed design and implementation.
Before this final design decision is taken it is essential for the Design Group to meet with the Steering Committee and for members to consult with their constituents.
Both Steering Group and constituents are likely to have an important point of view.
The Steering Group may be aware of company constraints which make a particular option difficult to introduce; the constituents may recognise that one option would prove less acceptable to a particular interest group than others.
The final choice of system will therefore reflect the careful diagnosis, objective setting and evaluation of organizational and technical options carried out by the Design Group; the views of the Steering Committee given from a broad company perspective, and the views of the users which are narrower and more influenced by sectional interests.
The selected system now has to be designed out in detail and this requires a knowledge of simple flow charting. The flow of information and documents through the design area must be recorded. Responsibilities must be allocated and tasks distributed to groups and individuals so as to provide an efficient work environment and the kinds of jobs which the job satisfaction survey has shown that staff are seeking.
Many of these tasks will be new and some, particularly those related to new technology, will require a period of training.
Particular attention must be paid to the creation of effective relationships and procedures across the boundary between the design area and adjoining departments with which it interacts.
The Design Group of the Purchase Invoice system which has been used as an example so far chose a mixture of autonomous groups, a specialist group and a group responsible for routine work to achieve its efficiency and job satisfaction objectives. The next two diagrams show how proposed work organisation was checked against objectives.
Detailed work design for Purchase Invoice was based on two principles:
1. The social importance of the work group, which ideally should have between five and eight members.
2. The importance to a group or individual of having responsibility for a logically integrated, meaningful and interesting set of tasks.
The different tasks associated with the new work system were therefore divided up into a number of unit operations. A unit operation being defined as an integrated set of tasks separated from other sets of tasks by some kind of discontinuity.
The mix of tasks within each unit operation was carefully selected to meet the job satisfaction needs of the different kinds of staff who would be working there. There would be simple, routine tasks, but there would also be complex, problem solving tasks. A work group could be given responsibility for more than one unit operation, but unit operations would not be split between groups.
This kind of group structure based on carefully thought out unit operations has a number of advantages.
The mix of tasks within each unit operation should be checked to establish that the following good job design principles apply.
Two other important points for a Design Group to note are:
1. Any new organizational structure should be kept as flexible as possible so that it can be easily changed. There are many ways of combining tasks together and the last example showed only one of these.
2. The notion of the job and of job ownership (this is my job) should not be encouraged. Staff should be expected to constantly review the set of tasks for which they are responsible and request, or be requested, that this be altered to meet new needs and new machines.
This kind of review can provide a valuable continuing role for a Design Group.
The Design Group's work is by no means over once the new organizational structure is created and work activities are allocated. It must also ensure that a high quality physical environment is created for those staff working within the design area.
In doing this it will almost certainly need to have expert advice, particularly if the new system involves the introduction of much new machinery.
The modern semi-automated office requires great thought to be given to temperature, lighting, seating, ducting for cables and the allocation of individual work space.
The best designed system is not going to achieve its objectives unless it is successfully implemented, and many good systems encounter problems at the implementation stage.
Just as a system needs to be systematically and creatively designed, so its implementation needs to be carefully planned and executed.
The Design Group can now usefully change its role to that of implementation group and some of the diagnostic tools which it used as designer can prove useful in its new role.
It must now ask the following questions:
1. What kinds of problems are likely to be encountered on implementation and how can these be avoided?
2. During the implementation period what activities will have to be co-ordinated, both within the design area and between it and other areas?
3. What training is necessary and how and by whom will this be provided?
4. How much time is required for implementation and how can progress best be monitored?
The answers to these questions will provide the Design Group with an understanding of what is likely to be the best implementation strategy. For example, whether to use a total change strategy in which the old system is removed and the new one substituted virtually overnight. Or whether to use a phased change strategy, with parts of the new system being introduced over an extended time period.
Here again is an opportunity for discussion between the Design Group and the Steering Committee and between the Design Group and its constituents.
Once the new system is in, working and has settled down, its ability to meet all the objectives set for it at the design stage must be tested. Two of the tools used at the diagnostic stage are now used again. These are variance analysis and job satisfaction analysis.
It will be recalled that system design was directed at achieving two things. Improved efficiency and improved job satisfaction. Improved efficiency can be checked by establishing the extent to which key variances — systemic weaknesses that cannot be eliminated—are now effectively controlled. And the extent to which operational variances—problems arising from the way the previous system had been organized—have been eliminated without new ones taking their place.
ETHICS has now been used in many firms in the UK and extensively by a major manufacturer in the United States. Experience has shown that it produces well designed "total" systems in which technology and work organisation fit easily together. The manner in which the two reinforce each other's positive characteristics increases company efficiency and employee job satisfaction and improves the quality of the work environment.
ETHICS requires no more man days than conventional systems design, but it does require more lead time. Design Group members and their constituents have to learn new analytical and design skills, and this takes time. It is time well spent, however. As professional systems designers become increasingly busy the ability of users to manage their own change becomes a valuable company resource.
It must be emphasised that the definition of a user in ETHICS is anyone who will be directly, or indirectly, in contact with a new technical system. Therefore all groups must be involved in the design process. In this, it is similar to the Japanese Quality Circle.
Finally, it must be recognised that today the design task never ends. As a new system becomes operational, so it becomes obsolescent. Design is now a merry-go-round which never stops. It is our need and challenge to keep up with it.
Designing Human Systems for New Technology - The ETHICS Method,
by Enid Mumford (1983) ISBN 0 903 808 285
Copyright © Enid Mumford (1983). All rights reserved.
Note to students: you are encouraged to refer to this book in your work but please remember to include full acknowledgement of the author. Failure to do this might mean you would be accused of plagiarism.Top of Page