Model Management by Means of Computer-Based Information Systems
in Managerial Contexts
Fons Wijnhoven
University of Twente, Fac. of Management Studies,
Dept. of Information Management, P.O. Box 217,
7500 AE Enschede, The Netherlands.
ABSTRACT
Models and computer-based information systems frequently meet resistance and suspicion by
management, because they often do not meet the knowledge demands of management in a company.
Solving this problem requires approaching modelling and information systems development as a
management discipline. This discipline’ involves the activities of developing, maintaining, effective
using, and conserving of models and systems. The paper concludes with a normative view of the
relation between management levels and model management activities, and considers the possible use
of computer-based information systems for effective model management.
THE PROBLEM
Modern organizations are confronted with increasing managerial complexity resulting from shorter
product life cycles, increasing market demands, increasing organizational size, increasing capital
intensity, and changing laws that demand a reduction of environmental pollution and market
liberalization: This growing complexity has traditionally been met by allowing slack resources (e.g.
large work-in-process capital and stocks) and independent tasks or working groups (Galbraith, 1973).
More recently, managerial complexity is also coped with by the formulation of specific company
strategies aiming at a reduction of the amount and diversity of markets and products to be managed
(e.g. focus and niche strategies, see Ansoff, 1987; Porter, 1985) and aiming at outsourcing of company
divisions.
The subject of ‘this paper is to explore the strategy of increasing the existing information processing
capacities in organizations. This interest is motivated by the growing technical opportunities of
computer-based information systems in recent years (lower priced and improved user-interfaces) and
developments in human intellectual capacities to make sense out of data, e.g. as a result of
improvements in system dynamics (Senge, 1990; Vennix, 1990). The problem with both developments
is that they frequently are means and technologies without a clear link to managerial needs. When
not first seriously considering what managerial requirements both can meet, and where its limits are,
history will probably repeat the. well known problems with M.I.S. and Operation Research, which both
overshoot demands and pretended much more than what could be considered realistic (Ackoff, 1967
and 1979).
The purpose of this paper is to derive criteria from insights in business management, by which we
can assess the value of models and the contributions of computer-based information systems for
management. To avoid a technology push, I will first describe management types that explain different
model needs. Second, I will investigate the possible role of computer-based information systems in
management (development, maintenance, use and conservation) of models. Both discussions will
converge to a normative view of types of models and computers for different management levels.
- 817 -
MANAGEMENT TYPES AND THE NEED FOR MODELS
Because of their ability to assess problems and find means-ends relations in a systematic way, models
can be of much value to management. The actual use of the models is, however, highly dependent on
their shape (mathematical or qualitative, complex or simple, formal or informal). Different types of
management require different types of models. Strangely enough, no explicit consideration about the
relation between management types and the shape of models is known to me. Therefore, I will first
describe a typology of management and then explore its consequences for models.
1. ‘Typology of management.
The most well known typology of management consists of a distinction between strategic planning,
tactical management, and operational planning and control (Anthony, 1965). Strategic management is
concerned with the definition of company goals, policies and general guidelines charting the course of
the organization. Tactical management is concerned with the acquisition of resources, the formulation
of acquisition tactics, determination of plant locations, and the arrangement and monitoring of budgets.
Operational management involves effective and efficient use of existing facilities and resources to
carry out activities, and monitoring and control of activities in relation to precisely formulated targets
and (budget) constraints.
2. Strategic management and models.
‘When formulating organizational goals, management aims at prescibing what the company should do
and what should be like for staying viable. This idea generating process mostly uses only implicit
models about goals and means. Sometimes, like in Delphi-sessions, such a management process is well
organized and specifically aims at finding consensus among its participants (Linstone and Turoff,
1975). This can be done by first finding consensus about the main factors at stake (definition of an
idiom, vocabulary) and then second, members can start discussing relations between these variables.
One step further, people can try to quantify these relationships. At the strategic level, however, models
stay quite soft and not much effort is done in making the models very precise. The most important
thing at this level is finding a common language which is politically acceptable (De Geus, 1988).
The emerging model must be motivating and directing people in the organization. Becaiise models at
this level stay rather imprecise (frequently with very imprecise definitions), they are flexible enough to
be used for legitimation or explanation of actions and they are sometimes improved after reflecting on
actions (Mintzberg, 1987). Alternatively, many business scientists have acquired knowledge which,
when used well, improves the chance for developing effective plans and strategies and making the
organization more proactive as well (Ansoff, 1987; Porter, 1985; Buzzell and Gale, 1987).
Besides these more or less scientific tools, also much information from informal media, news papers,
literature and intelligence services could be used. It is necessary to have creative minds which connect
all these impressions and make sense out of them with relevancy for finding a way the company
should go. In this case, management is clearly rational as well as intuitive (Mintzberg, 1976).
System dynamics modelling is very useful for strategic management, because it is a means for
organizing ideas in a systematic way (from very soft to very mathematical) and it supports the
understanding of unintended impacts of policies (Vennix 1990). One example of this way of reasoning
is given by Senge (1990: 99-100) in an assessment of quality circles. Senge states that quality circles
were initiated to improve problem solving by means of open communications. Open communications,
however, were regarded as threats to traditional union-management relationships, because they were
inconsistent with adversarial attitudes of unions to management and were a threat to management's
uni-lateral control. This means: the more successful quality circles are, the more resistance they will
meet, possibly leading to their failure at a later stage.
- 818 -
System dynamics models for strategic management will mostly be very soft, having no explicit
measurements of variables and no estimation of parameters. This type of soft modelling (Checkland,
1981) is precisely what is needed for effective strategic management.
3. Tactical management and modelling.
At this management level, many models are developed and improved on the basis of experience.
Business Science Schools are developing ideas especially in this area; e.g. econometric models for
planning fund needs (e.g. Du Pont analysis described. by Weston and Brigham, 1979; M. Aiso, 1973
for an econometric approach to financial forecasting). Many quantitative variables are involved here.
Just to mention a few: interest rates, inflation, growth expectations of production and markets, need for
capital and people, fiscal policy with respect to interests, dividends, taxes assets etc. On basis of these
variables, one could develop a model with which predictions can be made about fund needs and
determine the best type of funding (loans, stocks). These models can be used many times with some
adjustments on parameters involved (interest rates etc.).
For budgeting many models can easily-be developed by means of a spreadsheet. Moreover, one needs
a systems for evaluating budgets and supporting cost accounting, which contain a model with process
variables to be measured, goals variables and analysis tools for comparing performance against agreed-
on norms (c.f. Lawler and Rhode, 1976; Hofstede, 1981).
The determination of plant locations, contains decisions that can be supported by models as well.
Important variables are: human resources available in the neighbourhood, availability of transportation
and other infra-structures, availability of mineral resources, fiscal policy. political stability and support
from (local) government.
New products must be developed, produced and marketed. These activities require large sums of
money, with sometimes an uncertain contribution to a company’s profits. In this situation, market
information is mostly outdated before the product comes on the market. Therefore, a more long-term
view is required, in which characteristics of the aimed market(s) are systematically linked. This would
probably lead to a more or less qualitative view, containing ideas about cause-effect relationships.
At this level of management, system dynamics models can be helpful for assessment for decisions.
One example of this kind is given by Senge (1990: 316-320). A computer company had a very
effective sales group, realizing each year a 20 % growth in sales. Therefore, each year more salesmen
were appointed. Each year. the growth in sales personnel was realized by hiring experienced people
from other companies. At a certain moment, there were not enough experienced people to get from
competitors. The policy of personnel growth, therefore, did not work any more, The company had to
start training programmes for junior salesmen, led by seniors. The last group was not motivated for
this teaching job, because they received the main income from the contracts they realized. The
company had to chose between a different payment system for seniors or to decide to restrict growth
ambitions. Just hiring more junior salesmen would only lead to more management problems for the
sales division and could harm the reputation of the company.
Whereas at the strategic management level especially soft models were proposed, at this management
level more formal and quantitative models are applicable as well. I call these model medium soft/hard.
4. Operational management.
This management level is concerned with the effective and efficient use of existing facilities and
resources within given budget constraints. The essence of models at this level is to precise the
definitions of performance indicators, measurement instruments (for providing feedback information)
and ideas about effective actions to adjust behaviour. The model functions as a rigid set of rules and is
very far from the dynamics described earlier. Functional organizations with stable environments do not
suffer much from the rigidity of control systems. However, when more dynamics in organizational
environments exists, the organization can lose contact with its environment as a result of the
inflexibility of the steering models applied at operational planning and control. Feedback should be
applied to the principles of management (theory) which are used at this level and are imposed by
tactical management, System dynamics brings in important insights which should be added. It is not so
that the implementation of these dynamics is only computer-based. Systems around the control system
are essential for providing people with the necessary feedback. These "infra-structural” systems can
consist of open interactions among people to find better ways of learning, a management style that
supports learning and that is not based on uni-lateral control, a culture that aims at mutual support and
responsibility for the common goals of the group and a political atmosphere in which class struggles
are made subordinate to company performance should be created (cf. Argyris and Schon, 1978).
In an insurance company, management was based on the principle of hierarchy. This meant that
Teports were sent upward in the organizational hierarchy and inspected by management. Sometimes
actions were undertaken to motivate people and to adjust organizational behaviour. This way of
management did not fit with the view of the new president of the company. He suggested that the
company should become more flexible and more people should be concemed with management issues,
Therefore, a project called "MIS" was set up with the aim of measuring performance of people and
divisions. The information from this system has become available for division presidents and CEOs.
This meant that division presidents had access to data about their colleagues as well, which motivated
division presidents to be knowledgeable about data from their own division. In meetings of CEOs with
division presidents, performance problems were discussed on basis of the data. The open discussions
were not only a control on what is happening, but also led to supporting each other in finding
solutions. In this way mental models received many new useful ideas never thought about in. the
specific division.
In relation to the other management types these models are very quantitative, formal and precise, and
are called hard models. However, technical they may look like, they do not make human and social
Processes entirely obsolete.
COMPUTER-BASED INFORMATION SYSTEMS AND MODEL MANAGEMENT
In the previous section, the idea of different modelling needs related to different management types
has become obvious. Now, I shall study the role of computer-based information systems in these
modelling processes. The modelling process is aiming at supporting management. This can only be
realized when well-managed, leading to high satisfaction of the model user. Analogous to information
systems management, one can describe four functions in model management: development,
maintenance, use and conservation. The development process is well described in the literature on
information systems development and modelling. When finished, the system is said to be installed or
implemented. This means that it should be ready for use, which implies that it should be conserved
and available. Additionally, it must be maintained, meaning that it should be improved when required.
‘When costs for maintenance are too high in relation to the user’s satisfaction, the system or model
should be dismissed and the model management process can start again with developing a new model.
In the next section, I will explore these roles for model management in relation with computer systems
in more detail.
1 velopment.
Especially in the area of strategic management, the development of a model is important, because the
process of model development is recognized as an important learning result. When the knowledge is
acquired, the model is thrown away, because strategic management processes are mostly non-
- 820 -
repetitive, aiming at finding structure in fuzzy problems and agreeing on the way the problem(s)
should be perceived. Computer systems can be supporting this management level in three ways:
1. Use of existing external systems (data and models) by which comparisons with other companies are
possible.
One example of this kind is PIMS (Buzzell and Gale, 1985), a database containing data about
competitive positions, market characteristics, and capital and product structure. On basis of a large
database with market choices and business results of over 3,000 companies and strategic business
units, an econometric model has been developed with which data of a specific company subscribing to
PIMS, is compared with others and analyzed (anonymously). The aim of PIMS is to find the critical
success factors which are applicable to the company in its specific market and situation.
2. Frame works for structured thinking at the strategic level.
An interesting example of the support of structured thinking is developed in Ansplan, a software
package by which management thinking is organized by prompting the right questions and issues in a
systematic way (Ansoff, 1986). One part of Ansplan consists of the Strategic Position Analysis, which
is split up in an analysis of market attractiveness and an analysis of competitiveness. In the
attractiveness analysis predictions of demand and profits are made on basis of three methods:
* extrapolation of historical developments.
* developing expectations under "stable" market development. This means that changes that will
impact on success and that certainly are going to happen are also considered.
+ .developing expectations under “turbulent” conditions. Here the chance element is taken in
consideration and optimistic (opportunities) and pessimistic (threads) scenarios are developed.
The other part of Ansplan supports assessing strengths and weaknesses of existing human resources,
company competencies and other company assets. A comparison between market developments and
organizational strengths and weaknesses leads to the description of an organizational strategy.
3. Tools for organizing thought: System dynamics modelling.
Whereas PIMS is very useful for comparing the organization with others, Ansplan has the capability to
develop one’s own ideas in a systematic way. Nevertheless, both use many presumptions that are not
always regarded as relevant. Especially, when cause and effects, or means-ends relationships, are
considered, PIMS is restricted to the economic variables only, whereas Ansplan does not explicitly
support thinking in terms of means and ends. The importance of system dynamics is that it does not
requires heavy presumptions for management, except the idea that management is based on a theory of
actions that involves thinking about feedback processes, functions and dysfunctions of possible policy
alternatives, When companies meet disastrous environments, e.g. the oil crisis in 1972, new ways of
thinking must be developed that can cope with situations never encountered before (cf. De Geus,
1988). Our: knowledge and experiences, then can act as conditions making things even worse. In this
kind of situation brainstorming and Delphi-research are important processes in the board room.
Computers can support these processes with providing flexible tools for the organization of thought
only in terms of logic and not of assumptions. One of these tools is STELLA, which enables the _
flexible development of system dynamics models (cf. Vennix, 1990: 52-54 and 223).
In development processes, ways of reasoning are invented for structuring perceptions. This is typically
one of the major activities of strategic management. Models and model development tools. must,
therefore, contain a minimum of presumptions, or must have gained a superior high reputation, which
also could act as a legitimation for action (e.g. Ansplan and PIMS). Also in the case of reused models
the relation between management and models is very loose, leaving management with much tasks in
thinking and deciding themselves.
2. Maintenance.
At the strategic management level, models are used once for finding a way of thinking about the
problem at stake. The discovered method for thinking is institutionalized in data-gathering and analysis
departments or functions (cf. De Groot, Waalewijn, De Wolf, 1989). At the tactical management level,
complex problems having sometimes a repeating character, are handled, and it is worthwhile
considering to keep a model and do research on it to improve it as well. Two types of maintainable
knowledge are applicable here: quantitative types (based on statistics, mathematics and econometrics)
and qualitative types (based on knowledge based systems).
Quantitative models.
PIMS is again a good example for a quantitative model here. Knowledge from economics is
formalized in an econometric model and parameters are estimated on basis of the data received from
the 3,000 applicant companies, The Strategic Planning Institute now maintains the model and provides
individual companies with information about their position in relation to other companies via the
model (anonymous of course, and with the payment of an annual fee).
Some companies develop their own models for management. One case I met in a Dutch University
where the financial management office developed a model to distribute money among the several
faculties in a quantitative and precise way. In The Netherlands, it is generally so that the government
pays most funds for universities. The sum of money is applied as a lump sum, so that the university
board can apply its own policy to distribute it. The model was used as an instrument to make
discussions less political and speculative and make budget allocations more precise and predictable.
The result was that the model maker was asked each year to add more variables and parameters, so
thatthe model iss complex. now that only, a. few insiders con’ validate the;resilts and:'be serious
Participants in the discussion.
Qualitative models.
Qualitative models can have a very soft shape. At the moment knowledge engineers are studying the
application of computers for maintaining qualitative knowledge. Some major improvements in the field
of the application of knowledge based systems for managing manufacturing are mentioied by ‘Kerr
(1991). One of the basic features of the systems Kerr studied was that the principles in which a
Production process should be performed, can be formalized in so-called production rules. This means
that one can define precisely which substances should be added to the process in what quantities and
proportions, This has large implication for the planning of procurement, stocks and supplies.
Moreover, the implications of market strategies can be passed on in terms of needed capacities.
Essentially, the models developed are maintained not only because they are programmed, but also
because engineers and managers have access to them, can verify their validity and improve them when
necessary. :
Models at the tactical management level have a rather complex, formal and quantitative shape.
However, they are not used without a critical view and only when applicable. Own experiences and
new insights must be added to the model, and therefore (like a good computer based information
+ System) it must be open for maintenance activities.
3. Use.
At the operational management level, processes are monitored by comparing processes and activities
with norms, The norms are chosen on the basis of a theory of action, which management thinks will
give them most success. Therefore, the norms and the ideas behind them are a frame of reference by
which actions and data are interpreted and management actions are motivated. This frame of reference
is of course constrained by the general policy and objectives defined at the strategic and tactical
Management levels in the organization.. Again one could distinguish quantitative and qualitative
models.
Quantitative models.
i managers are frequently given some targets and short term goals, These are mostly defined
~- 822 -
in a precise quantitative form, e.g. the production of 1,000 computer mother board next week.
Performance control is simply the comparison of this norm with actual performance at certain point in
time. Nevertheless, there is more at stake. Knowing that a performance standard is not achieved is
only the start for finding out means how this problem can be solved and how management hypotheses
should be changed. It is sometimes very difficult to find the roots of the problem, because information
can be hidden for management, and there is no open learning orientation (Argyris and Schon, 1978).
Besides the organizational changes required for effective learning, also system dynamics modelling
tools can be useful for analyzing the situation and for finding the roots for the problems discovered.
For instance, system dynamics models are useful for extrapolating the consequences of not meeting
deadlines and from these insights and consequences, alternative actions can be found. The
organizational techno-structure can be useful in supporting the development of models and ideas for
solutions (Mintzberg, 1979). They are also the guards of models and knowledge for the problems
which are not used on a daily basis.
Qualitative models.
For analyzing the roots of problems and actions to be taken, the use of knowledge-based systems is
growing in importance. As stated before‘ production rules can be formalized and become part of an
expert system. Coupled with a monitoring system, analysis of problem causes is possible and actions
can be triggered, In monitoring of chemical processes, this can be part of the process control system.
In manufacturing it can be added to existing production planning and resource planning systems like
MRPII (Kerr, 1991). This means that computers are used as instruments for the interpretation of data.
For developing such models and systems, a clear view must be available of the production process and
about factors influencing its performance. Also feedback loops must be defined. Systems Dynamics
then is an integral part of systems development and the way processes are actually controlled.
Especially at the operational management level, hard models are introduced to be used not too critical.
Even not applying them could damage the consistency of management. This does not mean that
managers have no decision discretion any more. Applying one model does not mean that only one
solution is possible. Use of the systems applicable here (be it MIS, DSS, MRPII or what else) is
highly compulsory at this level.
4, Conservation.
Organizations always are in a tension between renewal and reusing of solutions. From one perspective
(efficiency), it is not smart to regard all problems as unique and, thus, having to reinvent solutions.
Therefore, it might be useful to conserve knowledge generated from experience and earlier research in
models, knowledge bases and data bases. Conservation can be realized by giving specific
organizational units this job. In strictly functional organizations this task is (implicitly) allocated to the
technostructure (saving knowledge about technology of production in the shape of soft or hard models)
and the support staff (saving administrative data about specific cases, aggregate or individual)
(Mintzberg, 1979). Besides, much of this knowledge can be saved on computer systems in model
bases and knowledge-bases (saving production rules, recipes etc.; cf. Kerr, 1991). When these models
and knowledge bases are not used frequently (especially in case of tactical management), knowledge
about them starts to erode, leading to underutilization of existing knowlegde. In this case, internal
consultants in the technostructure, the support staff of the organization, or specialists within a
functional unit or working group, could act as model librarians and consultants.-The model consultant
also should make models available for organizational groups that have not been involved in the
development of the model, or have never encountered the related problems before.
The conservation of models can be very effective and efficient at the operational management level
because the same models are used on a very frequent basis in recurring decisions. At the tactical
management levels, model conservation is useful, because of the costs of developing them anew. As
- 823 -
stated before, the model should then be open for amendments and improvements. At the strategic
management level, model conservation is frequently outsourced and done by highly respected institutes
that also take charge of improving the models on basis of new scientific insights.
However interesting the subject of model conservation may be, some warning is necessary because it
could easily lead to.a conservative mind not adapting to new and future problems. Therefore, a system
dynamic approach to knowledge conservation is absolutely essential for keeping a critical view on the
impact of historical solutions for the future.
SYNTHESIS
It was suggested that strategic management uses primarily the model development activity, with
specific computer support. At the tactical level critical reuse of models is proposed, demanding model
maintenance activities and models that are open for maintenance. At the operational level compulsory
use of models is proposed. The conservation activity is mainly outsourced at the strategic application
field, conserved by specialists in the technostructure and support staff of the organization at the tactical
application field, embedded in computer application that are frequently used at the. operational
management level. These propositions are summarized in next table.
Model management activity
|| Soft modelling.
| Computer conservation by
support by highly respected
Stella, Ansplan, institutes.
PIMS
able 1: Norma
use involved.
The table describes a normative view on model types and computer use for different management
levels. Diverging from these norms (the empty cells 1 through 6) will likely lead to the same problems
OR. and MIS have met in the past, because of a discrepancy between their features and management
demands. I will shortly describe the contents of these open cells.
Cell 1. Medium sofi/hard models used for strategic management can likely lead to resistance of use,
because strategic management will then feel too restricted by presumptions they do not agree or which
they find irrelevant.
Cell 2. Using too soft models for tactical management, will not give them enough constraints (e.g.
organizational goals and directives). They will feel uneasy with the fuzziness of the models and feel
- 824 -
unconfident for using them during decision making.
Cell 3. Using hard models for tactical management will lead to problems, because tactical management
will feel that the model is too formal and overprecise.
Cell 4, Applying medium soft/hard models to operational management will lead to failure because the
model does not give the support for making the precise decisions needed at this management level.
Cell 5. Hard models for strategic management are overshooting, discussing issues in a too precise and
formal way and therefore lead to severe resistance and disqualifications of the supplier or consultant
Proposing it.
Cell 6. Soft models for operational management will lead to serious disqualifications of the consultant
or supplier again.
Cells 5 and 6 are very obviously wrong. Cells 1 through 4 are more difficult to assess, because some
models and systems can be interesting for more then one management level. Besides, it needs a very
precise understanding of the way management is thinks and works.
CONCLUSIONS
In this paper I have described a link between modelling, computer-based information systems and
managerial needs for both technologies. The resulting insights lead to some propositions that could
explain historical problems in this area and give ideas for avoiding them in the future, Nevertheless,
for this proposed model counts’ the same as for all other models: it should not only be used and
conserved, but also be maintained and if necessary be replaced by better ones. For instance, one could
study in more detail the organizational environmental conditions (culture, management style, market
environment) involved with effective model and computer use in a management context. What should
be aimed at is not the optimization of the means (models and computers), but the optimization of the
interface between the means and the organizational goals.
ACKNOWLEDGEMENTS.
Many thanks to my colleagues Cees van Slooten and Karel Thinissen, who provided me with useful
suggestions and ideas for this paper.
REFERENCES
Ackoff, R. 1967. Management misinformation systems. Management Science 14: 147-156.
Ackoff, R.L. 1979. The future of operational research is.past. Journal of the operational research
society, 30: 93-104.
Aiso, M. .1973. Forecasting Technique. IBM Systems Journal 12:187-209.
Ansoff, Igor .1986. Competitive Strategy Analysis on the Personal Computer. Journal of Business
Strategy 6: 28-36.
Ansoff, Igor .1987. Corporate Strategy, second edition. London: Penguin Books.
Anthony, R.N. 1965. Planning and control systems: a framework for analysis. Harvard University
Press.
Argyris, Chris and Donald A. Schon. 1978. Organizational learning: a theory of action perspective.
Reading, Mass., Addison-Wesley publ.
Buzzel, Robert D. and Bradley T. Gale .1987. The PIMS Principles: linking strategy to performance.
The Free Press, New York.
Checkland, Peter .1981. Systems Thinking, Systems Practice. John Wiley and Sons, Chichester.
Galbraith, Jay R. 1973. Designing complex organizations. Reading, Massachuttes,
Geus, A. de .1988. Planning as Learning. Harvard Business Review: pp. 70-74.
Groot, E. de, Ph. Waalewijn and WJM. de Wolf .1989. Spiegel voor Strategen. [A study of PIMS
and Ansplan in 16 Dutch companies]. Dutch Society for Strategic Policy.
Hofstede, Geert .1981. Management Control of Public and Not-For-Profit Activities. Accounting,
Organizations and Society 6; 193-211.
Kerr, Roger .1991. Knowledge-based manufacturing management: applications of artificial intelligence
to the effective management of manufacturing companies. Addison-Wesly, Sydney Australia, ;
Lawler III, Edward E. and John R. Rhode .1976. Information and Control in Organizations. Goodyear
publ., Santa Monica. ‘
Linston, Harold A. and Murray Turoff .1975. (eds.) The Delphi Method; techniques and applications.
Reading Mass., Addison Wesley.
Mintzberg, Henry .1976. Planning on the left side and managing on the right. Harvard Business
Review July-August: 49-58.
Mintzberg, Henry .1979. The Structuring of Organizations. Englewood Cliffs, New Jersey, Prectice
Hall.
Mintzberg, Henry .1987. Crafting Strategy. Harvard Business Review, July-August, pp. 66-75.
Porter, Michael .1985. Competitive advantage: creating and sustaining superior performance. Free
press, New York.
Senge, Peter M. .1990. The Fifth Discipline: the art and practice of the learning organization.
Doubleday Currency, New York.
Vennix, Jac. A.M. .1990. Mental Models and Computer Models: design and evaluation of computer-
based learning environment for policy-making. University of Nijmegen.
Weston, J.F. and EF. Brigham. .1979. Essentials of Managerial Finance, fifth edition. The Dryden
- 826 -
