Applied models and the use of quantitative techniques
- PDF / 1,363,432 Bytes
- 2 Pages / 589.56 x 841.92 pts Page_size
- 119 Downloads / 188 Views
EDITORIAL
Applied models and the use of quantitative techniques From time to time there appears in the literature a survey of the use of OR techniques. Typical conclusions from these surveys are that statistical analysis and simulation are the most used techn- iques, but that other standard quantitative techniques usually included on OR courses appear to be relatively little used. The experience of many practitioners bears this out: 'We do lots of good work but we hardly
between elementary and 'expert' use of model
surveys need to be treated with caution since they can easily be criticised on methodological grounds. For example, surveys of identified operational researchers don't reveal what use is being made of OR techniques by other groups of people not included in the survey. Another difficulty is defining the nature of a given'technique' and what constitutes use of the technique. In the first place it is important to distinguish between the processes of problem
'tool bag' approach to managerial problems. Ac-
evaluation techniques. Surveys on the use made of techniques need to address these dimensions if they are to have any real value. What surveys on the use of techniques do highlight is that very little work appears to have done on why 'techniques' are not more widely used. Apart from obvious practical difficulties in exploring this issue, there may also be a concern to avoid advocacy of a
ever use the techniques discussed in most OR textbooks'. However, the conclusions of such
cepted wisdom is that, rather than look for problems to which a particular tool can be applied, one should primarily seek to address important problems. This assumes that the OR practitioner is familiar with a
wide variety of generic models and associated
evaluation techniques and will apply these efficiently and effectively as a situation warrants. Increasingly
this assumption is becoming unrealistic, as both techniques and associated software packages pro-
formulation and problem evaluation.
liferate and increase in sophistication. One re-
sponse is an increase in specialisation of individual
management scientists or groups. Another re-
Problem formulation typically involves model build-
¡ng of some description, whether quantitative or qualitative. The classic approach is to recognise that a given situation can be represented by a general OR model with suitable modifications. In
sponse is to continue offering a general service, but
one which emphasises a pragmatic approach
biassed towards simplicity, robustness and transparency in model formulation, together with a will-
ingness for qualitative as much as quantitative
principle, evaluation is then straightforward using the
body of (mathematical) methods that have been developed to evaluate the generic model form and its variations. For example, a production mixing problem might be represented by an LP model formulation. Depending on the size and particular
analysis. This is frequently what the client needs, at
least to start with. However, while this pragmatic
Data Loading...
