QUANTITATIVE MODELS AND METHODS

Hopefully, the examples in the previous sections have shown the relevance, as well as the limitations, of quantitative analysis for real-life business decisions. In the following chapters we cover a wide variety of tools, which could be somewhat confusing for the reader. Hence, it is a good idea to set a conceptual framework to classify and interpret the different approaches, which should not be regarded as a disordered array of technicalities. Applying a quantitative analysis typically requires two steps:

  • Building a quantitative model of a system, or part of it
  • Solving it by some suitable method

To see the difference, consider the linear programming model to find the optimal mix in Section 1.1.2. After you have written the model down, a numerical solution procedure should be applied to come up with the answer we need. Lightning fast software packages to solve linear programming are widely available; hence, emphasis should be placed on model building, rather than model solving. In fact, data collection, a correct assessment of objectives and constraints, and the ability to interpret the solution are definitely more critical to success than plain number crunching.

Indeed, the title of the book is arguably somewhat misleading, even though quantitative methods is the standard name of courses covering these subjects. Nevertheless, it is often essential to have at least a rough idea of the inner working of solution methods, because

  1. The way the model is formulated may have an impact on available methods, as well as their efficiency.
  2. The solution method may be valid only if some assumptions about the data hold; ignoring these assumptions may result in gross mistakes if a method is applied to the wrong problem.

Another important healthy principle that we should keep in mind is21

All models are wrong, but some are useful.

Indeed, any model is a simplified representation of reality, and it need not be quantitative. Qualitative models are very useful in business process reengineering (BPR), and are often based on some graphical formalism to clarify relationships between actors, chain of events by activity diagrams, etc. Quantitative models, on the contrary, are based on numerical information.


Comments

Leave a Reply

Your email address will not be published. Required fields are marked *