How to build a theoretical model

My note in March 2008 (from my research diary)

A model is a simplified (often mathematical) description of a system etc., to assist calculations and predictions.

Steps in building a theoretical model:-

1.  The variables to be used in characterising and understanding the process must be specified.
2. The forms of the relationships connecting these variables must be specified.
3. Ignorance and the need for simplicity will ensure that all relationships other than identities are subject to error and so, for purposes of efficient statistical estimation, these error terms must be specified.
4. The parameters of the model must be estimated and the extent of its identification ascertained; if this is inadequate, the model must be reformulated.
5. Finally, the model must be kept up to date and used, so that an impression can be formed of its robustness and reliability.

Theoretical models are of many kinds: static or dynamic; partial or complete; aggregated or disaggregated; deterministic or stochastic; descriptive or optimising.