Abstract
In Chapter 5 we introduced some basic notions regarding the specification of marketing models. In this chapter we will study a number of criteria which a model structure should satisfy in order to stand a good chance of being implemented. The likelihood of model acceptance depends also, as was indicated in Chapter 4, on a number of other criteria which have little or nothing to do with model structure. These will be analysed in Chapter 13.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
See, for example, Leeflang and Koerts (1973).
The reader may wonder whether one will be able to estimate the parameters in the model wanted by the user. It is at this point that the combination of subjective with data-based estimation will come in. This is elaborated upon in Chapter 11.
For some authors, for example Larréché and Montgomery (1975), ease of testing is itself an implementation criterion related to model structure.
See, for example, Montgomery and Urban (1969, p. 5).
Communication is also becoming easier because students in many business schools, often potential users, are given at least some superficial exposure to model building jargon. Also, potential model builders, are introduced to issues of implementation.
Amstutz (1970) proposes something which is conceptually similar.
See also Somermeyer (1967), Leeflang (1974), and Leeflang and Koerts (1975).
A simple way to combine a laboratory experiment to determine price elasticity with historical data analysis on other marketing instruments is proposed by Naert (1972).
Of course, there remains the problem of coordinating the various subproblems in order to make sure that subunits do not pursue objectives that are in conflict with overall company objectives. Internal or transfer pricing can be successfully applied to overcome such potential conflicts. See, for example, Baumol and Fabian (1964), and Hass (1968).
See, also Leeflang (1977b, 1977c).
On the application of express see, for example, Book and Latour (1977).
To avoid confusion, it should be made clear that robustness has a totally different meaning in statistics and econometrics. According to Theil (1971, p. 615), a statistical test is called robust if it is insensitive to departures from the assumptions under which it is derived.
Put in mathematical terms, first- and second-order derivatives should have the right signs.
The rest of this chapter closely follows Naert (1974).
In a sense we are then placing ourselves within the realm of positive economics. See, for example, Friedman (1953).
For simplicity of exposition we will not write error terms.
Constraints for the general linear model (with constant term) were derived by Schmalensee (1972) and Naert and Bultez (1973). See also McGuire and Weiss (1976) and Weverbergh (1976). For a further discussion, see Chapter 8.
In fact Beckwith was concerned with the simultaneous estimation of a number of brands. Here we report the ordinary least squares (OLS) results for one of the brands. In brackets under the coefficients are the t-statistics.
We will see in Section 6.5 that (6.7) is in fact only robust in a limited way.
See Section 5.3.1.4.
An alternative would be to estimate equation (6.7) directly by means of nonlinear estimation techniques.
For a more detailed description, see Naert (1974). Both α and ß were ultimately varied in steps of 5,000. No finer steps were judged meaningful because of the low sensitivity of R 2 to α and ß in the neighbourhood of the optimum.
See, for example, Theil (1965b, pp. 7–9). See also Chapter 12 on validation.
For evidence that low sensitivity is also frequently observed in reality, see Naert (1973), and Bultez and Naert (1977).
An index j has been added for clarity of the subsequent discussion.
See also Leeflang (1977d).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1978 H. E. Stenfert Kroese B. V.
About this chapter
Cite this chapter
Naert, P.A., Leeflang, P.S.H. (1978). Implementation criteria with respect to model structure. In: Building Implementable Marketing Models. Springer, Boston, MA. https://doi.org/10.1007/978-1-4615-6586-4_6
Download citation
DOI: https://doi.org/10.1007/978-1-4615-6586-4_6
Publisher Name: Springer, Boston, MA
Print ISBN: 978-90-207-0674-1
Online ISBN: 978-1-4615-6586-4
eBook Packages: Springer Book Archive