Discrete Choice Modelling is a specialised method with applications across a number of academic disciplines and myriad practical uses.
What is Discrete Choice Modelling?
Discrete Choice Modelling (DCM) analyses the choice behaviour of individuals and groups who face discrete economic alternatives:
- buying a TV;
- taking out a home loan;
- choosing a holiday;
- voting for a political party;
- betting on a horse, etc.
The theory and methods were pioneered by Dan McFadden at the University of California, Berkeley. The first practical application (1972) was spectacularly successful and accurate (see BART Case Study) and McFadden won the 2000 Nobel Prize in Economics for his work.
A Choice Model, like a prism separating a beam of light into its component spectrum, will identify the ‘value’ of each feature of a product.
We use these values to calculate the probability an individual will purchase a product with certain features and price.
Applied across the population, this probability translates to Market Share.
We can compare two product configurations, to predict which features and price will produce the largest Market Share.
Features can be product attributes (colour, size), or information attributes (advertising, product reviews).
Furthermore we can use Choice Models to calculate:
- (1) Willingness to Pay
- (2) The effect of different advertising campaigns on sales
- (3) How consumers trade off between features
- (4) The value of a Brand
|
