My curiosity, here is how the researcher intend to use the 5ps or 6ps of marketing (i assume these are the inputs) against a single out(price). In a distribution free, non-parametric environment like DEA without using the price function or Malmquist Production Index (MPI)or will the MPI be decomposed to a CRS/VRS - DEA frontier(s)
i would like to know what kind of DEA model that will be appropriate for the research.
with a sungle output, you might rather use stochastic frontier analysis. It is easier to sell and doesn’t assume all inputs being really necessary.
I upload DEAwithCrossEfficient.pdf that is one of first chapter of DEA with CCR and Inverted DEA models and crossEficient published from Amazon.
I explain how to build LINGO SEts model and DEA.
I think there are problems even with the first 4 P's of the marketing mix
Product
Price
Place
Promotion
Only two are likely to be able to be reduced to numerical data easily. Price and Promotion. Place is likely to moderate these first too so if you are selling the same product into two countries or via two distribution channels the price and promotion of each may differ. Note you may also vary the exact product on offer too eg if the product is a book and place is used to distinguish between ebook sales in US and bookstore sales. Product is different and promotion is also likely to differ as the physical sales promotion will include promotion to the distributors and book store chain buyers but would be numerical variables in DEA would suggest multiple variable for price and promotion instead.
- Badges
- Science topic