Premium
Paired Comparison Product Testing When Individual Preferences are Stochastic: An Alternative Model
Author(s) -
Horsnell Gareth
Publication year - 1977
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.2307/2347024
Subject(s) - econometrics , product (mathematics) , computer science , mathematics , geometry
Summary Wierenga (1974) has asserted that the preference of an individual in a paired comparison product test is stochastic, i.e. the probability p, that he will prefer one of the two alternatives is not 0 or 1, but lies somewhere between these values and is itself variable over the population. This paper shows that Wierenga's data are adequately fitted by a composite model where some responses are deterministic and some are random with a common probability of choice between the two products. It is also demonstrated that the composite model provides a better basis for inference and decision‐making and it is argued that Wierenga's proposed stochastic distribution is an unnecessary assumption in the conditions of product testing.