As mixture regression models increasingly receive attention from both theory and practice, the question of selecting the correct number of segments gains urgency. A misspecification can lead to an under- or oversegmentation, thus resulting in flawed management decisions on customer targeting or product positioning. This paper presents the results of an extensive simulation study that examines the performance of commonly used information criteria in a mixture regression context with normal data. Unlike with previous studies, the performance is evaluated at a broad range of sample/segment size combinations being the most critical factors for the effectiveness of the criteria from both a theoretical and practical point of view. In order to assess the absolute performance of each criterion with respect to chance, the performance is reviewed against so called chance criteria, derived from discriminant analysis. The results induce recommendations on criterion selection when a certain sample size is given and help to judge what sample size is needed in order to guarantee an accurate decision based on a certain criterion respectively.
Download Info
To download:
If you experience problems downloading a file, check if you have the
proper application to
view it first. Information about this may be contained
in the File-Format links below. In case of further problems read
the IDEAS help
file. Note that these files are not on the IDEAS
site. Please be patient as the files may be large.
Find related papers by JEL classification: M31 - Business Administration and Business Economics; Marketing; Accounting - - Marketing and Advertising - - - Marketing C30 - Mathematical and Quantitative Methods - - Multiple or Simultaneous Equation Models; Multiple Variables - - - General
This paper has been announced in the following NEP Reports: