Abstract:
|
Repeated measures are of interest to marketers (Lessne & Hanumara, 1988). Many marketing issues deal with the study of change in marketing variables based on analyses of repeated measurements, or at different levels of an independent variable. Growth analyses of demographic variables such as population demographics are of utmost importance, especially in the United States, which is fast becoming an older nation. Traditional statistics in studies of demographic change analyzing longitudinal data use Ordinary Least Squares regression pooled across repeated measurements (Steenkamp & Baumgartner, 2000), and ANOVA; which are not appropriate, because such longitudinal data are autocorrelated (Timm, 1980). So, data from the AGing Integrated Database (AGID) was used to run growth curve models using AMOS 20.0. Results suggested that for the baby boomer population, a two-factor unspecified LGM was the best fit. This showed a linear growth in the overall baby boomer population. Among the Hispanic population, a three-factor polynomial LGM was the best fit. A chi-square difference test and acceptable values of fit suggested a possibility of quadratic growth in the Hispanic population.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.