Premium
Exact group comparisons using irregular longitudinal data
Author(s) -
Maritz J. S.,
Lombard C. J.,
Morrell C. H.
Publication year - 1998
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 - 10.1111/1467-9876.00114
Subject(s) - statistics , mathematics , variable (mathematics) , permutation (music) , longitudinal data , group (periodic table) , random variate , algorithm , computer science , random variable , data mining , mathematical analysis , physics , chemistry , organic chemistry , acoustics
Longitudinal data on an experimental subject arise when observations of a dependent variable are made at several successive values of an indexing variable x ; often the indexing variable is time. When such data are recorded for a number of subjects, typically animal or human, the x ‐configuration may vary from subject to subject, producing irregular longitudinal data. Exact permutation methods are considered for comparing two or more groups of subjects. This approach entails choosing appropriate descriptive and test statistics and generating their exact distributions. The method is applied to two data sets, one relating to pigs and one to baboons. The second of these data sets involves a co‐variate and the results are compared with those yielded by fitting a parametric mixed effects model. In the first example differences between groups were clearly established, and the graphical representations proposed were useful for describing the differences. In the second example the agreement between the results of the two methods of analysis was found to be very close.