Statistical Vector Field Analysis
Applied to Mixed Cross--sectional and Longitudinal Data
Steven M. Boker
J. J. McArdle
Department of Psychology
The University of Virginia
Charlottesville, Virginia 22903
Combined cross--sectional and longitudinal data often present complex patterns of change. Growth functions representing the change in some measure as a function of chronological age can themselves be a function of initial values of the measurement. Individual differences in developmental age with respect to chronological age tend to distort attempts to fit a single growth curve through combined cross--sectional and longitudinal data. This paper presents a method by which these data can be visualized along with several examples from a data set comprising measurements of intellectual abilities with respect to aging. We call this new method a statistical vector field ( svf) plot. An svf plot simulateously allows the visualization of cross--sectional information summarizing sampling densities, and longitudinal information summarizing how individuals at each particular age and ability are likely to change over time.
Vector field, cross-sectional and longitudinal data, growth curve, scientific visualization, statistical vector field, svf software.
The C source code and an IBM-PC compatible compiled version of the svf software described in this paper will be made available from the University of Virginia anonymous ftp archive server ( ftp.Virginia.EDU).
Steven M. Boker
Sun Feb 12 18:20:50 EST 1995