INVESTIGATING ESTIMATOR CONSISTENCY OF NONLINEAR GROWTH CURVE PARAMETERS

Abstract

Researchers often model biological growth data as a nonlinear function of time. This function involves unknown parameters that have intrinsic meaning to the researchers. Maximum likelihood provides a way to estimate these parameters. We present asymptotic results for the maximum likelihood estimators when adopting the generalized logistic function. Under increasing-domain asymptotics, the maximum likelihood estimators are consistent. Under infill asymptotics, evidence suggests they are not.

Keywords:

• Infill asymptotics
• Maximum likelihood
• Spatial process

ACKNOWLEDGMENTS

The author thanks Fred Dahm for his assistance and guidance during the research that evolved into this paper. The author also thanks John Crown for his constructive comments during the preparation of this paper.