The initial conditions problem in longitudinal binary process: A simulation study

In this paperwe report some simulation studies to compare twobasic approaches inmodelling the random effects for longitudinal binary data. The first, naive approach, treats the endogenous initial conditions as exogenous and the second, correct approach,models the initial conditions as endogenous. The initial conditions problem in binarydata,with unobservedheterogeneity, arises when the process has a Markov property. To see the effect of the error variance and Markov effect, different values for these parameters are considered.We will apply the proposed approach to the data on depression. We show that, in the presence of heterogeneity and stat dependence, ignoring the initial conditions results in biased estimates and misleading interpretation. 2005 Elsevier B.V. All rights reserved.

[1]  R. Crouchley,et al.  A comparison of GEE and random effects models for distinguishing heterogeneity, nonstationarity and state dependence in a collection of short binary event series , 2001 .

[2]  Cheng Hsiao,et al.  Estimation of Dynamic Models with Error Components , 1981 .

[3]  Richard B. Davies,et al.  The Empirical Analysis of Housing Careers: A Review and a General Statistical Modelling Framework , 1991 .

[4]  Virginia A. Clark,et al.  Parameter Estimation for Mover-Stayer Models , 1983 .

[5]  Robert Crouchley,et al.  The Mover-Stayer Model , 1986 .

[6]  Bengt Muthén,et al.  A Structural Probit Model with Latent Variables , 1979 .

[7]  Marco Alfò,et al.  Random coefficient models for binary longitudinal responses with attrition , 2000, Stat. Comput..

[8]  Cheng Hsiao,et al.  Analysis of Panel Data , 1987 .

[9]  Cheng Hsiao,et al.  Formulation and estimation of dynamic models using panel data , 1982 .

[10]  Alok Bhargava,et al.  Estimating Dynamic Random Effects Models from Panel Data Covering Short Time Periods , 1983 .

[11]  R R Frerichs,et al.  Prevalence of depression in Los Angeles County. , 1981, American journal of epidemiology.

[12]  Nariida C. Smith,et al.  Estimating automobile utilisation with panel data: An investigation of alternative assumptions for the initial conditions and error covariances , 1990 .

[13]  P. Diggle.  Analysis of Longitudinal Data , 1995 .

[14]  J. Lindsey.  Models for Repeated Measurements , 1993 .

[15]  Marc Nerlove,et al.  Further evidence on the estimation of dynamic economic relations from a time series of cross-sections , 1971 .

[16]  Robert Crouchley,et al.  A comparison of population average and random‐effect models for the analysis of longitudinal count data with base‐line information , 1999 .

[17]  P. Thall,et al.  Some covariance models for longitudinal count data with overdispersion. , 1990, Biometrics.