A series of well-child visits maintained during the first 2 years of life has a positive effect on health outcomes as indicated by a decrease in avoidable hospitalizations among poor and near-poor children, regardless of race, level of poverty, or health status. National efforts to improve the quality of child health services for young children should focus on increasing compliance with periodic preventive care for young children in addition to improving immunization levels.

When direct measures of response error are not available in multiwave panel data, latent Markov chain models can provide a useful framework for analysis of response errors. These models permit the estimation of real change and response error even when the observations are widely spaced. This methodology is illustrated with Social Security Administration panel data from the years 1971, 1972, and 1974 (see Frohlich 1975 and Social Security Administration 1979a on self-reported disability status. The model shows much less real change than that indicated from inspection of raw turnover tables. Direct estimates of measurement error from a later survey (1978) give a proximate confirmation of the model results.

A MIMIC model is developed that contains a one-dimensional latent variable measured by ordinal indicators. The latent variable can be either discrete and ordered or continuous. Parameters for both the measurement and structural components of the model are estimated simultaneously using the EM algorithm. An example of a sociological application is presented.

No abstract

Parameter estimation and the computation of standard errors in social science models often require the assumption that observations are independent. This assumption is frequently violated with pooled cross-section and time-series data and household survey data. A recent article by Liang and Zeger (1986) shows that classical estimation methods retain good statistical properties in a wide variety of analyses where observations are not independent, and that correct standard errors of estimated model parameters are not difficult to compute. This article describes one of Liang and Zeger's results and presents its applications to the estimation of linear and logistic regression models from pooled cross-section and time-series data.