A note on a manova model applied to problems in growth curve

A Comparison of Efficiencies of Longitudinal, Mixed Longitudinal, and Cross-Sectional Designs

The choice among a longitudinal, mixed longitudinal, or cross-sectional design is often called for in educational and psychological research. The problem of choosing the most efficient design to estimate polynomial parameters for time-structured data is considered, and the comparison of the efficiencies shows that the assumed degree of the polynomial is crucial for the selection of the most efficient design. When the degree is much smaller than the number of time points and the correlations between adjacent time points are not too large, cross-sectional and mixed longitudinal designs are more efficient than a longitudinal design.

Regression for randomly sampled spatial series: the trigonometric case

The model Y(t) = s(t | θ) + ε(t) is studied in the case that observations are made at scattered points τ j in a subset of Rp and θ is a finite-dimensional parameter. The particular cases of 0 = (α, β) and (α, β, ω) are considered in detail. Consistency and asymptotic normality results are developed assuming that the spatial series ε(·) and the point process {τ j } are independent, stationary and mixing. The estimates considered are equivalent to least squares asymptotically and are not generally asymptotically efficient.

Contributions of the paper include: study of the Rp case, management of irregularly placed observations, allowance for abnormal domains of observation and the discovery that aliasing complications do not arise when the point process {τ j } is mixing. There is a brief discussion of the construction and properties of maximum likelihood estimates for the spatial-temporal case.

Nutritional Intervention

A secondary analysis of a nutrition experi ment in Cali, Colombia was conducted. The theme of the reanalysis was the notion of competing methods of analysts. Specifically, the effect of nutritional supplementa tion on the physical development of malnourished children was analyzed by means of univariate and multivariate methods for the analysis of repeated measures. It was concluded that the nutritional treatment was successful, but to assess the precise nature of the effect, it was necessary to resort to the multivariate approach.

Comment on the asymptotics of a distribution-free goodness of fit test statistic

In a recent article Jennrich and Satorra (Psychometrika 78: 545-552, 2013) showed that a proof by Browne (British Journal of Mathematical and Statistical Psychology 37: 62-83, 1984) of the asymptotic distribution of a goodness of fit test statistic is incomplete because it fails to prove that the orthogonal component function employed is continuous. Jennrich and Satorra (Psychometrika 78: 545-552, 2013) showed how Browne's proof can be completed satisfactorily but this required the development of an extensive and mathematically sophisticated framework for continuous orthogonal component functions. This short note provides a simple proof of the asymptotic distribution of Browne's (British Journal of Mathematical and Statistical Psychology 37: 62-83, 1984) test statistic by using an equivalent form of the statistic that does not involve orthogonal component functions and consequently avoids all complicating issues associated with them.

A new method to compare statistical tree growth curves: the PL-GMANOVA model and its application with dendrochronological data

Growth curves are monotonically increasing functions that measure repeatedly the same subjects over time. The classical growth curve model in the statistical literature is the Generalized Multivariate Analysis of Variance (GMANOVA) model. In order to model the tree trunk radius (r) over time (t) of trees on different sites, GMANOVA is combined here with the adapted PL regression model Q = A · T+E, where for b ≠ 0 : Q = Ei[-b · r]-Ei[-b · r1] and for b = 0 : Q  = Ln[r/r1], A =  initial relative growth to be estimated, T = t-t1, and E is an error term for each tree and time point. Furthermore, Ei[-b · r]  = ∫(Exp[-b · r]/r)dr, b = -1/TPR, with TPR being the turning point radius in a sigmoid curve, and r1 at t1 is an estimated calibrating time-radius point. Advantages of the approach are that growth rates can be compared among growth curves with different turning point radiuses and different starting points, hidden outliers are easily detectable, the method is statistically robust, and heteroscedasticity of the residuals among time points is allowed. The model was implemented with dendrochronological data of 235 Pinus montezumae trees on ten Mexican volcano sites to calculate comparison intervals for the estimated initial relative growth A. One site (at the Popocatépetl volcano) stood out, with A being 3.9 times the value of the site with the slowest-growing trees. Calculating variance components for the initial relative growth, 34% of the growth variation was found among sites, 31% among trees, and 35% over time. Without the Popocatépetl site, the numbers changed to 7%, 42%, and 51%. Further explanation of differences in growth would need to focus on factors that vary within sites and over time.

Limited Information Estimation in Binary Factor Analysis: A Review and Extension

Based on the Bayes modal estimate of factor scores in binary latent variable models, this paper proposes two new limited information estimators for the factor analysis model with a logistic link function for binary data based on Bernoulli distributions up to the second and the third order with maximum likelihood estimation and Laplace approximations to required integrals. These estimators and two existing limited information weighted least squares estimators are studied empirically. The limited information estimators compare favorably to full information estimators based on marginal maximum likelihood, MCMC, and multinomial distribution with a Laplace approximation methodology. Among the various estimators, Maydeu-Olivares and Joe's (2005) weighted least squares limited information estimators implemented with Laplace approximations for probabilities are shown in a simulation to have the best root mean square errors.

A comparison of information functions and search strategies for sensor planning in target classification

This paper investigates the comparative performance of several information-driven search strategies and decision rules using a canonical target classification problem. Five sensor models are considered: one obtained from classical estimation theory and four obtained from Bernoulli, Poisson, binomial, and mixture-of-binomial distributions. A systematic approach is presented for deriving information functions that represent the expected utility of future sensor measurements from mutual information, Rènyi divergence, Kullback-Leibler divergence, information potential, quadratic entropy, and the Cauchy-Schwarz distance. The resulting information-driven strategies are compared to direct-search, alert-confirm, task-driven (TS), and log-likelihood-ratio (LLR) search strategies. Extensive numerical simulations show that quadratic entropy typically leads to the most effective search strategy with respect to correct-classification rates. In the presence of prior information, the quadratic-entropy-driven strategy also displays the lowest rate of false alarms. However, when prior information is absent or very noisy, TS and LLR strategies achieve the lowest false-alarm rates for the Bernoulli, mixture-of-binomial, and classical sensor models.

A Semiparametric Bayesian Approach to Multivariate Longitudinal Data

We extend the standard multivariate mixed model by incorporating a smooth time effect and relaxing distributional assumptions. We propose a semiparametric Bayesian approach to multivariate longitudinal data using a mixture of Polya trees prior distribution. Usually, the distribution of random effects in a longitudinal data model is assumed to be Gaussian. However, the normality assumption may be suspect, particularly if the estimated longitudinal trajectory parameters exhibit multimodality and skewness. In this paper we propose a mixture of Polya trees prior density to address the limitations of the parametric random effects distribution. We illustrate the methodology by analyzing data from a recent HIV-AIDS study.

Does noise reduction matter for curve fitting in growth curve models?

In this paper, we discuss the efficiency of noise reduction for curve fitting in nonlinear growth curve models. We use singular spectrum analysis as a nonlinear-nonparametric denoising method. A set of longitudinal measurements is used in considering the performance of the method. We also use artificially generated data sets with and without noise for the purpose of validation of the results obtained in this study. The results show that noise reduction is important for curve fitting in growth curve models and also, that the singular spectrum analysis technique can be used as a powerful tool for noise reduction in longitudinal measurements.