Using Adaptive Tests for the Analysis of Repeated Measurements

Tw, O'Gorman

doi:10.1080/10543400802051859

Cited by 4 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The random parameters in Equation 20were predicted using the best linear unbiased predictor (EBLUP) [46,71]…”

Section: Random Parameters Predictionmentioning

confidence: 99%

“…They contain both fixed effects parameters and random effects parameters. Fixed effects parameters account for covariate or treatment effects as in traditional regression, while random effects parameters explain the different sources of stochastic variability [46][47][48]. Mixed-effects models are therefore extensively used in forestry, such as diameter-height models [49,50], crown models [51,52], self-thinning models [53][54][55], and growth models [56,57].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Development of a Mixed-Effects Individual-Tree Basal Area Increment Model for Oaks (Quercus spp.) Considering Forest Structural Diversity

Wang

Zeng

Wang

et al. 2019

Forests

View full text Add to dashboard Cite

In the context of uneven-aged mixed-species forest management, an individual-tree basal area increment model considering forest structural diversity was developed for oaks (Quercus spp.) using data collected from 11,860 observations in 845 sample plots from the 7th (2004), 8th (2009), and 9th (2014) Chinese National Forest Inventory in Hunan Province, south-central China. Since the data was longitudinal and had a nested structure, we used a linear mixed-effects approach to construct the model. We also used the variance function and an autocorrelation structure to describe within-plot heteroscedasticity and autocorrelation. Finally, the optimal mixed-effects model was determined based on the Akaike information criterion (AIC), Bayesian information criterion (BIC), log-likelihood (Loglik) and the likelihood ratio test (LRT). The results indicate that the reciprocal transformation of initial diameter at breast height (1/DBH), relative density index (RD), number of trees per hectare (NT), elevation (EL) and Gini coefficient (GC) had a significant impact on the individual-tree basal area increment. In comparison to the basic model developed using least absolute shrinkage and selection operator (LASSO) regression, the mixed-effects model performance was greatly improved. In addition, we observed that the heteroscedasticity was successfully removed by the exponent function and autocorrelation was significantly corrected by AR(1). Our final model also indicated that forest structural diversity significantly affected tree growth and hence should not be neglected. We hope that our final model will contribute to the scientific management of oak-dominated forests.

show abstract

“…The random parameters in Equation 20were predicted using the best linear unbiased predictor (EBLUP) [46,71]…”

Section: Random Parameters Predictionmentioning

confidence: 99%