Linear quantile regression models for longitudinal experiments: an overview

Marino, Maria Francesca; Farcomeni, Alessio

doi:10.1007/s40300-015-0072-5

Cited by 36 publications

(35 citation statements)

References 121 publications

(156 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other references on quantile regression in the longitudinal data framework include conditional fixed effect models [8][9][10][11] and the proposal by Liu et al 12 for handling (short) longitudinal sequences of Gaussian responses subject to (possibly) non-ignorable missingness. For a general review, see Marino and Farcomeni 16 . In this paper, we will focus on random coefficients models.…”

Section: Extensions Of This Model Are Discussed By Liu and Bottaimentioning

confidence: 99%

Mixed hidden Markov quantile regression models for longitudinal data with possibly incomplete sequences

Marino

Tzavidis

Alfó

2016

Stat Methods Med Res

Self Cite

View full text Add to dashboard Cite

Quantile regression provides a detailed and robust picture of the distribution of a response variable, conditional on a set of observed covariates. Recently, it has be been extended to the analysis of longitudinal continuous outcomes using either time-constant or time-varying random parameters.However, in real-life data, we frequently observe both temporal shocks in the overall trend and individual-specific heterogeneity in model parameters. A benchmark dataset on HIV progression gives a clear example. Here, the evolution of the CD4 log counts exhibits both sudden temporal changes in the overall trend and heterogeneity in the effect of the time since seroconversion on the response dynamics. To accommodate such situations, we propose a quantile regression model where time-varying and time-constant random coefficients are jointly considered. Since observed data may be incomplete due to early drop-out, we also extend the proposed model in a pattern mixture perspective. We assess the performance of the proposals via a large scale simulation study and the analysis of the CD4 count data.

show abstract

Section: Extensions Of This Model Are Discussed By Liu and Bottaimentioning

confidence: 99%

Mixed hidden Markov quantile regression models for longitudinal data with possibly incomplete sequences

Marino

Tzavidis

Alfó

2016

Stat Methods Med Res

Self Cite

View full text Add to dashboard Cite

show abstract

“…Chen et al [39] proposed a marginally-conditional quantile regression approach to deal with the longitudinal dataset by including the random observing times, and the consistency and asymptotic normality for the estimators were developed. Other productive studies were found in Marino and Farcomeni, and Li et al [40,41]. Galarza and Lachos [37] used the stochastic approximation of the expectation-maximization (EM) algorithm to derive the maximum likelihood estimates of the fixed effects and variance component of the quantile regression for the linear mixed-effects model, and this method was constructively implemented by the R (Vienna, Austria) package qrLMM [37,42,43].…”

Section: Introductionmentioning

confidence: 99%

Using Linear Mixed-Effects Models with Quantile Regression to Simulate the Crown Profile of Planted Pinus sylvestris var. Mongolica Trees

Sun¹,

Gao²,

Li³

2017

Forests

View full text Add to dashboard Cite

Abstract:Crown profile is mostly related to the competition of individual trees in the stands, light interception, growth, and yield of trees. A total of 76 sample trees with a total number of 889 whorls and 3658 live branches were used to develop the outer crown profile model of the planted Pinus sylvestris var. mongolica trees in Heilongjiang Province, China. The power-exponential equation, modified Kozak equation, and simple polynomial equation were used and the model which showed the best performance was used as the basic model. The dummy variable approach was used to analyze the effect of stand age and stand density on the crown profile. Quantile regression for linear mixed-effects model, where the correlations between the series measurements on the same subject were considered, was used to model the outer crown profile. The results indicated that the power-exponential equation had the smallest error and was used as the basic model. Based on the dummy variable approach, stand age and stand density showed significant effects on the crown profile on the whole. Thus, they were directly included into the linear form of the power-exponential equation by a natural logarithm transformation to develop the quantile regression for the linear mixed-effects model. The 0.95th quantile regression model performed best in modeling the outer crown profile when compared to other quantiles. The prediction accuracy of the 0.95th quantile regression model by adding the random effects increased when compared to the quantile regression without random effect. The quantile regression for the linear mixed-effects model also showed an excellent performance in the largest crown radius prediction when compared to the quantile regression model. From suppressed trees to dominant trees, the crown radius increased, with tree size increasing for the same stand age and stand density increases. The crown radius of the suppressed trees from 21 to 40 year stands was the largest and the smallest was from older (>40 years) stands. The crown radius for both the intermediate and dominant trees from 21 to 40 year stands were similar and were larger than the younger (10-20 years) stands. The crown radius increased with tree size when the stand variables were constant. Furthermore, the crown radius increased with the increase of stand age, decreased with increasing stand density, and decreased with increased ratio of tree height to diameter at the breast height (HD) for trees with the same tree variables. Stand density had a weaker effect on the crown profile when compared to the HD. The growth rate of the crown radius of planted Pinus sylvestris var. mongolica trees increased with increasing stand age, and decreased with decreasing stand density. The growth rate of the crown radius decreased with increasing HD.

show abstract

“…With r considered both conditional and marginal models to deal with repeated measures in QR. 28 In the former category, we find random effects models, either linear [29][30][31][32][33] or nonlinear, 34,35 which, however, can be computationally demanding. In the marginal models category, the weighted approach 36 is perhaps the easiest to implement.…”

Section: Introductionmentioning

confidence: 99%

“…In our analytic approach, we allow for the following: (i) the exclusion of particular observation times if these are related to transitions that are not of interest; (ii) one or more absorbing states, that is, terminal states which, once reached, cannot be departed from (eg, death); (iii) states to which transitions are impossible (eg, from healthy to graft failure); and (iv) repeated measures on the same subjects who visit the same state multiple times. With r considered both conditional and marginal models to deal with repeated measures in QR . In the former category, we find random effects models, either linear or nonlinear, which, however, can be computationally demanding.…”

Section: Introductionmentioning

confidence: 99%

Multistate quantile regression models

Farcomeni

Geraci

2019

Statistics in Medicine

Self Cite

View full text Add to dashboard Cite

We develop regression methods for inference on conditional quantiles of time‐to‐transition in multistate processes. Special cases include survival, recurrent event, semicompeting, and competing risk data. We use an ad hoc representation of the underlying stochastic process, in conjunction with methods for censored quantile regression. In a simulation study, we demonstrate that the proposed approach has a superior finite sample performance over simple methods for censored quantile regression, which naively assume independence between states, and over methods for competing risks, even when the latter are applied to competing risk data settings. We apply our approach to data on hospital‐acquired infections in cirrhotic patients, showing a quantile‐dependent effect of catheterization on time to infection.

show abstract

Linear quantile regression models for longitudinal experiments: an overview

Cited by 36 publications

References 121 publications

Mixed hidden Markov quantile regression models for longitudinal data with possibly incomplete sequences

Mixed hidden Markov quantile regression models for longitudinal data with possibly incomplete sequences

Using Linear Mixed-Effects Models with Quantile Regression to Simulate the Crown Profile of Planted Pinus sylvestris var. Mongolica Trees

Multistate quantile regression models

Contact Info

Product

Resources

About