Assessing relationships between genetic evaluations using robust regression with an application to holsteins in Uruguay

Globalization of dairy cattle breeding has created a need for international sire proofs. Some early methods for converting proofs from one population to another are based on simple linear regression. An alternative robust regression method based on the t-distribution is presented, and maximum likelihood and Bayesian techniques for analysis are described, including the situation in which some proofs are missing. Procedures were used to investigate the relationship between Holstein sire proofs obtained by two Uruguayan genetic evaluation programs. The results suggest that conversion equations developed from data including only sires having proofs in both populations can lead to distorted results, relative to estimates obtained using techniques for incomplete data. There was evidence of non-normality of regression residuals, which constitutes an additional source of bias. A robust estimator may not solve all problems, but can provide simple conversion equations that are less sensitive to outlying proofs and to departures from assumptions.

Determinantes da evolução do peso e altura em crianças de 3 meses a 6 anos assistidas em creche: Análise por modelo linear não hierarquizado em ensaio quase-experimental

Objective. To assess factors determining growth in a group of children between 3 months and 6 years old enrolled in a public municipal (i.e., government-supported, not private) day-care center, in comparison to a group of children with similar characteristics but who were not enrolled in the center. Methods. A quasi-experimental study was designed to observe 444 children aged 3 to 72 months from a low-income neighborhood in the city of Sorocaba, in the state of São Paulo, Brazil. Two groups were studied: 164 children enrolled in a local municipal day-care center (intervention group) and 280 not receiving care at the center (nonintervention, comparison group) but instead being cared for at home. Both groups were seen four times over a period of 16 months. At each observation session, the children's weight and height were measured. Information was also collected on the mother's sociodemographic characteristics and the illnesses she had suffered as well as the child's weight and other health characteristics at birth, the child's illnesses in the 15 days before each observation, and any hospitalizations. Results. The children in both groups were from low-income families, with 65% of the families having an average monthly income below US$ 100; 80% of the mothers had received 8 years of schooling or less. Multivariate linear regression analysis showed that at the first observation (just before enrollment in the day-care center), birth weight was the only factor that explained the nutritional differences between the two groups. Subsequent analyses showed that being in day care was the factor that best explained the differences between the groups, especially in terms of the adequacy of weight for age, after controlling for birthweight, sex, age at the beginning of the study, and illnesses in the 15 days before an observation session. The nutritional impact of the intervention was significant as early as 3 months after being enrolled in day care. Conclusions. The nutritional benefits of the care provided at the center outweighed the negative effects sometimes seen in such centers, such as the greater morbidity that children in day-care centers often experience in comparison to children receiving care at home.

Uso de regressão linear para estimar parâmetros físico-químicos relacionados à qualidade do jerked beef

Jerked beef is officially defined as salted, cured and dried beef. Water activity (Aw), moisture, ash and residual nitrite are the physicochemical parameters that define this product identity and quality standards. In this work, the behavior of these parameters during jerked beef processing was evaluated and a significant correlation among them was revealed. These results allowed the establishment of statistical equations that enable the estimation of all the physicochemical parameters from the results obtained in the measurement of just one of them.

Correlação linear e espacial da produtividade de forragem de milho com atributos físicos de um latossolo vermelho distrófico

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Comparação entre os estimadores de mínimos quadrados ordinários e mínimos desvios absolutos em modelos de regressão linear simples -: uma aplicação na energia na agricultura

Pós-graduação em Agronomia (Energia na Agricultura) - FCA

Modelagem da produção acumulada de biogás em biodigestores tipo batelada segundo a porcentagem de inóculo adicionada utilizando os modelos de regressão não-linear de Gompertz e expoencial

Pós-graduação em Agronomia (Energia na Agricultura) - FCA

Non-destructive linear model for leaf area estimation in Vernonia ferruginea Less

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Linear and undulating periodized strength plus aerobic training promote similar benefits and lead to improvement of insulin resistance on obese adolescents

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

GIS-based analytical tools for transport planning: spatial regression models for transportation demand forecast

Considering the importance of spatial issues in transport planning, the main objective of this study was to analyze the results obtained from different approaches of spatial regression models. In the case of spatial autocorrelation, spatial dependence patterns should be incorporated in the models, since that dependence may affect the predictive power of these models. The results obtained with the spatial regression models were also compared with the results of a multiple linear regression model that is typically used in trips generation estimations. The findings support the hypothesis that the inclusion of spatial effects in regression models is important, since the best results were obtained with alternative models (spatial regression models or the ones with spatial variables included). This was observed in a case study carried out in the city of Porto Alegre, in the state of Rio Grande do Sul, Brazil, in the stages of specification and calibration of the models, with two distinct datasets.

Bartlett corrections in Birnbaum-Saunders nonlinear regression models

Lemonte and Cordeiro [Birnbaum-Saunders nonlinear regression models, Comput. Stat. Data Anal. 53 (2009), pp. 4441-4452] introduced a class of Birnbaum-Saunders (BS) nonlinear regression models potentially useful in lifetime data analysis. We give a general matrix Bartlett correction formula to improve the likelihood ratio (LR) tests in these models. The formula is simple enough to be used analytically to obtain several closed-form expressions in special cases. Our results generalize those in Lemonte et al. [Improved likelihood inference in Birnbaum-Saunders regressions, Comput. Stat. DataAnal. 54 (2010), pp. 1307-1316], which hold only for the BS linear regression models. We consider Monte Carlo simulations to show that the corrected tests work better than the usual LR tests.

Influência da idade no comportamento da frequência cardíaca na transição repouso-exercício: uma análise por deltas e regressão linear

INTRODUÇÃO: As modificações da frequência cardíaca (FC) durante a transição repouso-exercício podem ser caracterizadas por meio da aplicação de cálculos matemáticos simples, como: deltas 0-10 e 0-30s para inferir sobre o sistema nervoso parassimpático, e delta e regressão linear aplicados no intervalo 60-240s para inferir sobre o sistema nervoso simpático. Assim, o objetivo deste estudo foi testar a hipótese de que indivíduos jovens e de meia-idade apresentam diferentes respostas da FC em exercício de intensidade moderada e intensa, com diferentes cálculos matemáticos. MÉTODOS: Homens aparentemente saudáveis, sendo sete de meia-idade e 10 jovens, foram submetidos a testes de carga constante de intensidade moderada e intensa. Foram calculados os deltas da FC nos períodos de 0-10s, 0-30s e 60-240s e a regressão linear simples no período de 60 a 240s. Os parâmetros obtidos na análise de regressão linear simples foram: intercepto e inclinação angular. Utilizou-se o teste Shapiro-Wilk para verificar a distribuição dos dados e o teste t não pareado para comparação entre os grupos. O nível de significância estatística considerado foi 5%. RESULTADOS: O valor do intercepto e do delta 0-10s foi menor no grupo meia-idade nas duas cargas e a inclinação do ângular foi menor no grupo meia-idade no exercício moderado. CONCLUSÃO: Os indivíduos jovens apresentam retirada vagal de maior magnitude no estágio inicial da resposta da FC durante exercício dinâmico em carga constante nas intensidades analisadas e maior velocidade de ajuste da resposta simpática em exercícios moderados.

ROBREG: Stata module providing robust regression estimators

robreg provides a number of robust estimators for linear regression models. Among them are the high breakdown-point and high efficiency MM-estimator, the Huber and bisquare M-estimator, and the S-estimator, each supporting classic or robust standard errors. Furthermore, basic versions of the LMS/LQS (least median of squares) and LTS (least trimmed squares) estimators are provided. Note that the moremata package, also available from SSC, is required.

Modeling absolute differences in life expectancy with a censored skew-normal regression approach.

Parameter estimates from commonly used multivariable parametric survival regression models do not directly quantify differences in years of life expectancy. Gaussian linear regression models give results in terms of absolute mean differences, but are not appropriate in modeling life expectancy, because in many situations time to death has a negative skewed distribution. A regression approach using a skew-normal distribution would be an alternative to parametric survival models in the modeling of life expectancy, because parameter estimates can be interpreted in terms of survival time differences while allowing for skewness of the distribution. In this paper we show how to use the skew-normal regression so that censored and left-truncated observations are accounted for. With this we model differences in life expectancy using data from the Swiss National Cohort Study and from official life expectancy estimates and compare the results with those derived from commonly used survival regression models. We conclude that a censored skew-normal survival regression approach for left-truncated observations can be used to model differences in life expectancy across covariates of interest.

A novel method for evaluating an interaction effect of correlated continuous covariates in a linear model

Interaction effect is an important scientific interest for many areas of research. Common approach for investigating the interaction effect of two continuous covariates on a response variable is through a cross-product term in multiple linear regression. In epidemiological studies, the two-way analysis of variance (ANOVA) type of method has also been utilized to examine the interaction effect by replacing the continuous covariates with their discretized levels. However, the implications of model assumptions of either approach have not been examined and the statistical validation has only focused on the general method, not specifically for the interaction effect.^ In this dissertation, we investigated the validity of both approaches based on the mathematical assumptions for non-skewed data. We showed that linear regression may not be an appropriate model when the interaction effect exists because it implies a highly skewed distribution for the response variable. We also showed that the normality and constant variance assumptions required by ANOVA are not satisfied in the model where the continuous covariates are replaced with their discretized levels. Therefore, naïve application of ANOVA method may lead to an incorrect conclusion. ^ Given the problems identified above, we proposed a novel method modifying from the traditional ANOVA approach to rigorously evaluate the interaction effect. The analytical expression of the interaction effect was derived based on the conditional distribution of the response variable given the discretized continuous covariates. A testing procedure that combines the p-values from each level of the discretized covariates was developed to test the overall significance of the interaction effect. According to the simulation study, the proposed method is more powerful then the least squares regression and the ANOVA method in detecting the interaction effect when data comes from a trivariate normal distribution. The proposed method was applied to a dataset from the National Institute of Neurological Disorders and Stroke (NINDS) tissue plasminogen activator (t-PA) stroke trial, and baseline age-by-weight interaction effect was found significant in predicting the change from baseline in NIHSS at Month-3 among patients received t-PA therapy.^

Selection bias and the cross-validation of regression models for prediction

Strategies are compared for the development of a linear regression model with stochastic (multivariate normal) regressor variables and the subsequent assessment of its predictive ability. Bias and mean squared error of four estimators of predictive performance are evaluated in simulated samples of 32 population correlation matrices. Models including all of the available predictors are compared with those obtained using selected subsets. The subset selection procedures investigated include two stopping rules, C$\sb{\rm p}$ and S$\sb{\rm p}$, each combined with an 'all possible subsets' or 'forward selection' of variables. The estimators of performance utilized include parametric (MSEP$\sb{\rm m}$) and non-parametric (PRESS) assessments in the entire sample, and two data splitting estimates restricted to a random or balanced (Snee's DUPLEX) 'validation' half sample. The simulations were performed as a designed experiment, with population correlation matrices representing a broad range of data structures.^ The techniques examined for subset selection do not generally result in improved predictions relative to the full model. Approaches using 'forward selection' result in slightly smaller prediction errors and less biased estimators of predictive accuracy than 'all possible subsets' approaches but no differences are detected between the performances of C$\sb{\rm p}$ and S$\sb{\rm p}$. In every case, prediction errors of models obtained by subset selection in either of the half splits exceed those obtained using all predictors and the entire sample.^ Only the random split estimator is conditionally (on $\\beta$) unbiased, however MSEP$\sb{\rm m}$ is unbiased on average and PRESS is nearly so in unselected (fixed form) models. When subset selection techniques are used, MSEP$\sb{\rm m}$ and PRESS always underestimate prediction errors, by as much as 27 percent (on average) in small samples. Despite their bias, the mean squared errors (MSE) of these estimators are at least 30 percent less than that of the unbiased random split estimator. The DUPLEX split estimator suffers from large MSE as well as bias, and seems of little value within the context of stochastic regressor variables.^ To maximize predictive accuracy while retaining a reliable estimate of that accuracy, it is recommended that the entire sample be used for model development, and a leave-one-out statistic (e.g. PRESS) be used for assessment. ^