6
views
0
recommends
+1 Recommend
1 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      ¿Cómo corregir la heterocedasticidad y autocorrelación de residuales en modelos de ahusamiento y crecimiento en altura? Translated title: How to correct the heteroscedasticity and autocorrelation of residuals in taper and height growth models?

      research-article

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Resumen: En la modelación del ahusamiento y del crecimiento en altura dominante con datos de series de tiempo, es muy común la presencia de heterocedasticidad y autocorrelación de los errores. Funciones de varianza (varFunc) y estructuras de correlación (corStruct) para corregir la heterocedasticidad y modelar dependencia de los errores, respectivamente. Estas fueron combinadas y evaluadas en ecuaciones de ahusamiento y crecimiento en altura de Pinus teocote en Durango, México. La base de datos se obtuvo de 51 análisis troncales con 768 observaciones de ahusamiento y 634 de altura. Las varFunc utilizadas fueron: 1) función de potencia (varPower); 2) función exponencial (varExp); 3) función constante y de potencia (varConstPower); y 4) función combinada de potencia y exponencial (varComb). Las corStruct incluyeron: simetría compuesta (corCompSymm), autorregresiva de orden 1 (corAR1), autorregresiva continua (corCAR1), autorregresiva de media móvil (corARMA2-0), corARMA1-1, corARMA2-1, corARMA2-2, corARMA3-1 y corARMA3-2. Las ecuaciones se ajustaron por mínimos cuadrados generalizados no lineales; y se evaluaron con un sistema de calificación con los estadísticos de ajuste: RMSE, R2, AIC, BIC, LogLik, CV y sesgo promedio. Con base en la calificación, las mejores combinaciones para el ahusamiento y crecimiento en altura fueron 1-9, 2-5, 3-8 y 4-6 y 1-6, 2-9, 3-7 y 4-4, respectivamente. En el ahusamiento solo la combinación 2-5 fue homocedástica con residuales independientes al igual que las ecuaciones de altura seleccionadas y las varFunc y corStruct presentaron influencia en la trayectoria de las curvas de índice de sitio construidas.

          Translated abstract

          Abstract: In modeling of taper functions and dominant height growth with time series data, the presence of heteroscedasticity and autocorrelation in residuals is common. Variance Functions (varFunc) and correlation structures (corStruct) were used to correct heteroscedasticity and autocorrelation; both were combined and evaluated through taper and height growth equations for Pinus teocote in Durango, Mexico. A dataset of 51 stems analysis with 768 taper observations and 634 height growth observations was used. The varFuncs applied were: 1) power function (varPower); 2) exponential function (varExp); 3) constant plus power function (varConstPower); and 4) a combination of power and exponential functions (varComb). The corStructs were: compound symmetry (corCompSymm), autoregressive of order 1 (corAR1), continuous-time autoregressive of order 1 (corCAR1), autoregressive-moving average (corARMA2-0), corARMA1-1, corARMA2-1, corARMA2-2, corARMA3-1 and corARMA3-2. To fit the equations, the generalized nonlinear least squares method was used and evaluated with a rating system through: RMSE, R2, AIC, BIC, LogLik, VC and average bias. According to the rating system, the best combinations for taper and height growth equations were 1-9, 2-5, 3-8 and 4-6 and 1-6, 2-9, 3-7 and 4-4, respectively. In the taper equation, only the combination 2-5 was homoscedastic with independent residuals, and the selected height growth equations were homoscedastic with independent residuals; the varFunc and corStruct had influence on the trajectories of site index curves.

          Related collections

          Most cited references32

          • Record: found
          • Abstract: found
          • Article: not found

          Regression analysis of spatial data.

          Many of the most interesting questions ecologists ask lead to analyses of spatial data. Yet, perhaps confused by the large number of statistical models and fitting methods available, many ecologists seem to believe this is best left to specialists. Here, we describe the issues that need consideration when analysing spatial data and illustrate these using simulation studies. Our comparative analysis involves using methods including generalized least squares, spatial filters, wavelet revised models, conditional autoregressive models and generalized additive mixed models to estimate regression coefficients from synthetic but realistic data sets, including some which violate standard regression assumptions. We assess the performance of each method using two measures and using statistical error rates for model selection. Methods that performed well included generalized least squares family of models and a Bayesian implementation of the conditional auto-regressive model. Ordinary least squares also performed adequately in the absence of model selection, but had poorly controlled Type I error rates and so did not show the improvements in performance under model selection when using the above methods. Removing large-scale spatial trends in the response led to poor performance. These are empirical results; hence extrapolation of these findings to other situations should be performed cautiously. Nevertheless, our simulation-based approach provides much stronger evidence for comparative analysis than assessments based on single or small numbers of data sets, and should be considered a necessary foundation for statements of this type in future.
            Bookmark
            • Record: found
            • Abstract: not found
            • Article: not found

            Testing for Serial Correlation in Least Squares Regression. III

              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Assumptions of Multiple Regression: Correcting Two Misconceptions

              In 2002, an article entitled "Four assumptions of multiple regression that researchers should always test" by.Osborne and Waters was published in PARE. This article has gone on to be viewed more than 275,000 times.(as of August 2013), and it is one of the first results displayed in a Google search for "regression assumptions". While Osborne and Water's efforts in raising awareness of the need to check assumptions when using regression are laudable, we note that the original article contained at least two fairly important misconceptions about the assumptions of multiple regression: Firstly, that multiple regression requires the assumption of normally distributed variables; and secondly, that measurement errors necessarily cause underestimation of simple regression coefficients. In this article, we clarify that multiple regression models estimated using ordinary least squares require the assumption of normally distributed errors in order for trustworthy inferences, at least in small samples, but not the assumption of normally distributed response or predictor variables. Secondly, we point out that regression coefficients in simple regression models will be.biased (toward zero) estimates of the relationships between variables of interest when measurement error is uncorrelated across those variables, but that when correlated measurement error is present, regression coefficients may be either upwardly or downwardly biased. We conclude with a brief corrected summary of the assumptions of multiple regression when using ordinary least squares. Accessed 36,189 times on https://pareonline.net from September 06, 2013 to December 31, 2019. For downloads from January 1, 2020 forward, please click on the PlumX Metrics link to the right.
                Bookmark

                Author and article information

                Contributors
                Role: ND
                Role: ND
                Role: ND
                Journal
                remcf
                Revista mexicana de ciencias forestales
                Rev. mex. de cienc. forestales
                Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (México, Distrito Federal, Mexico )
                2007-1132
                October 2018
                : 9
                : 49
                : 28-59
                Affiliations
                [1] orgnameInstituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias orgdiv1Campo Experimental Valle del Guadiana Mexico
                [3] orgnameUniversidad Autónoma de Nuevo León orgdiv1Facultad de Ciencias Forestales Mexico
                [2] orgnameUniversidad Autónoma de Nuevo León orgdiv1Posgrado en Ciencias Forestales Mexico
                Article
                S2007-11322018000500028
                10.29298/rmcf.v9i49.151
                6b63e7d7-c3ee-47d9-a7f8-0a60b7ee3bbe

                This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

                History
                : 10 December 2017
                : 30 July 2018
                Page count
                Figures: 0, Tables: 0, Equations: 0, References: 37, Pages: 32
                Product

                SciELO Mexico

                Categories
                Artículos

                Ahusamiento,altura dominante,estructuras de correlación,funciones de varianza,Pinus teocote Schiede ex Schltdl. & Cham.,residuales,Taper,dominant height,correlation structures,variance functions,residuals

                Comments

                Comment on this article