• Communications for Statistical Applications and Methods
• /
• v.8 no.2
• /
• pp.311-317
• /
• 2001

It is necessary to check the linearity of selected covariates in regression diagnostics. There are various graphical methods using residual plots such as partial residual plots, augmented partial residual plots and combining conditional expectation and residual plots. In this paper, we propose the modified pseudolikelihood ratio test statistics based on these residual plots to test linearity of selected covariate. These test statistics which measure the distance between the nonparametric and parametric models are derived as a ratio of quadratic forms. The approximate distribution of these statistics is calculated numerically by using three moments. The power comparison of these statistics is given.

• Communications for Statistical Applications and Methods
• /
• v.9 no.3
• /
• pp.787-797
• /
• 2002

Partial residual plots, augmented partial residual plots and CERES plots are basic diagnostic tools for dealing with curvature as a function of specific predictors in regression problem. However, it is known that these plots can miss a curve or show a false curve in some cases such as predictors are related each other. Dynamic display of these plots is developed and applied. Examples demonstrate that dynamic plots are useful for obtaining additional Information on the curvature.

• Communications for Statistical Applications and Methods
• /
• v.11 no.3
• /
• pp.529-537
• /
• 2004

This article concerns dynamic graphical methods for visualizing a curvature in regression problem in which some predictors enter nonlinearly. A sequence of augmented partial residual plot or partial residual plot updated by the change of linear combination of two predictors are constructed. Examples demonstrate that the suggested methods can be used to reduce the dimension of explanatory variables as well as to capture a curvature.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2000.04a
• /
• pp.373-376
• /
• 2000

We investigate some properties of commonly used residual plots in linear regression and provide some systematic insight into the relationships among the plots. We discuss three issues of linear regression in this stream of context. First of all, we introduce two graphical comparison methods to display the variance inflation factor. Secondly, we show that the role of a suppressor variable in linear regression can be checked graphically. Finally, we show that several other types of standardized regression coefficients, besides the ordinary one, can be obtained in residual plots and the correlation coefficients of one of these residual plots can be used in ranking the relative importance of variables.

• Communications for Statistical Applications and Methods
• /
• v.7 no.1
• /
• pp.37-46
• /
• 2000

We investigate some properties of commonly used residual plots in linear regression and provide some systematic insight into the relationships among the plots. We discuss three issues of linear regression in this stream of context. First of all we introduce two graphical comparison methods to display the variance inflation factor. Secondly we show that the role of a suppressor variable in linear regression can be checked graphiclly. Finally we show that several other types of standardized regression coefficients besides the ordinary one can be obtained in residual plots and the correlation coefficients of one of these residual plots can be used in ranking the relative importance of variables.

• Communications for Statistical Applications and Methods
• /
• v.4 no.1
• /
• pp.119-128
• /
• 1997

Added variable plot and component-plus-residual plot are very useful for studying the role of a predictor in classical regression analysis. The former is usually used to check the effect of adding a new variable to existing model. The latter has been suggested as computationally convenient substitutes for the added variable plots, however, this plot is found to be better in detecting nonlinear relationships of a new predictor. By combining these two plots dynamically, we can take advantages of two plots simultaneously. And even further, we can get some knowledge of collinearity between a new predictor and predictors already in the model, and more accurate information about the possible outliers.

• Journal of the Korean Data and Information Science Society
• /
• v.11 no.1
• /
• pp.73-81
• /
• 2000

The variance inflation factor can be expressed by the square of the ratio of t-statistics associated with slopes of partial regression and partial residual plots. Disagreement of two sides in the interpretation can be occurred, and we analyze it with some illustrations.

• Communications for Statistical Applications and Methods
• /
• v.11 no.2
• /
• pp.345-353
• /
• 2004

Graphical methods for the specification of the curvature as a function of two predictors are animated to see the effect of an added variable to the model. Through a 3D animated plot it might be difficult to find a sequence of interpretable plots. But examples demonstrate that useful information can be obtained by using rotation technique in 3D plot. Besides 3D plots, an example of 2D animated plot applied to the case of high correlation between predictors and an added predictor is also given. It implies that speed of the convergence to a certain image in a dynamic plot may be understood as an influence of collinearity.

• 한국데이터정보과학회:학술대회논문집
• /
• 2003.10a
• /
• pp.93-100
• /
• 2003

It is necessary to check the curvature of selected covariates in regression diagnostics. There are various graphical methods using residual plots based on least squares fitting. The sensitivity of LS fitting to outliers can distort their residuals, making the identification of the unknown function difficult to impossible. In this paper, we compare combining conditional expectation and residual plots(CERES Plots) between least square fit and robust fits using Huber M-estimator. Robust CERES will be far less distorted than their LS counterparts in the presence of outliers and hence, will be more useful in identifying the unknown function.

• Communications for Statistical Applications and Methods
• /
• v.16 no.3
• /
• pp.495-500
• /
• 2009

Two diagnostic plots, added variable plot and partial residual plot, are proposed when a new explanatory variable is linearly added to constrained regressions. They are useful for investigating the effect of adding an explanatory variable to the constrained regression. They visually give an overall impression of the strength of linear relationship between response variable and added variable. A numerical example is provided for illustration.