Residual plots, Advanced Statistics

Residual plots are the plots of some type of residual which might be helpful in assessing the assumption made by the fitted model. In regression analysis there are various methods of plotting the residual values which can be helpful in assessing particular components of the regression model. The most useful plots when diagnosing the linear regression models are given below;

* A boxplot or probability plot of the residuals can be useful in checking for symmetry and specifically for normality of the error terms in the regression model.

* Plotting the residuals against the corresponding values of the explanatory variable.

Any sign of curvature in the plot might suggest that say a quadratic term in the explanatory variable should be included in the model.

* Plotting residuals against fitted values of the response variable. If the variability of the residuals appears to increase with the increase in the size of the fitted values, a transformation of the response variable former to fitting is indicated.

Figure shows some idealized residual plots that indicate particular points about models.

* Figure (a) is what is looked for to confirm that the fitted model is appropriate,

* Figure (b) suggests that the assumption of constant variance is not justified so that some transformation of the response variable before fitting might be sensible,

* Figure (c) implies that the model requires a quadratic term in the explanantory variable.

 

Posted Date: 8/1/2012 12:46:34 AM | Location : United States







Related Discussions:- Residual plots, Assignment Help, Ask Question on Residual plots, Get Answer, Expert's Help, Residual plots Discussions

Write discussion on Residual plots
Your posts are moderated
Related Questions
Normal 0 false false false EN-US X-NONE X-NONE

Normal approximation : Normal distributions which approximate other distributions; such as, a normal distribution with the mean np and variance np(1 - p) which acts as an approxima

Chebyshev's inequality: A statement about the proportion of the observations which fall within some number of the standard deviations of the mean for any of the probability distri


Designs which permits two or more questions to be addressed in the investigation. The easiest factorial design is one in which each of the two treatments or interventions are p

Lagging indicators: The part of a collection of the economic time series designed to give information about the broad swings in measures of the aggregate economic activity known a

A study not involving the passing of time. All information is collected at the same time and subjects are contacted only once. Many surveys are of this type. The temporal sequence

Quantile regression is an extension of the classical least squares from estimation of the conditional mean models to the estimation of the variety of models for many conditional q

Compound symmetry : The property possessed by the variance-covariance matrix of the set of multivariate data when its chief diagonal elements are equal to each other, and in additi

The probability distribution, f (x), of largest extreme can be given as    The location parameter, α is the mode and β is the scale parameter. The mean, variance skewn