0 Bewertungen0% fanden dieses Dokument nützlich (0 Abstimmungen)
112 Ansichten6 Seiten
This document discusses normal probability plots and their use in assessing the normality of residuals in simple linear regression. It provides examples of normal probability plots that correspond to:
1) Normally distributed residuals that are approximately linear, suggesting normal errors.
2) Normally distributed residuals with one outlier, appearing linear except for one point.
3) Positively skewed residuals, appearing nonlinear and suggesting non-normal errors.
4) Residuals with heavy tails, appearing nonlinear and suggesting non-normal errors.
This document discusses normal probability plots and their use in assessing the normality of residuals in simple linear regression. It provides examples of normal probability plots that correspond to:
1) Normally distributed residuals that are approximately linear, suggesting normal errors.
2) Normally distributed residuals with one outlier, appearing linear except for one point.
3) Positively skewed residuals, appearing nonlinear and suggesting non-normal errors.
4) Residuals with heavy tails, appearing nonlinear and suggesting non-normal errors.
This document discusses normal probability plots and their use in assessing the normality of residuals in simple linear regression. It provides examples of normal probability plots that correspond to:
1) Normally distributed residuals that are approximately linear, suggesting normal errors.
2) Normally distributed residuals with one outlier, appearing linear except for one point.
3) Positively skewed residuals, appearing nonlinear and suggesting non-normal errors.
4) Residuals with heavy tails, appearing nonlinear and suggesting non-normal errors.
Here are the characteristics of a well-behaved residual vs.
fits plot and what they suggest about
the appropriateness of the simple linear regression model: The residuals "bounce randomly" around the 0 line. This suggests that the assumption that the relationship is linear is reasonable. The residuals roughly form a "horizontal band" around the 0 line. This suggests that the variances of the error terms are equal. No one residual "stands out" from the basic random pattern of residuals. This suggests that there
Normal Probability Plot
Note that the relationship between the theoretical percentiles and the sample percentiles is approximately linear. Therefore, the normal probability plot of the residuals suggests that the error terms are indeed normally distributed. We did what seems like an awful lot of work to create our normal probability plot of residuals. That's because we effectively created it "by hand." The good news is that most statistical software, including Minitab, will create the normal probability plot of the residuals for us. In fact, Minitab offers three different versions of the plots. The first one looks just like the above plot. The second version of the plot:
draws a line (with a confidence band) through the points to help you decide if the relationship is linear. And, the third version of the plot the one we will probably use the most often looks like:
The advantage of this version of the plot is that it includes (in the lower right hand corner) the P- value for the Ryan-Joiner test: for testing the null hypothesis H0: The error terms are normally distributed against the alternative hypothesis HA: The error terms are not normally distributed. The Ryan-Joiner test basically assesses the correlation (r = 0.9811) between the sample percentiles and the theoretical normal percentiles. Since the P-value is large (P > 0.10), we fail to reject the null hypothesis. There is insufficient evidence to conclude that the error terms are not normally distributed. We can proceed assuming that the error terms are normally distributed. Let's take a look at examples of the different kinds of normal probability plots we can obtain and learn what each tells us. Normally distributed residuals The following histogram of residuals suggests that the residuals (and hence the error terms) are normally distributed:
The normal probability plot of the residuals is approximately linear supporting the condition that the error terms are normally distributed.
Normal residuals but with one outlier The following histogram of residuals suggests that the residuals (and hence the error terms) are normally distributed. But, there is one extreme outlier (with a value of 10):
Here's the corresponding normal probability plot of the residuals:
This is a classic example of what a normal probability plot looks like when the residuals are normally distributed, but there is just one outlier. The relationship is approximately linear with the exception of the one data point. We could proceed with the assumption that the error terms are normally distributed upon removing the outlier from the data set. Skewed (positive) residuals The following histogram of residuals suggests that the residuals (and hence the error terms) are not normally distributed. On the contrary, the distribution of the residuals is quite skewed (to the right).
Here's the corresponding normal probability plot of the residuals:
This is a classic example of what a normal probability plot looks like when the residuals are positively skewed (to the right). Clearly, the condition that the error terms are normally distributed is not met. Heavy-tailed residuals The following histogram of residuals suggests that the residuals (and hence the error terms) are not normally distributed. There are too many extreme positive and negative residuals. We say the distribution is "heavy tailed."
Here's the corresponding normal probability plot of the residuals:
The relationship between the sample percentiles and theoretical percentiles is not linear. Again, the condition that the error terms are normally distributed is not met.