topic Re: reading a univariate plot in SAS Procedures

reading a univariate plot

amartinez401 — Thu, 14 Mar 2013 13:43:01 GMT

I need to answer the following questions based on a proc univariate normal plot and need help asap.

Can anyone help?

a. Violation of equal mean assumption using a residual plot

b. Violation of equal variance assumption using a residual plot.

c. Violation of normality assumption

 Using the normal probability plot,

 Using the box plot

 Using a stem and leaf plot

 conduct a statistical test (Ho, Ha, TS, pvalue and conclusion)

d. Flag potential outliers looking at studentized and jackknife residuals. (State the critical value you used ) What does this mean?

e. Flag any high leverage points. (State the critical value you used) . What does this mean?

f. Look at Cook’s D values and make conclusions. State critical value.

g. Make a conclusion about the PRESS statistic.

Re: reading a univariate plot

Cynthia_sas — Sun, 17 Mar 2013 22:16:19 GMT

Hi:

Usually, folks post the code they've tried, any error messages they get and a sample of the data they need to analyze. You have only posted questions, without any data or program. Saying that the plot is a UNIVARIATE plot is useful, but without anything concrete to go on, it's hard to comment.

a,b,c: Testing the assumptions of linear regression and Does your data violate linear regression assumptions?

d http://www-stat.wharton.upenn.edu/~waterman/Teaching/701f99/Class04/class04.pdf

e http://pages.stern.nyu.edu/~churvich/Undergrad/Handouts2/31-Reg6.pdf

f http://www.ats.ucla.edu/stat/sas/webbooks/reg/chapter2/sasreg2.htm

g http://www.business.latech.edu/~jcochran/QA622/pdf%20note%20files/checking%20the%20fit%20in%20multiple%20regression.pdf

There are quite a few .EDU resources here. Google is your friend.

cynthia

Re: reading a univariate plot

Rick_SAS — Mon, 18 Mar 2013 18:48:37 GMT

These are questions about regression diagnostic plots. Submit

ODS GRAPHICS ON;

and run PROC REG. Several of these plots are created automatically.