Programming the statistical procedures from SAS

Error term and Residuals

Reply
Regular Contributor
Posts: 165

Error term and Residuals

What is the difference between error term and residuals in modeling?

Trusted Advisor
Posts: 1,444

Re: Error term and Residuals

An error term exists in the model. It is a mathematical construct.

The residual is the actual value of one data point (deviation from the value predicted by the model).

Regular Contributor
Posts: 165

Re: Error term and Residuals

I understood residuals now.

How does an error term exists in the model? May I request you to give an simple example please.

Trusted Advisor
Posts: 1,444

Re: Error term and Residuals

A simple regression model looks like this

y = b0 + b1*X + e

e is the error term, b0 is the intercept and b1 is the slope

Regular Contributor
Posts: 165

Re: Error term and Residuals

Thanks.

Where that 'e' comes from?

Grand Advisor
Posts: 16,941

Re: Error term and Residuals

FYI the Stats course for SAS is free via the SAS Analytics U homepage.

From wikipedia:
http://en.wikipedia.org/wiki/Regression_analysis


Theoretical Model with e : y_i=\beta_0 +\beta_1 x_i +\varepsilon_i,\quad i=1,\dots,n.\!
Fitted model:
 \widehat{y_i} = \widehat{\beta}_0 + \widehat{\beta}_1 x_i.

The residual,  e_i = y_i - \widehat{y}_i , is the difference between the value of the dependent variable predicted by the model,  \widehat{y_i}, and the true value of the dependent variable, y_i.

Respected Advisor
Posts: 2,655

Re: Error term and Residuals

I would recommend reading some good introductory regression and linear modeling texts.  You may find this one to be helpful:

Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis, by Frank Harrell, Jr.

Steve Denham

Ask a Question
Discussion stats
  • 6 replies
  • 306 views
  • 4 likes
  • 4 in conversation