Overview for Orthogonal Regression

Use Orthogonal Regression, also known as Deming regression, to determine whether two instruments or methods provide comparable measurements. Orthogonal regression examines the linear relationship between two continuous variables: one response (Y) and one predictor (X). Unlike simple linear regression (least squares regression), both the response and predictor in orthogonal regression contain measurement error. In simple regression, only the response variable contains measurement error. If you use simple regression to determine comparability when both variables contain measurement error, the results depend on which variable the calculations assume have no measurement error. Orthogonal regression addresses this problem so the roles of the variables have little influence on the results.

For example, an engineer at a medical device company wants to determine whether the company's new blood pressure monitor is equivalent to a similar model produced by another company.

Where to find this analysis

To perform orthogonal regression, choose Stat > Regression > Orthogonal Regression.

When to use an alternate analysis

If you have one continuous predictor but it does not contain measurement error, use Fitted Line Plot.

By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy