Lab 11 Exercises - Multicollinearity

Lab 11 Exercises - Multicollinearity - Lab 11:...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Lab 11: Multicollinearity Objectives: In today’s lab we will investigate the existence of a problem in the sample data - Multicollinearity . Multcollinearity is the existence of linear association among two or more independent variables. The mere existence of multicollinearity does not mean you have a problem; you only have a problem when the degree of multicollinearity is high. The exercises below illustrate diagnosing the problem, and applying a couple of possible remedies. Data : The data for today’s lab are those that we used last week, the Roses sales data, and the data in the Minitab worksheet Chickendata.mtw . You should have a Minitab project file for the Roses data. Multicollinearity – diagnosis of the problem in the roses data. 1. Check for multicollinearity in the model we estimated for the sales of roses. These are time series data collected quarterly. We always suspect times-series data of multicollinearity because common inflationary trends are often part of price and income data. Thus, the variables prose, pcarn and dinc may be highly correlated. In addition, we created the variables D2 and ProseD2 . These may also have strong correlation. Open your Minitab roses project from last week and let’s check for multicollinearity problems. 2. Estimate the following model and ask Minitab to calculate the variance inflation factors (VIFs). You’ll find a check box for Variance Inflation Factors in the Options menu of Regression . 01 2 3 + t Sales + Prose Pcarn Disinc + D2 ProseD2 + u β βββδ γ = ; 3. First, review your regression results. Look for the “tell-tale signs” of multicollinearity. ± What is the R 2 for the model? Is it high suggesting a good model? ± What is the calculated F-statistic for the model? Is it large suggesting statistical significance? ± These two measures indicate how well the independent variables explained the dependent variable. A high R 2 and a large F-statistic indicate the model did well explaining the dependent variable.
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 12/08/2011 for the course ECON 312 taught by Professor Daniellass during the Winter '10 term at UMass (Amherst).

Page1 / 3

Lab 11 Exercises - Multicollinearity - Lab 11:...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online