Here's an example I use in the summer ESTIMATE course at MSU. It's based on an actual contingent valuation survey. There are two prices, one of regular apples the other of "ecologically friendly" apples. The prices were randomly assigned as a pair, (PR, PE).
#metricstotheface
#metricstotheface
https://twitter.com/jmwooldridge/status/1483493723233259527
Individuals were then asked to choose a basket of regular and eco-friendly applies. A linear regression for QE (quantity of eco-labeled) gives very good results: strong downward sloping demand curve, an increase in the competing price shifts out the demand curve.
Now, the prices were generated to be highly correlated with, corr = 0.83. Not VIF > 10 territory but a pretty high correlation. If PR is dropped from the equation for QE, the estimated price effect for PE falls dramatically -- because there's an important omitted variable, PR.
When I ask at ESTIMATE, many have said PR should be dropped because of collinearity with PE -- clearly the wrong answer. PR must be included precisely because it is correlated with PE.
The data set is APPLE.RAW (or APPLE.DTA) with my intro book.
The data set is APPLE.RAW (or APPLE.DTA) with my intro book.
Looks like ESTIMATE will be June 10-13 in 2022. :-D
• • •
Missing some Tweet in this thread? You can try to
force a refresh
