Correction to: Testing for Threshold Effects in Regression Models (Journal of the American Statistical Association, (2013), 106, (220-231), 10.1198/jasa.2011.tm09800)

Publication Type:
Journal Article
Journal of the American Statistical Association, 2017, 112 (518), pp. 883 - ?
Issue Date:
Filename Description Size
Correction.pdfPublished Version496.07 kB
Adobe PDF
Full metadata record
© 2017 American Statistical Association. As an empirical illustration, Lee, Seo, and Shin (2011) used a subsample of the dataset originally constructed by Card, Mas, and Rothstein (2008) to test whether there existed a tipping point for racial segregation. When using the same dataset for different purposes, we found that there was an indexing error in the coding. In addition, there was a minor typographical error in the article. On page 224, the average marginal effect should have been defined as (Formula presented.), where Ф(·) is the CDF of the normal distribution. Tables 1 and 2 replace the old tables on page 224 and provide corrected empirical results. The corrected results indicate that there exists a tipping point in Chicago but not in other cities. The estimated tipping point for Chicago is similar between probit and median regression models. (Table presented.).
Please use this identifier to cite or link to this item: