misclassification in binary variables
Misclassification of binary variables is the first case of non-classical measurement error considered. Similar to the classical errors-in-variables result, misclassification of a binary regressor leads to attenuation of slope coefficient estimates in linear regression. Classical instrumental variables will not address the problem. Bounds results under a number of different sets of assumptions can be derived. When the dependent variable is binary, misclassification also leads to slope attenuation. Some identification results are available in this case.
|This chapter was published in: Steven N. Durlauf & Lawrence E. Blume (ed.) , , pages , 2010, 3rd quarter update.|
|This item is provided by Palgrave Macmillan in its series The New Palgrave Dictionary of Economics with number v:4:year:2010:doi:3832.|
|Contact details of provider:|| Web page: http://www.palgrave-journals.com/|
|Order Information:|| Web: http://www.dictionaryofeconomics.com/help/faq#_Toc198623697 Email: |
When requesting a correction, please mention this item's handle: RePEc:pal:dofeco:v:4:year:2010:doi:3832. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Sheeja Sanoj)
If references are entirely missing, you can add them using this form.