Saturday, May 3, 2014

Nate Silver Uses Stata, Doesn't Care About Getting Standard Errors Right

Both propositions seemingly demonstrated convincingly here

Even putting aside heteroskedasticity, with 29 observations, I think you have to at least type ", small" at the end of that command line, don't you?  

 

http://lawprofessors.typepad.com/law_econ/2014/05/nate-silver-uses-stata-doesnt-care-about-getting-standard-errors-right.html

| Permalink

TrackBack URL for this entry:

http://www.typepad.com/services/trackback/6a00d8341bfae553ef01a73dbb19be970d

Listed below are links to weblogs that reference Nate Silver Uses Stata, Doesn't Care About Getting Standard Errors Right:

Comments

I get stuck even before getting to the regression results. Everything depends on how Forbes is doing its valuation. Is Forbes looking at the value of the team in terms of the value of its present cash flow, or in terms of predicted cash flow based on city growth, or the value it would sell for on the market? Since sports teams are fun to hold regardless of return, their market value doesn't necessarily correspond to the value of their profits. Thus the importance of the number of local billionaires. And I don't know what his rate-of-return variable means, as a result. If it's about market value, and markets are efficient, then past rates of return are no guide to future rates.

Posted by: Eric Rasmusen | May 6, 2014 8:05:40 AM

Post a comment