Crikey, catch up! I've mentioned the research numerous times. Have a jolly read... http://egov.ufsc.br/portal/sites/default/files/anexos/32969-41234-1-PB.pdf
I asked you. You haven’t explained it, nor explain it’s use. Like all the evidence (other studies that agree with) you refer to, you merely refer to the critique of others. Can the use of dummy variables in regression analysis have a functional roll? If so, why? If not why? By the way, I am familiar with the article you posted, long before this exchange. I don’t accept any article as gospel without following lines of criticisms.
Yep and I gave you a whole paper to show it. No need to thank me! Dummy vriables are used as a matter of fact in regression. This is about a empirically flawed methodology. Read the paper and then get back to me with a real question
Meaning that the matter is not properly understood or comprehended on the part of yourself, otherwise elaboration would be possible on the part of yourself, without the need to cite a paper as the default answer. There has been no effort presented on the part of yourself to break down the findings of the paper, and explain how it proved the so-called "dummy variable" by John Lott, what it amounts to, and how it changes the entire nature of his findings and renders them moot as a result.
Most people do not understand very complex problems with multiple relationships, gaps, voids, fuzzy areas and sub sets. There are unknowns or X - factors, and these receive substituted imaginary or random values otherwise known as "dummy variables" and this is absolutely necessary in complex algorithms and just plain anywhere complex problems exist. These dummy variables aid in construction a working armature or platform, and in computer constructs of any proposed design, are eventually replaced by actual proved numbers. Without dummy variables, it would be impossible to proceed with most medical, engineering and high technology projects.
Crikey, what whinge! It's not difficult. Dummy variables are a crucial part of the regression methodology. They are of course vital for aspects such as controlling for socio-economic variable. However, the authors show that Lott introduces empirical bias as the impact of concealed weapons is predicted to differ across counties. Referring to the paper is perfectly correct. This isn't just acknowledgement of the empirical difficulties created by Lott's simplistic approach. It's also about acknowledging that, even though the data used is the same, elimination of the empirical bias substantially changes the estimates (including the sign, destroying Lott's conclusions) Thus made me laugh. I used this example to show why your anti-education approach is feeble. That you must compare and contrast across papers and that, by adopting literature review methods, you will achieve genuine critique. There's a delicious irony that you're moaning that I haven't referred enough to the paper. Cheers!
The above does not actually refute the points being raised in criticism of the argument being presented by yourself. All is does is ultimately admit that it is not possible for yourself to explain what the paper actually says, or how it demonstrates that John Lott does not know what he is talking about. It merely demonstrates that it is yourself who does not know what they are talking about.
Keep up! The dummy variable approach could only be credible if the concealed weapon effect does not vary. The authors show that isn't the case. Essentially Lott assumes, via a dummy variable, a linear shift in the crime relationship. That is cobblers. The authors show the folly theoretically and empirically. Their theoretical analysis builds on deterrence analysis from the likes of Ehrlich. It confirms that the hypothesis test is dependent on a non-linear complex relationship. To further demonstrate the stupidity of assuming otherwise, it replaces the dummy variable approach with a more flexible empirical methodology. It confirms that the impact of empirical bias determines Lott's findings.
