Sunday, January 27, 2008

More NORC fisking...

To further strengthen the argument that the NORC report is biased in it's conclusions on firearms, I did even more research into the data sets and variables used. Just to make sure. Jadetroll, in his usual SOP, claimed I didn't even read the report (even though I quoted and sourced it) and that they "controlled" for the differences in population for their report. When asked for these variables they used, he did his usual avoidances. Why would he do that you ask? Because of pages listing this on the GSS:

Dataset: General Social Surveys, 1972-2006 [Cumulative File]
Regression analysis:
Dependent variable:
No dependent variable added
Independent variable:
No independent variable added

Listwise correlation analysis: [Change to pairwise]

No variables added

Double checking the data sets here and here, the report percentages are based off of straight response numbers. No controlling for population or variables. So their claim that "

"As Figure 2 shows gun ownership has been declining over the last 35 years" is disingenuous at best and shows that there is reason to suspect this packet as politically biased as they base their conclusions off of other biased surveys.

No comments: