2 ms·
Here's a dirty little secret … and I know this is going to upset many of the statist types … but vast majority of those OECD types of datasets are so utterly po
by frankfrankfrank 5y ago
Here's a dirty little secret … and I know this is going to upset many of the statist types … but vast majority of those OECD types of datasets are so utterly polluted that they are essentially useless beyond the broadest of strokes. Just alone the definition and the application of that definition and the capacity to accurately capture the data alone produces excessive inaccuracies that in essence rely on the formality of the process and the institutional perception as a means to claim or even just project assumed authority and accuracy of the data.
To put it in a different context; the self-delusional con job that was the lie about how wonderful and amazing the Afghan national army was for the last about 10 years minimum, is a perfect example of just how broad, deep, and long a blatant fraud can be perpetuated in such an open and overt manner, all underpinned by all kinds of shiny presentations and back rubbing soliloquies.
It's all fake. I know that makes people's cognitive dissonance flare up, but that's the uncomfortable truth. The emperor in fact has no clothes on, nor do any of his courtiers. We will see how long the charade can last.
- phdelightful 5y agoIf your point is "lies, damn lies, and statistics," then I guess I can't argue with that. If your point is something more specific about OECD self-employment data, or the processes and practices for a specific or category of "OECD types of datasets" or general practices of society-level datasets...I encourage you to make that point, and give others something to work with.
- ElemenoPicuares 5y agoDirtier and Secreter: People's cognitive dissonance would flare way up if you shared convincing evidence that the OECD or the similar bodies fake their self-employment data.