Slashdot Mirror
OOXML Vote and the CPI Corruption Index
Tapani Tarvainen writes "It turns out there's an interesting correlation between Transparency International's 'corruption perceptions index' and voting behavior in ISO's OOXML decision. Countries with a lower score (more corruption) on the 2006 CPI were more likely to vote in favor of OOXML, and those with a higher score were less likely. According to the analysis, 'This statistics supports with a P value of 0.07328 the hypothesis that the corrupted countries were more likely to vote for approval (one-tailed Fisher's Exact test). In other words, simplified a bit: the likelihood that there was no positive correlation between the corruption level and probability of an approval vote, that is, this is just a random effect, is about 7%.' Of course, correlation doesn't prove causality."
I've put up a little website with some initial thoughts, and I'd appreciate feedback from the slashdot community please.
Countries like India and China ... home to over 33% of the world's population - have voted "NO". Countires like Cote' de Ivorie and Cyprus have equal voting rights.
This population index anomaly must be rectified, before the ISO can regain any credibility as an International standards organisation.
If you keep throwing chairs, one day you'll break windows....
Most research in the social sciences considers the threshold for statistical significance below .05. Since this is above, few would have confidence that this result is not random chance. So, reporting this result, while informative if you are trained in statistics, is likely misleading to the average reader...
Given the inappropriate use of the Fisher's test, questionable use of a one-tailed model and p > 0.05, I'd start with worrying about having proven correlation.
What I'm listening to now on Pandora...
Another intersting point coming out of all this the role of ECMA International
formerly European Computer Manufactureres Association - dont see many of them around these days).
ECMA is fully accredited by ISO and in ists search for a new role as a standards body did
a nice job producing a standard for the orphaned Javascript ( except for changing the name
to the disease like ECMAscript).
However since then other "standards" developed by ECMA have been:-
-- the programming language C# ( C "sharp")
-- a Common Language Infrastructure (CLI)
-- a CLI binding for C++
-- Office Open XMLOffice
Anyone spot a pattern here?
The other problem is that ECMAInternational is essentialy a club of computer software and
hardware manufacturers and unlike national standards organisations (ASA, BS, DN etc.)
does not have any public interest mandate; it exists only to serve its members and
to join you need to be a large software or hardware manufacturer.
I have no problem with any industry forming a club to standardise things among themselves
but for an industry association to be the main sponser of an ISO standard seems plain
wrong.
Microsoft for one seems to have spotted an ideal vehicle for turning proprietary products into standards.
Old COBOL programmers never die. They just code in C.
Are you suggesting that correlations cannot be run without someone "looking for something" to prove causality? If so, why would anyone ever use correlation in any sort of statistical analysis because it's merely a means to an end in the eye of the person running the correlation.
Bogus statistics example: 78% of 16-18 year old children consume large amounts of carbonated soda. 93% of 16-18 year old children attend high school. It therefore follows that there is a direct correlation between 16-18 year old children who drink carbonated soda and those that attend high school.
Please forgive my abysmal example of a correlation (because I'm really bad at doing real math and statistics) but it's there to show that anyone can create a correlation and assign it some number without having an ulterior motive. The implication of your statement above was that this correlation shouldn't have even been run except that it was to further the agenda of the author.
I think the correlation is interesting on its face, but I'm not about to use that as evidence in an international court to point fingers and shout "corruption!"
...and that's the way the cookie crumbles.
"OOXML Vote and the Corruption Perceptions Index Corruption Index"
OK, that's it for now - I have to run down to the ATM machine and put in my PIN number...
Appending "Of course, correlation doesn't prove causality." to the end of an article strongly implying causality in every sense [...]
It wasn't the article that said that, it was the Slashdot summary. A bit of a weasel word though, it should be clarified as "correlation doesn't always prove causality, but in this case we believe based on evidence A, B, C that..." or removed.
[...]doesn't absolve the reporter from the false conclusions he/she implies throughout the rest of the article.
Speaking of weasel words... What conclusions do you believe are false then, and why?
That the correlation was run at ALL implies that someone was 'looking for something' - suspect 1.
We HAVE to look for SOMETHING, both in statistics and other science. It is pretty much impossible to do as Shelock Holmes said - "It is a capital mistake to theorize before you have all the evidence. It biases the judgment." How do you know that you have all the evidence if you don't even know what you are looking for?
Sorry, this is FUD passed off as news supported by phony statistics.
You may not agree with the conclusions, but how are the statistics phony?
Being bitter is drinking poison and hoping someone else will die