Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.
Display:
Let's be more precise. There is no statistically significant difference between the two data sets analysed even though the individual data points would suggest that... ;)
by vladimir on Wed Mar 18th, 2009 at 10:04:28 AM EST
[ Parent ]
Well, not all children can be above average.

There is a case to be made that the most extreme outliers by one measure are the ones for the Kosovo war, in the direction you imply. Still, I don't think the difference is statistically signnificant. And the figure of 8,000 Serb civilians happens to be the largest estimate you could possibly find, and it is taken from a different source to the others. Wikipedia lists 5,000 as the estimate of the total number of Albanian casualties including KLA combatants.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Wed Mar 18th, 2009 at 10:08:54 AM EST
[ Parent ]
I used 3 400 civilians... which implies 1 600 KLA to get to the 5 000 cited in Wikipedia. So its coherent. The total estimated dead in Kosovo is around 12 000 - which would imply 7 000 dead Serbs. So again, the numbers I used seem coherent.

It also looks as if the Croats are getting off the hook for their "activities" in Croatia.

by vladimir on Wed Mar 18th, 2009 at 10:20:40 AM EST
[ Parent ]
vladimir:
It also looks as if the Croats are getting off the hook for their "activities" in Croatia.
Oh, really?

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Wed Mar 18th, 2009 at 11:13:02 AM EST
[ Parent ]
vladimir:
The total estimated dead in Kosovo is around 12 000
That is the "largest estimate you could find" that I was referring to. For instance, the Wikipedia article on theKosovo War has the following summary in their infobox:
KLA: 5,000+ killed [3]
NATO: 2 non-combat deaths[4]
1200 killed [5][6]
Around 100 Albanian civilians killed by NATO forces [7]
NATO bombings: Human Rights Watch was only able to verify 500 civilian deaths throughout Yugoslavia, [8][9] with other sources stating from 1,200 to 5,700 [8]
If you added the estimate of 8,000 to the number of Serb civilian casualties you'd get "Human Rights Watch was only able to verify 500 civilian deaths throughout Yugoslavia,  with other sources stating from 1,200 to 5,700 or even 8,000". If you replace 8,000 with 5,700 the data point doesn't look like an obvious outlier any more. Like I waid above not all children can be above average or some data point has to have the largest deviation.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Wed Mar 18th, 2009 at 11:42:06 AM EST
[ Parent ]
If my memory serves me well Madame Secretary Albright, the NYT and the BBC were screaming back in 1999 that there were 500 000 missing... down to a Human Rights Watch estimate of 500. At last we've found something of statistical significance!
by vladimir on Wed Mar 18th, 2009 at 01:08:05 PM EST
[ Parent ]
That's the same Albright sho said 500,000 Iraqi deaths through sanctions was "a price we are willing to pay" to topple Saddam.

Look, the fact that there are no NATO people on trial for war crimes is neither here nor there regarding the bias of the ICTY regarding the different Yugoslav ethnic groups.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Wed Mar 18th, 2009 at 01:49:08 PM EST
[ Parent ]
I don't see any figure that would suggest that the croats are an outlier in the graph and so have been particularly let off /badly treated, so how do you come to that conclusion?

Any idiot can face a crisis - it's day to day living that wears you out.
by ceebs (ceebs (at) eurotrib (dot) com) on Wed Mar 18th, 2009 at 12:36:09 PM EST
[ Parent ]
I use the base ratios of indicted and convicted per enemy civilian casualties... because the objective of the statistical analysis completed by Migeru and Jake was to search for a correlation - of which, I agree, there was none. But there was none partly because they had insufficient data points to play with.
by vladimir on Wed Mar 18th, 2009 at 01:12:43 PM EST
[ Parent ]
The convicted : civs ratio is evident.
The indicted : civs ratio is not conclusive.
by vladimir on Wed Mar 18th, 2009 at 01:17:21 PM EST
[ Parent ]
As has been explained to you already, you cannot use the number convicted:indicted (roughly equivalent to convicted:casualties since indicted:casualties is not significantly biased) as long as there are cases outstanding. You need to use convicted:(acquitted+dismissed).

You're grasping at straws - no amount of evidence will convince you that the court is not biased against Serbs.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Wed Mar 18th, 2009 at 01:29:38 PM EST
[ Parent ]
This is the correct graph.

by vladimir on Wed Mar 18th, 2009 at 01:48:09 PM EST
[ Parent ]
That's a two-way contingency table and, again, comparing rowwise ratios is not the proper way to do things. In addition, you need to aggregate rows or columns as appropriate if you have expected numbers below 5 in order to have some hope of statistical significance. So, for instance, for the Bosnia Muslim, Albanian, Macedonian and Croatia Croat rows, having less than 10 indictees each, you're not going to be able to prove much, statistically. This is not unlike when I said
With only 6 points it is really difficult to argue anything. For instance, what is the chance that all three "Serb indictee" points are above the line? 1 in 8. This is not sufficient to show bias at 90% confidence (you would need the probability to be less than 1 in 10) let alone 95% confidence (1 in 20).
So maybe you can show (given the large numbers) that the Bosnia Serbs are being shafted, but we're throwing out the 4:0 conviction to acquittal rate of Bosnia Muslims.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Wed Mar 18th, 2009 at 01:56:48 PM EST
[ Parent ]
That is an utterly bizzare argument, lacking any semblence of logic. You yourself admit that the figures do not show any bias along a lines that you have suggested.

If there was a consistent bias against the Serbs to the extent that you claim, then I would argue that it should be visible even at low levels, with few data points to play with.

Any idiot can face a crisis - it's day to day living that wears you out.

by ceebs (ceebs (at) eurotrib (dot) com) on Wed Mar 18th, 2009 at 01:58:22 PM EST
[ Parent ]
The smaller the number of data points, the more pronounced the difference between them must be in order for it to be statistically significant.
by vladimir on Wed Mar 18th, 2009 at 02:52:24 PM EST
[ Parent ]
But why particularly should the Croat points show up as worse than any others, for them to have "got off the hook"?

Any idiot can face a crisis - it's day to day living that wears you out.
by ceebs (ceebs (at) eurotrib (dot) com) on Wed Mar 18th, 2009 at 03:00:14 PM EST
[ Parent ]
Because the ratio of Croat convicted to Serb civilian casualties in Croatia DOES show up as being particularly lenient to Croats - the ratio is 0 to some 2 300 dead. Then, of course, we can argue whether this is the right metric or not... but on this metric, the Croats seem to have "gotten off the hook".
by vladimir on Wed Mar 18th, 2009 at 03:11:58 PM EST
[ Parent ]
Especially beware of highlighting - I'm sure highlighting single data points has legitimate uses, but off the top of my head, I cannot think of a single one. A very good indication that Someone Is Up To No Good.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Wed Mar 18th, 2009 at 04:42:34 PM EST
[ Parent ]
OK Jake, I get your point.... BUT if your data set is too small to work on statistical significance testing... and if you're not allowed to highlight single data points (if you don't want to be accused of being up to no good) then what can you do with small sets figures?
by vladimir on Thu Mar 19th, 2009 at 03:13:40 AM EST
[ Parent ]
Then you go get a bigger data set with a better resolution.

If that's not possible, you sit down and cry.

And when you're done crying, you stop trying to prove abuses that are impossible to prove, and concentrate on the abuses that are possible to prove - such as the excessive durations of the trials (for all the accused).

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Thu Mar 19th, 2009 at 04:58:45 PM EST
[ Parent ]
Or the lack of indictments against NATO officials.

Sweden's finest (and perhaps only) collaborative, leftist e-newspaper Synapze.se
by A swedish kind of death on Fri Mar 20th, 2009 at 04:20:44 AM EST
[ Parent ]
vladimir:
BUT if your data set is too small to work on statistical significance testing... and if you're not allowed to highlight single data points (if you don't want to be accused of being up to no good) then what can you do with small sets figures?
Then you don't use statistical arguments.

Just a simple question. How many coin tosses do you need to reject the hypothesis that a coin is unbiased with 99% confidence? 95%? 90%? And if your coin is used fewer times than that and then is lost, how are you going to use statistics to argue it was biased?

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Fri Mar 20th, 2009 at 04:48:46 AM EST
[ Parent ]
The ideal result should be heads 50% of the time and tails 50% of the time.

99% confidence means you are accepting a 1% error on the ideal.

95% means that you are accepting a 5% error on the ideal.

It would be hard to get within 1% or 5% of the ideal with very few tosses. If you do a few tosses you may actually conclude that the coin is biased even if it's not. I suspect that within 10 or 20 tosses you should seriously approach your ideal 50-50.

by vladimir on Fri Mar 20th, 2009 at 07:39:29 AM EST
[ Parent ]
The right answer is that with 3 coin tosses, even if all 3 are Heads, the chance of that happening is 12.5% so you cannot reject the hypothesis that the coin is fair at 90% confidence.

With 4 coin tosses, HHHH has a probability of 6.25% which allows you to reject at 90% but not 95%.

With 5 coin tosses, HHHHH has a probability of 3.125% which allows you to reject at 95% but not at 99%.

The point is that, with less than 4 coin tosses you cannot show bias, no matter what. Sometimes you simply don't have enough data to argue statistically.

And statistics can only suggest where to look for actual evidence, it can't prove (or disprove) bias all by itself.

For instance, the contingency table analysis I did yesterday suggests looking for actual (not statistical) evidence of bias in the duration or the trials, not in the result. JakeS posted a theory that indictments were issued in the hopes of gathering sufficient evidence by the time the cases came to trial, which in some cases hasn't happened, resulting in prolongued imprisonments without trial rather than dismissals for lack of evidence. But a theory consistent with statistical suggestions is not evidence.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith

by Migeru (migeru at eurotrib dot com) on Fri Mar 20th, 2009 at 07:55:14 AM EST
[ Parent ]
vladimir:
The ideal result should be heads 50% of the time and tails 50% of the time.
You urgently need to go read the first chapter of Feller's An Introduction to Probability Theory and its Applications which covers coin-tossing.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Fri Mar 20th, 2009 at 07:57:04 AM EST
[ Parent ]
It's only been about 20 years since my last statistics course.
by vladimir on Fri Mar 20th, 2009 at 08:06:54 AM EST
[ Parent ]
Because the ratio of Croat convicted to Serb civilian casualties in Croatia DOES show up as being particularly lenient to Croats - the ratio is 0 to some 2 300 dead.
There you go again. There are 0 convicted, 1 acquitted and 5 ongoing cases. These figures are evidence of nothing.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith
by Migeru (migeru at eurotrib dot com) on Wed Mar 18th, 2009 at 05:14:33 PM EST
[ Parent ]

Display:

Occasional Series