Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.
Display:
To put the issue of conviction rates to rest, consider the following table
	    Total Serb Croat Bosniac Albanian Unknown Macedonian
Indicted      157  101	  32	   9	    9	    4	       2
Convicted      69   48	  13	   4	    3	    0	       1
Dismissed      18   11	   2	   0	    1	    4	       0
Ongoing        16    9	   6	   1	    0	    0	       0
Acquitted      14    2	   4	   3	    4	    0	       1
Died	       14   12	   1	   1	    0	    0	       0
Unknown        13    9	   4	   0	    0	    0	       0
Transferred    11    8	   2	   0	    1	    0	       0
At Large	2    2	   0	   0	    0	    0	       0
under the hypothesis that the court is not biased, this is the expected distribution (obtained by taking the "total" column and the "indicted" row and multiplying them)
	    Total   Serb  Croat Bosniac Albanian Unknown Macedonian
Indicted      157 101.00 32.000   9.000    9.000   4.000     2.0000
Convicted      69  44.40 14.100   3.960    3.960   1.760     0.8790
Dismissed      18  11.60  3.670   1.030    1.030   0.459     0.2290
Ongoing        16  10.30  3.260   0.917    0.917   0.408     0.2040
Acquitted      14   9.01  2.850   0.803    0.803   0.357     0.1780
Died	       14   9.01  2.850   0.803    0.803   0.357     0.1780
Unknown        13   8.36  2.650   0.745    0.745   0.331     0.1660
Transferred    11   7.08  2.240   0.631    0.631   0.280     0.1400
At Large	2   1.29  0.408   0.115    0.115   0.051     0.0255
At this point one would have to aggregate some of the rows and columns on order to ensure that the "expected" number in each cell is at least about 5. But we won't do this yet. First, for each cell in the table (except for the "Total" column and the "indicted" row) we compute (observer - expected)^2/(expected) as a measure of the deviation from the null hypothesis. We get
	    Total Serb Croat Bosniac Albanian Unknown Macedonian
Total	     73.4  7.5	 5.9	 8.6	 15.8	 30.9	     4.7
Convicted     2.4  0.3	 0.1	 0.0	  0.2	  1.8	     0.0
Dismissed    29.4  0.0	 0.8	 1.0	  0.0	 27.3	     0.2
Ongoing       4.0  0.2	 2.3	 0.0	  0.9	  0.4	     0.2
Acquitted    28.8  5.5	 0.5	 6.0	 12.7	  0.4	     3.8
Died	      3.6  1.0	 1.2	 0.0	  0.8	  0.4	     0.2
Unknown       2.7  0.0	 0.7	 0.7	  0.7	  0.3	     0.2
Transferred   1.4  0.1	 0.0	 0.6	  0.2	  0.3	     0.1
At Large      1.1  0.4	 0.4	 0.1	  0.1	  0.1	     0.0
Where we have added the numbers by row and by column into the "Total" row and column. The total deviation is 73.4 which should be compared with a chi-square variable with 47 degress of freedom (for 6 ethinicities x 8 trial status - 1). The p-value is 0.00819 so appears to be a significant deviation. The trouble is that the single largest deviation comes from the 4 "dismissed" cases of "unknown" ethnicity, and also from the fact that there are a lot of cells with expected numbers of observations below 1 which makes the approximation (observed - expected)^2/expected a very bad approximation to the true likelihood function, which is what we're estimating here.

So, what we need to do is to aggregate the "unknown" (I choose to do this with the smallest category, "Macedonian" into "other") and also the "al large" - which I will aggregate with the "Ongoing".

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 06:46:40 PM EST
It should be clear from the preceding discussion that there are still too many zeros and the table needs to be further aggregated. We can do the following aggregation
Dismissed + Acquitted -> Acquitted
Ongoing + Died + Unknown + Transferred + At Large -> Unknown
Bosniac + Albanian + Unknown + Macedonian -> Other
	  Total Serb Croat Other
Indicted    157  101	32    24
Convicted    69   48	13     8
Acquitted    32   13	 6    13
Unknown      56   40	13     3
Now, under the null hypothesis of that ethnicity and trial status are independent, the expected number of observations is
	  Total  Serb Croat Other
Indicted    157 101.0 32.00 24.00
Convicted    69  44.4 14.10 10.50
Acquitted    32  20.6  6.52  4.89
Unknown      56  36.0 11.40  8.56
The observed deviations are
	       Serb	 Croat	    Other
Convicted 0.2938299 0.08045151	0.6154036
Acquitted 2.7954551 0.04182424 13.4398968
Unknown   0.4384904 0.22037435	3.6118488
add up to 21.5 for 4 degrees of freedom (3-1)x(3-1). So this is clearly significant, with a p-value of 0.025%. The largest deviation is in the acquittal of the "Other". This might be because more "Other" were indicted than should have been, or because more have been acquitted than should have been, and that cannot be answered from the data.

(by the way, the number of degrees of freedom in the previous comment was 35 = (8-1)x(6-1), not 47)

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 07:23:41 PM EST
[ Parent ]
So, if my reading is correct:
> others are getting acquitted at significant rates
> Croats are getting off the hook at significant rates
by vladimir on Thu Mar 19th, 2009 at 03:10:06 AM EST
[ Parent ]
(My emphasis)
So, if my reading is correct:
> others are getting acquitted at significant rates
> Croats are getting off the hook at significant rates
No. The expected number of acquittals under the null hypothesis (which is that trial status is independent of ethnicity) is
	  Total  Serb Croat Other
Acquitted    32  20.6  6.52  4.89
The observed number of acquittals is
	  Total Serb Croat Other
Acquitted    32   13	 6    13
The normalised size of the deviations is
	       Serb	 Croat	    Other
Acquitted 2.7954551 0.04182424 13.4398968
The significant difference is between the acquittal rates of Serb indictees (low) and "Other" indictees (high). Croats are being indicted at the average rate for the whole court.

Now, it does not follow that Serbs are being convicted at a higher rate, though one should check. The expected number of convictions under independence is

	  Total  Serb Croat Other
Convicted    69  44.4 14.10 10.50
And the observed number is
	  Total Serb Croat Other
Convicted    69   48	13     8
which (although it has a 7% higher Serb conviction rate than expected) is not significantly different from the expectation for any group (basically, 7% deviations are to be expected in a random experiment)
	       Serb	 Croat	    Other
Convicted 0.2938299 0.08045151	0.6154036
So, the "missing" acquitted Serbs are maybe (but only partly) in "excess convictions" with the rest in the "unknown" category (which includes ongoing/transferred cases, dead suspects, "at large" suspects and "unknown"). If you go back to the top-level post you will find that the largest deviation among these categories is the "Dead", of which 9 would be expected and 12 are observed.

So, another difference is that more "Serb" (and Croatian) cases are dragging out than those of the "Other" ethnic group. The expected number of "unknown" cases is

	  Total  Serb Croat Other
Unknown      56  36.0 11.40  8.56
and the observed number is
	  Total Serb Croat Other
Unknown      56   40	13     3
with normalised deviations of
The observed deviations are
	       Serb	 Croat	    Other
Unknown   0.4384904 0.22037435	3.6118488
In fact, given the deviation, the only "effect" is a speedier case resolution for the "Other" ethinic group.

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 Thu Mar 19th, 2009 at 03:54:02 AM EST
[ Parent ]
Croats are being indictedacquitted at the average rate for the whole court.


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 Thu Mar 19th, 2009 at 03:57:48 AM EST
[ Parent ]
ok. That's clear now.
by vladimir on Thu Mar 19th, 2009 at 04:00:14 AM EST
[ Parent ]
And thanks for the number crunching effort! Hip Hip Hip...
by vladimir on Thu Mar 19th, 2009 at 03:16:39 AM EST
[ Parent ]

Display: