CLPEX Chat Board

The Detail this week has Mr. Cole outlining some bad IDs. So there are 22 cases with bad IDs and there are millions of cases with IDs that are not bad. If we gathered statisics, would it really be that bad to say in court "The error rate among practioners is X out of X" "And my known error rate is zero" (hopefully)

Any jury that hears that the chance of an error is 1 in a million or in 10 million or whatever it might be will still side with the examiner in that it is an ID.

I agree with the fact that there should not be a problem in disclosing the fact that errors have and do occur, to the courts and juries (if asked). At the end of the day, when presenting evidence in court, it is not our job to convict but to present our evidence in an objective, professional manner with the facts at hand.

The only problem is that, as Mr. Cole has pointed out, we do not know what the error rate is on either side, and by that I mean erroneous inclusions and exclusions.

Anonymous wrote: The only problem is that, as Mr. Cole has pointed out, we do not know what the error rate is on either side, and by that I mean erroneous inclusions and exclusions.

I think there is another dimension to this problem. Some defense experts want to lump erroneous inclusions and exclusions together. To me that is like having your cake and eating it too (whatever that means).

I think a lot of erroneous exclusions come from an examiner being overly cautious. A miss at that point happens because they are not looking for agreement. They are actually practicing what the defense wants us all to do! They are looking for disagreement, and in doing so they failed to make a match that was somewhat obscure.
I like Kasey's terminology... Type I and Type II errors. One being serious, and to me Type II being less serious. However, all too often defense experts want to ignore this distinction because a high all inclusive error rate will be scarier to a jury. In that case it is not about the truth... its about a vigorous defense.

With all that being said I think we should take the individual erroneous identifications very seriously even if they are statistically small. It is the same principle as distinguishing the two error types. Type I almost always carries with it heavier ramifications so we need to do everything in our power to reduce this possibility. Presenting a laissez-faire attitude about such things I think hurts our credibility in court and may even effect our work product.

Mike

I think a lot of erroneous exclusions come from an examiner being overly cautious. A miss at that point happens because they are not looking for agreement. They are actually practicing what the defense wants us all to do! They are looking for disagreement, and in doing so they failed to make a match that was somewhat obscure.

I must say that I fail to see the difference. Both identification and exclusion are unambiguous conclusions that place the source of the latent in a defined subset of humans. The fact that one subset has one element, while the other is the set of all people who are not the unique source make no difference.

It seems to me that the the two conclusions require equal rigor. And, without having thought deeply about it, it may be that, since exclusion implies that I can confidently say that conditions have been met:

1 - The latent is of sufficient clarity and quantity of detail that it would be possible to associate it with some matching portion of the source skin. if that source was available.

AND

2a - The entire friction ridge surface is available for comparison, and it can be shown that the latent is not an impression of any portion. OR

2b - The latent CAN be absolutely identified as representing a particular subset of skin, and all those portions of the potential source are available, and it can be shown that the latent is not an impression of that subset.


Thus, the cases in which ALL POSSIBLE potential matching portions of the potential source are not available do not permit Exclusion and can only be Inconclusive.

On one end of the range is the latent that is clearly a finger or even a particular finger. One the other end is the clear and detailed modest size impression of some portion of skin that could be anywhere. One would have to have the most pristene and comrehensive major case prints to pronouce exclusion in the later case. And, of course, a most exhaustive examination.

SWGFAST articulates this view in Standards for Conclusions, conditions that must be met for Exclusion,

Further, I do not accept the notion that an error in exclusion is inherently less harmful than an erroneous identification. I can easily propose a scenario in which an erroneous exclusion could be as significant. The fact that errors of identification are more often harmful may be true, but that fact is no help to the victim of the more uncommon harmful erroneous exclusion. And, in the realm of having and eating cake, if examinations properly are conducted without reference to the facts of the cases, discussion of relative weights of exclusion error and identification error are meaningless.

Overcaution cannot, absent other error, result in exclusion. Inconclusive, yes. But not an absolute statement placing the source in a defined subset of those other than the compared known.


Which raises the question of whether there is such a thing as an erroneous Inconclusive. I think not.