CLPEX Chat Board

Another perspective:
Simultaneity must be proved with supporting information in order to be used as a foundation for individualization. However, that is the very issue. There is a desire to use two separate impressions where each is insufficient and cannot support individualization alone. The solution is to sum their probabilities to effect an individualization? I am not sure this is correct within the laws of probability specifically for independent events. Whereas two [independent events] should multiply in probability rather than sum. It is like the old assumption that "two Roman half truths would make a whole truth", when they should represent something like 75%.

In reality, we do not have the information to understand the actual probability of each impression as having been made by a particular person. Thus, they must be viewed as independent impression events. The probability that they would both be (1) from the same person, (2) made simultaneously, and (3) support individualization should be [less probable] as the conditions multiply rather than sum. Of course, if the simultaneous impressions could be proven on their own, an alternate solution would not be needed. We must realize the limits of our science in order to maintain accuracy and the methodological integrity of ACE-V. This notion of simultaneity of insufficient impressions supporting individualization seems dubious.

There's certainly considerable presumption brought to bear when one declares two distinct impressions to be simultaneous. And I think that opens a wider than usual door to criticize an identification made from a single print in a cluster in which another impression excludes the first source OR is insufficient on its own but contains detail inconsistent with the finger it would be, if it were being viewed as simultaneous. In other words:

Consider. Impressions 1, 2, 3, and 4, all having the positional character of being four fingers impressed simultaneously and in a place and orientation consistent with natural contact. 1 and 3 are marks consistent with being finger impressions but have no detail. 2 is deemed sufficient for identification and is matched to an individual. 4 has some clear detail but is deemed insufficient for identification. The clear detail in 4 excludes as its source the respective finger that would have been impressed if they were all simultaneous and natural.

An advocate for the identification of 2 will argue that they must not, therefore, be simultaneous, so the match of 2 is valid. The opposing advocate will argue that if simultaneous impressions of limited detail may be viewed together to effect identification, a partial impression deemed simultaneous by the same reasoning of relative position and orientation must be used to invalidate the identification match of 2.

If you can use portions of different impressions to identify, you are viewing them as a single impression of portions of the total friction ridge skin of the source. If one had found among a presumed set of partial impressions that a partial impression needed to conclude identification a clear inconsistency, one would not have concluded an identification. This is not different from the usual case. In this sense, all presumed single impressions can be viewed as simultaneous impressions of portions of skin. You cannot ignore inconsistent details in a single impression in order to support an identification.

"But 2 was identified!" Was it? We have long held that, although a sufficient quantity of detail supports an identification, without qualification, a clear inconsistency precludes identification. It's the safety net we must recognize on account of the elastic nature of sufficiency. Can we say, on one hand, that we can effect an identification across simultaneous impressions, and, on the other hand, say that an inconsistency under the same conditions merely means they were not simultaneous? If you're going to recognize as useful this class of simultaneous impressions, and you have a personal standard for concluding that they are simultaneous, you must, when that standard is met, consider the totality of detail. And I would suggest that your standard for deeming them simultaneous cannot be that being simultaneous supports the identification.

When the hypothetical situation I present above exists, how much higher must be the threshold for identification of 2? Is there an effective threshold? And if there is, does that imply that the same threshold would allow one to conclude identification of a single impression with that same amount of detail and a clear inconsistency? (By declaring them simultaneous, you're doing exactly what you do when you declared all parts of the single impression to have the same source.) Does a simultaneous impression like my example invalidate any identification of impression 2, no matter how detailed? If one answers, "No." is one then admitting that judging the value of others in the cluster is distinctly biased in favor of identification?

This, I think, is the dilemma of trying to use this sort of simultaneous impression. That if one accepts this principle, one must commit to the same consideration of possible simultaneous impressions in every case and not pick and choose. And, does that mean that you could be called upon to know if any single impression could have been part of a simultaneous cluster of which you were only shown the clearest? I do not reject the principle of using this class of cluster. I just think it's open to vigorous challenge and alternative theories of deposition. Not altogether a bad thing in my opinion, since I believe all conclusions should be considered open to vigorous challenge. But I think one must be prepared to answer the hard questions.

Any idea when SWGFAST will post this to their website? I have an older version of the draft and I can't follow along with the comments above. Please help!

Before conducting a forensic examination of a simultaneous impression, examiners shall have completed
training in the examination of latent print simultaneous impressions.

Answers to a few questions would be appreciated. Thankyou.

Does anyone currently offer a training course dedicated to simultaneous impressions?
To what extent are simultaneous impressions covered in other training courses people have attended?
What proportion of labs actually sanction the release of identifications based on aggregated detail;
a) when there is a stand-alone print and one in aggregate?
b) when no prints stand alone and all prints are considered in aggregate?

David,

We (SWGFAST) also envisioned that an agency wishing to conduct these examinations could meet this requirement by including a module in their training program. Demonstration of competency could also be achieved after the formal instruction with exercises created in-house, documenting accurate and appropriate application of the method. This would be a relatively straightforward in-house process. However, the question of "who trains the trainers?" and "who is initially competent to perform the exams", can be a matter of debate. But agencies believing they meet such requirements, need only to justify/document how they meet it (i.e. this person is competent to perform the exams b/c they took XYZ courses, demonstrated XYZ knowledge, and successfully completed XYZ examinations specifically related to this method). I think the "successfully completed XYZ examinations" (where the ground truth is known) will go a long way to demonstrating competence.

g.

Sorry to drag this one back from the archives but my efforts to get my agency to sanction SI examination are still ongoing and I need to do some more background reading.

Can anyone supply me with a .pdf or other copy of the following paper, referenced in the SWGFAST standards?

Black, J. P. Pilot Study: The Application of ACE-V to Simultaneous (Cluster) Impressions. Journal
of Forensic Identification. 2006, 56 (6), 933 - 971.

Thanks in advance.

David