CLPEX Chat Board

Hello All

I was under the impression that the IAI does not support qualified opinions on comparisons where (traditionally) there would have been less points than would have been accepted.

In line with the more modern holistic approach and, given the uniqueness of every comparison, is there still such a thing as a qualified opinion?

I believe there could be if it was allowed.

But by not allowing the term 'Probability' to enter into the conclusion then it is kind of behind the curtains. It may never come out in a report or is reported as Non-Suitable or Inconclusive, but the idea is still there----just supressed.

As long as there are issues of quality then you are going to have qualified opinions (view points) that are not offered up as front line conclusions.

I have never seen the term "unique comparison" before and it is interesting that you used it in that manner.

I guess everything in life is unique-----never duplicated.

Not to create further confusion in the matter but I didn't think that unique and comparison was used inappropriately in the same sentence.

I recall having read the idea in various places and publications that ...every comparison / conclusion is a unique set of events...loosely paraphrased in my own way.

Surely no two comparisons are identical in every way? Maybe my choice of words were not the normal or generally used terminology but that is how I understand it.

Anton

Whenever a conclusion is based on observations that can't be quantified, every conclusion could be said to be qualified. It's an area where our argument that the known history of examinations and our own experiences lead to a high degree of certainty can be viewed in a somewhat different way. I could state, in a very rough sort of way that "when I see this much, and I see nothing that clearly contradicts it, I have a high confidence in this conclusion." I am, in effect, saying that there is some undefined but very small chance that, if I could see more, I would not accept that conclusion. I am saying that the weight of what I can reliably interpret leads me to conclude that, if I had more surface well impressed, perhaps if I could see what was in blank or black or hopelessly smudged areas, and there I observed a clear contradiction, I might not reach that conclusions and might well add that experience to my previous experience and could apply that in the future.

We cannot make conclusions of numerical probability. We don't have the numbers. But we are always really assigning an unspoken probability in every conclusions. That is true of every kind of conclusion in every discipline where findings are stated in numerical values, and it is the numerical finding that is debated as to its interpretation. DNA results are stated in absolute results, and the statistical likelihoods of identification are presented, and their magnitude is cited to assign probative power to the finding. Of course, some observational comparisons are a mix. A tire mark examiner may declare absolutely that an impression is that of a specific make and model of tire, because the absurd case of someone being diabolical enough to craft a fake is cast out. The remainder of the observations are of wear, damage, and embedded material that cannot be quantified and are used as we use latent impression observations.

Like latent examination, those tire mark conclusions are subject to highly unlikely but conceivable reservations. For instance, a tire of an identified make and model is compared to an impressions. The wear and accidental marks are found to match to a degree that would satisfy a conclusion of high certainty. But, if it is found that data impressed on the tire during manufacture shows that particular tire was not molded until after the impression was made, the examiner would have to abandon the conclusion and would add the experience to his professional store and would likely seek to provide other tire mark examiners with the same experience.

I think we have seen the same process, just moving in the other direction, at work in latent print examination. While I haven't done a study of the early literature, it seems to me that early latent print examinations demanded a great deal of available detail, quite high numbers of details. Equally strongly stated conclusions are today made from far less impressed area and far fewer available impressed features. (I realize I might just be seeing a tendency for early writers to show only the slam-dunk cases out of their desire to validate the discipline.) But if my impression from the literature is correct, it's simply a matter of experience gained and shared.

It is inevitable, when the observations cannot be quantified and specific quantities used to conclude, that some general sense of what amount and kind of reliably interpreted features can be observed be propagated through the discipline. The high certainty of conclusion is really made by applying a general sense of what the community of practitioners accept as conclusive, rather than the tested experience of the individual examiner alone. In rendering a conclusion, I am not saying that I base my conclusion on my personal experience of comparing my previous observations against my certain knowledge of sources proven independently to be the sources of the latents I identified with them. I am mostly saying "agreement to this degree is never shown to lead to an erroneous conclusion."

We do not much see battling latent print examiners in court in the same way we see some other experts disputing conclusions. But I think this is largely because we don't see much spirited adversarial effort to present it and because we are more prone as a discipline to agree to agree. Were we to see more adversarial approaches to latent print identification, the arguments would be over degrees of certainty. Each expert would present their observations, and I suspect those would not differ too much, but they might argue for a greater or lesser probability that the source could be identified with certainty. But both would have to, if pressed, allow that their conclusion carried the caveat of "if I am right." When this happens with such experts as psychologists and psychiatrists, the statement of confidence is really (even if they don't always say it this way) "I am right, because I am right, and he is mistaken, and I know I am right because if I were not right, I would be wrong."

It seems to me that there are indeed degrees of certainty in latent print examination conclusions. Since we cannot make absolute conclusions of identification (Having to admit that there is always some small possibility of error), I think it would be foolishly dogmatic to insist that there is not some greater certainty when a perfectly impressed nail-to-nail image is available than when the latent is just sufficient for identification. We would have to admit that there is some possibility that when an impression contains a blank area, it could contain a fatal feature. What we shrink from is quantifying the difference, either by numerical probability or by applying a verbal statement of degree. It is our convention to, in effect, say "I believe." And when the affair is not one of plain faith, belief is always qualified, when expressed by anyone who has not declared their mind closed, locked, and sealed.

If I am the examiner who holds that the latent cannot be conclusively identified, your conclusion who be stated as qualified by your extravagant assumptions. If I am the examiner who holds that the latent is sufficient for certainty, my conclusion is not much "qualified", because I hold that what I see takes what you view as "mere assumption" to be things I can also conclude based on my other observations. Do we not always, as a necessary part of our examination, conclude that no fatal feature is hiding just outside the impressed area?

I think the well-intentioned discouragement from making qualified opinions may well have tended to close the door to adversarial testimony on identification conclusions and, in the long-run, led to much of the criticism of the discipline.

Antonroland

Not to create further confusion in the matter but I didn't think that unique and comparison was used inappropriately in the same sentence.

It was not used inappropriately. You used it correctly. I just have never seen it stated that way, or if I did I never noticed it before. It probably has been used a lot and I just never noticied it.

What caused my eyebrows to go up is the word unique. I am a long time beliver that everything is unique and attributing somthing like friction ridge detail as unique is to me from my humble POV --redundant.

To me it is like saying I want to eat that red apple that is red. Or I want to drive that yellow mustang that is yellow.

When people say that friction ridge detail is 'Unique' and 'Permanent', the thought that comes to my mind is "Whoop-ti-doo everything is unique".

You said nothing wrong---my mind just fixated on the word 'Unique'.

Good morning gentlemen, Anton, Charles and Gerald,

Thanks for an interesting insight into the world of fingerprint communications. How we express our conclusions should have some standardised policy similar to what SWGFAST has published. These also need the scientific capability of being defended using all the details and physical relationships that fingerprints possess by virtue of its biological nature. I agree that how we conduct our comparisons will depend on the mind of the examiner but the process should be objective and the results reproducible. As we witness daily with computer algorithms the basis of their programming cannot incorporate all the variables associated with skin formation or transfer. The human brain and eye have far exceeding abilities to discriminate and confirm details over that of a limited mathematical program. When it comes to supporting what the human brain has concluded then we can turn to the computer to produce quantitative detail about the unique geometric relationships inherent to the source skin. The variables of this algorithm have been determined since the examiner has decided the loci and the extent of the pattern arrangement. The computer is no longer searching for a matching arrangement but rather tabulating empirical data which the human brain has already analysed. To make this process work effectively the examination is structured with the end product in mind such that the hypothesis and the testing of the hypothesis must be clearly separated. What Gerald has written about sufficiency really highlights what is needed in the SWGFAST policy. I was dissappointed when they defined sufficiency with another equally elusive term 'adequate' both of which dont assist the examiner in understanding the balance required between objective and subjective components in an examination. Maybe more will come in the future. Regards. Les

No worries, Charles

I think we spoke a bit past each other there as I really intended "unique" to describe the set of events forming the comparison rather than the friction ridge skin itself.

I am sure that has now been clarified and I want to thank all for your very interesting inputs here.

Much food for thought.

Anton