Suitability or "What is Enough"

[Charles Parker]

In my office there was a recent discussion that I hope maybe this message board can assist.

The issue is "suitability". One of the first steps in Analysis is to determine if a latent image is suitable. Some may call this "Value or No Value assessment" or if you go back a number of years, some may call it "Identifiable or Non-Identifiable". In my office today we use the term "Suitable or Non-Suitable".

My question is in your office what does "suitable or non-suitable" mean.

1. Does Non-Suitable mean that the latent image does not contain sufficient quality or quantity of friction ridge detail to be individulized.

2. Does Non-Suitable mean that the latent image does not contain sufficient quality or quantify of friction ridge detail to be compared.

Or any comment you may have please add to this post.

[Andrew Schriever]

Charles,

Though I dont use the term "suitable", I think that the terminology that the terminology I use is similar to yours.

I use the term comparison quality. During my analysis I determine if the latent (or inked) print is comparison quality. If a print is not comparison quality that means that, in my opinion, the print is not sufficient for individualization. Not every print that I determine (during analysis) to be of comparison quality is ultimately identifiable.

Hope that helps.

Andrew

[clpexco]

Charlie,

In my opinion, for the least conflict, you have to define the term as value/no value "for comparison". Here is why:

If you refer to "no value" (and mean for identification), then you have a problem when you have an exclusion on a "no latent of value". You end up either not reporting it (intentionally overlooking an exclusion), or changing your "no value" to "value"... for exclusion. (a confusing prospect)

In my opinion, a better alternative is to refer to "value" (and mean for comparison), then you are not stuck when you have an exclusion (you report it), and you are not stuck when you can't exclude (you report inconclusive).

If your goal is to never report "inconclusive" then you have to refer to value for identification, and some day you will be forced to either overlook an exclusion or reverse your "value" decision.

These are the opinions of a humble young examiner who sometimes goes against the grain. Ultimately, it comes down to a matter of your policy/procedure. But it sounds like you are discussing how to define those now, so I would have to recommend reporting for purposes of comparison, and give guidance as to when to use the "inconclusive" determination.

Let us know how it goes! And enjoy the ride.

-Kasey

[Andrew Schriever]

Kasey kind of took the words right out of my mouth. I tend to use the value/no value for comparison because it gives me more options. I dont tie myself down to saying the print is not identifiable, even though (through analysis) I doubt that it is identifiable.

Id rather spend more time comparing more prints (that may not be identifiable in the end), than take the chance of overlooking an identifiable print because it looks to be just below my 2QA threshold upon initial analysis.

The above strategy has paid off in the past as I have performed comparisons on latents where my initial thought was "I'll look at it, but it is borderline identifiable at best" and ended up with an identification. If I tied myself down to saying "is this print identifiable or not" from my analysis phase, I may have determined "no value for identification" and set it aside to work on other prints.

[David L. Grieve]

Charlie,

If one follows the ACE model, the purpose of analysis is to determine if the unknown has sufficient quality and quantity to proceed to comparison. Therefore, it has always seemed logical to me to use suitable or not suitable for comparison. Evaluation is the determination that a suitable print can be individualized to the known, was not made by the known or neither individualization nor elimination cannot be determined.

A latent may be suitable to compare but not individualize. During analysis, it is not always possible to predict a definitive conclusion is possible, nor, in my opinion, is that the purpose of analysis. I do believe that based upon evaluation a suitable latent print can be declared as unidentifiable. But as long as suitability determination is confined to comparison purposes, this is not a conflict.

[Cindy Rennie]

We have two categories: "Insufficient quality of detail to effect an individualization" and "Suitable for comparison only".

These two categories grew out of the examination of the latent fingerprint to determine if it was "good" enough for an AFIS search. If it wasn't, but there was still sufficient ridge detail to individualize, it was marked as "comparison only". If there wasn't enough detail to individualize, it was marked as "insufficient".

Here's the rub: what if the latent does not have enough detail to individualize but has enough to exclude? Do you tell the investigator that the latent could not have been made by suspect A because I can tell that the latent is a whorl (for example) and suspect A has no whorls? What if suspect B has whorls but there is not enough detail to individualize suspect B? (If you tell that investigator, "It could be suspect B but there is not enough detail to be sure", he's going to run out and arrest suspect B on your say-so and call you to court to testify.)

If you compare and make an identification, what do you tell the defense counsel in court? Does not "suitable for comparison only" lead one to believe that you need the fingerprints of a suspect to 'assist' with the comparison process? Does this not lead into the quagmire of 'observer bias'? ("You had the mind-set that this latent belonged to my client because you relied on the known fingerprint of my client to compare to the "comparison only" latent.")

When I analyze, compare and evaluate a possible identification, always in the back of my mind is the question, "Am I willing to go to court on this?".
If I have any doubts at all, I mark it as 'insufficient quality of detail' and move on. I tell myself two things:(1) It's not worth my career and the reputation of my Police Service (not to mention the science of fingerprint analysis) if I make a mistake; and (2) We'll get him next time.

It's a tough question, though.

[Charles Parker]

To those that have responded to this post and those that have voted. Thank-You for your input. It is interesting as I write this, that the vote was split 50-50.

I find the discussion of suitability most intriguing because it is (I believe) the hardest concept to teach, the hardest concept for new examiners to master, one of the most difficult to explain to lay people, and probably has the highest level of disagreement between examiners (especially what is comparable and what is not).

I use the word suitability because of the SWGFAST guidelines “Friction Ridge Examination Methodology for Latent Print Examiners”, section 3.1 on Analysis that says “Analysis is the assessment of a friction ridge impression to determine suitability for comparison.” In this document the word comparison is there so determining suitability for individualization is not a choice that SWGFAST has allowed for others to choose. If your section SOP states to determine suitability for individualization you can never say you’re following SWGFAST guidelines, because obviously you do not. I find it interesting that SWGFAST does not list in its terms and definitions anything that has to do with “Suitability, Value-No Value, Identifiable Non-Identifiable”, or the new one I learned this week “Comparable and Not Comparable”. Perhaps the concept is too broad and subjective to put a definition to it. But it is one of the first and a very crucial step involved in latent print analysis (I believe).

Kasey got the first lick in and responded with what is actually the meat of this post. Suitability for comparison or individualization. The term “Value or No-Value” that he proposes I have answered that above as to why I use suitability because of SWGFAST. In the old days I did use Value or No Value. One could say that Suitability is an 11 letter word for a 5 letter concept.

I understand the argument about exclusions from a latent print that can never be identified to a single source. I have seen it done, and I have done it in a particular case. However those are very rare and I believe should be taken on a case by case basis and is defined by the site of the latent print. In other words does it have probative value? There is a difference in an exclusion from the exterior of a vehicle on an auto burglary versus an exclusion of a possible suspect from a latent that is in the victim’s blood on the murder weapon. Should all latent prints be treated equally in each and every case? Perhaps that is left to another post and discussion.

I do not think it should be any latent print examiners goal to not report “inconclusive”. I report them routinely. If someone tells me that they do not report inconclusive, then I think they were not trained very well.

Kasey, you made a good post and you have never gone against my grain. My section has defined this issue a long time ago and I am not likely to change the writing on that stone tablet.

You are right about having to backtrack on occasion if you evaluate suitability for individualization instead of comparison. But I have seen the other (evaluate for comparison) also have its problems in some cases (distortion) as well as the courts (motion to exclude report) and as well as with investigators (how can we make it match). But hey, we are not governed or our rules set by them are we? Both concepts (philosophies) have their own little quirks and problems.

Andrew, I am curious if you are doing up front assessments of the latent prints when they first come in to your section? We do no up front assessments for suitability (AFIS is a different area) when latent prints come in. The assessment (evaluation or analysis whichever side of the bed you are on) is done only by the examiner when it is assigned. Doing it when it is assigned puts it under one hat and one hat only.

David thanks for jumping in. Correct me if I am wrong but did not this very same topic (along with methodology) consume two years of SWGFAST time? A lot of good discussion on this one. I just have a couple of questions. What if during the analysis of the unknown it was determined that the unknown did not have sufficient quality and quantity to proceed to comparison? What would be the conclusion?

Cindy, like Kasey, Andrew and David you have hit on some key elements. I hope for further discussions on this topic.

Thank you for your time. You have given me more to ponder on.

[David L. Grieve]

Charlie,

I apologize for the delay but I have been on the road again lately. Yes, we had some most interesting discussions about this topic when John Vanderkolk proposed the sufficiency curve that turned out to be an effective depiction of the issues. During analysis, the curve defines the separation between sufficient quality and quantity and insufficient quality and quantity. There are no numeric values to the axes, and the location of the curve in relationship to the axes defines the relative skill level of the examiner. As you probably recall, the first version had a sharp division between sufficient and insufficient, but we added the gray area. The reality is that a gray band will always exist, for latents with minimal information but with the potential for individualization will always exist. This gray zone also reveals a philosophical approach to suitablity determination. Those who are motivated to get the maximum potetial from evidence will be mostly in the lower region of the gray band, near insufficiency. Those a little more conservative will usually be near the top of the gray band near sufficiency. But I think, in some cases, this varies with each person, whether it is early morning or late afternoon, whether it is Monday or Friday, or as Pat has pointed out, whether you have a cold or allergies or watched the late, late show the night before.

I don't mean to imply this is strictly a subjective decision. I think we have reasonably objective guidelines but interpretation of what is observed is subject to numerous factors. I think there will be very few exceptions to sufficiency and insufficiency determination, but shift can occur in the gray zone.

If the examiner determines the unknown has insufficient quality and quantity of information, the latent print is deemed not suitable for comparison. In that event, that is the end of the process. But this critical element has to be monitored by some mechanism that ensures drift does not occur. By drift, I mean the entire curve is moved away from the axes or the gray band becomes broader and broader. We do it with random reanalysis in which an examiner serves as a quality review coordinator for a certain time and reworks cases selected at random from the other examiners. It isn't foolproof, but this can reveal when an individual, or even an entire section, has drifted outside the "norm."

A combination of circumstances provided an opportunity several years ago to apply the curve to analysis and test the results. Our quality review coordinator had discovered that drift was occurring to the extent that an unacceptable level of inconsistency was going on. We decided to hold a workshop for all examiners just dealing with the subject of analysis. We prepared a series of exercises, most containing nothing but 20 borderline impressions. We limited the class to about ten people and tried to have a cross section of all our laboratories in each class. For nearly a week, all we did was analyze latents and discuss them in depth. Each impression was projected on a screen and class members had to state their determination, then defend it. Others were encouraged to jump in with observations of detail not mentioned, or to present arguments for the opposite call.

Some most interesting observations were made. The first set of exercises consistently showed the greatest variation. By the end of the week, using with set nine or ten, we had the greatest consistency. But we never had all ten examiners agree on all twenty borderline latents, even at the end. But the more people talked about what they saw, what significance they placed in each feature, what third level detail was there, the more narrow the gray band became. I doubt it is possible to have 100% agreement on all suitability calls, but I am convinced that promoting discussion about suitability in a section has a tangible benefit.

[Andrew Schriever]

Andrew, I am curious if you are doing up front assessments of the latent prints when they first come in to your section? We do no up front assessments for suitability (AFIS is a different area) when latent prints come in. The assessment (evaluation or analysis whichever side of the bed you are on) is done only by the examiner when it is assigned. Doing it when it is assigned puts it under one hat and one hat only.

Charles,

If, by up front assesment, you mean "are the latents evaluated prior to case assignment"? The answer is no, the latents are not analyzed until the examiner cracks open the case.

[Rebecca Stokes]

I am curious if the majority of other agencies do not do up front assessments of their latents? My agency rates them when they come in each day. We are unsure if this is a good practise or not after reading some of your posts.
On the other topic, we use Value/No Value. The No Value rating is used for those prints that do not have any valuable information for comparison, whether that means individualization or just elimination. Then Of value or AFIS quality is used to rate others. These ratings however are subjective and could depend on the examiner. Differences of opinion among our staff is not unusual. This, however, does not cause any problems, just some good natured ribbing of the examiner if they were too liberal in their initial rating. But these daily ratings are done by visually looking at the latent card and not on latents enlarged or enhanced, so initial opinions about a latent can change when the case is worked. Which leads me back to my intial question. So far this hasn't caused a problem, just wondering what other agencies do and why.

[Andrew Schriever]

I know that both agencies that I have worked for have never done up front assessments of the latents when cases are first submitted for examination.

I dont really know how your SOP's work Rebecca, but it seems like this practice could lead to bias the examiner the case is assigned to. Say the supervisor or senior examiner that does the up front assessment says that latent X is of value (suitable) for comparison, but when the examiner working the case looks at it they may disagree. The case examiner may feel pressure to be more liberal with their analysis of the latent because the more senior or more experienced examiner has already said that the latent is suitable. I'm not implying at all that this happens at your agency, just throwing out the hypothetical that this could happen and could lead to mistakes.

In my opinion, it seems like the initial assessment should be done by the examiner working the case. Then, during review, any errors regarding the assessment of suitability can be addressed with the case examiner by the reviewer.

Does that make sense? Is my logic flawed? I am interested to know because this is really the first that I have ever heard about latents being analyzed prior to case assignment.

[Rebecca Stokes]

Latents are submitted each day by officers in the field. One of the techs in our office receives them, opens the sealed envelope and rates the latents inside. The AFIS prints are set aside on the AFIS desk and the rest are rated as No Value or Of Value. The latents then go in file until we receive a request by detectives or the DA to compare them. The request is assigned to a tech and that tech then pulls the latent envelope in file. Even though a latent envelope my have No Value rated, we still will re-examine the latents inside. This has been our SOP or routine for many years. We have talked about the very issues that you have raised. So far it hasn't been a problem but this has started a good discussion among our group, (only 5) which is always good for us. I am still curious about how other agencies work, knowing we are all a little differently run, but you never know when you might find a better way of doing things. Thanks for the reply!!

[Andrew Schriever]

OK, that is a little different SOP than I had imagined. Seeme like it works very well for you, but I just have one question.

Why review them initially if you are going to review them all again anyhow? Do you do that to you can report to the officer/detective quickly not to expect any comparisons performed because of a lack of sufficient quality prints (if that be the case)? If so, I can definately see the value.

[Rebecca Stokes]

Yes, although the dets don't always pay attention to our notices. Sometimes they submit requests even on cases where no latents have been turned in. But that's another disucussion for a different day!! If they get a notice that the print is AFIS then they know that we will enter it automatically. It is just another piece of information we can give them about the evidence found at the scene.
This is routine for us and I guess we will continue until we find good reason not too. But the discussion has been good and I would still like to hear from other agencies on their protocol.

[Charles Parker]

Rebecca,

I have had over the years the opportunity to work at 3 different agencies. Two PD’s and the State Lab. Also I have been an instructor at the State Academy since 1992, talked to a number of people and have visited about 12 agencies in this state.

I cannot speak for other states, but there are a number in this state that does the very thing you are talking about. I have seen two agencies that do not receive their latent lifts in envelopes but attached to an evidence tracking sheet. The examiners after determining what is identifiable and non-identifiable will then file the Non-IDs in one cabinet and the ID’s in a separate cabinet. Latent prints in the same case separated and filed separate. That is rare and I have not seen many do that.

My first agency conducted those receipting evaluations and it did cause problems on occasions. None that ever caused the sky to come crashing down in on us, but some heated discussions among the examiners. I do not remember as Andrew suggested that mindset came into to play. Probably because we had too many A-Type personalities and people were ready to go to war over what they believed to be the only right choice. The reasoning for doing it was to let the investigators know that latent prints were of no value so they would not be submitting unnecessary requests. As you can imagine that did create some problems, especially with some of the high profile cases. I attempted on a couple of occasions to eliminate that but just like oil on raw wood, once something like that sets in place it is hard to get out. Then I went to the State Lab. I believe State and Federal Labs are a lot different from PD’s and SO’s. If something is going to a State or Federal Lab it is going to be assigned sooner or later. Why else send it to them. Whereas PD’s and SO’s only a percentage of their latent cases are assigned. For example the agency I am at now, probably only 50% to 60% of the total number of latent cases we receive are actually assigned for comparison. A good number stay sealed and in the cabinet. When I arrived at my current agency I was placed in a position of having again to conduct up front assessments or evaluations. Not only did they determine if it was AFIS quality, but identifiable quality, and we had to indicate if palm prints were in the latent envelope. That information was put into a property report, again to let the detective know if latent prints were suitable in his case. One examiner was assigned each month and the duty rotated. Finally we were able to dump that process. The initial reaction from the old heads here was “Our work is going to increase as the detectives are going to submit comparison requests on latent prints when they are of no value”. Well that did not happen and it freed up some man hours. Now the value or no value is determined when that case is assigned. Cuts out some of the problems that Andrew has stated in his posts, and he is absolutely correct. Why review them twice.

For any agencies that do up front assessments for comparison or identification value, I would suggest to look at that seriously and do away with it. You would be better off doing Administrative and Technical Reviews on the finished product instead of trying to stop a few comparison requests from hitting your section.


Dave:

Thanks for the post. I too have just come back from a trip and I am trying to catch up at home and at work. I never realize how wide this state is until I have to travel it from one end to the other.

I love your idea of the workshop. In the last two advanced classes I have started something similar. I break them into groups and give them 15 sets of borderline, difficult, and what I call BNH (Best Non-Hits). I ask them as a group to go over them, and then state their observations as to distortion, background, conclusion, etc, and then ask the other groups to come in with any other observations. I have only used it twice, but each class seems to like the exercise a lot. It helps them see what others are thinking and doing. It is not as extensive as the one you listed, but then I am teaching a different group of people. I am thinking about expanding it out to the Intermediate Class, but not with the same level of difficulty. I think your idea should be used as an in-service for every latent print section every two or three years. I am going to see if we can get something similar here.

Back to the current post. After reading a couple of e-mails this morning, I believe I have found my problem and the way to correct it. I have been confusing results with conclusions. I have been trying to arrange my results with the three conclusions in the SWGFAST Guidelines, and it was not working for me. After the e-mails and reading the SWGFAST material I realized that in actuality there is a fourth response. Those three conclusions are only there if you have a latent print suitable for comparison. There was not a stated fourth conclusion for that situation where you have a latent print that is not suitable for comparison. You see for me when I make a determination that what I am looking at has no information what so ever to lead me to making a conclusion of individualization, exclusion, or inconclusive, it is just that “No Value For Comparison”. To me it has as much value in the overall analysis as the three stated conclusions. Also the definition of Inconclusive was throwing me a little: “The standard for an inconclusive finding is the absence of sufficient friction ridge details to effect a conclusion of individualization or exclusion.” It does not state or mean those that were determined of “Of Value for Comparison” only. At least for me you have to read that into it. It may not seem that way to those who were there for the SWGFAST discussions or who correspond on a daily basis with other highly educated and well informed LPE, but I think for some of us it is misleading when only three conclusions or results are stated but in actuality there are four.

Anyway I have now worked it out in my head (I think), so I want to thank you for your response.