CLPEX Chat Board

Hi all,

Please respond to the poll by indicating whether or not your agency retains (as part of case documentation) the AFIS search candidate lists in cases. If your agency does retain the candidate lists, please indicate whether the list is retained in all cases, regardless of the search results, or only in cases where a latent hits.

If your agency retains (as part of case documentation) the candidate lists, what is the justification in doing so? For those agencies that retain candidate lists, does your AFIS system provide the means to retrieve candidate lists at a later date?

Thanks for all of your help!

For those that do retain the AFIS Candidate List-----what is the reasoning behind retaining it?

I believe part of the intent of retaining the candidate lists is for court issues. There needs to be an indication of how the suspect's fingerprint card came to the examiner's attention. Secondarily, the candidate lists are retained as an aid to QA procedures. If the QA person gets a hit on a case that the original examiner did not, having the candidate list can show whether it was a missed hit (i.e. the suspect appears on the original examiner's list). Also, if you have the exact dates of searches (rather than a range of time over which the case was worked), you can pinpoint whether an identification was missed because the suspect wasn't in the database at the time of the search (this happened to me).

Amy I thought Court Issues would be the top of the list.

I should explain a little of what we do because there might be some confusion. In the 3 AFIS systems that we run we can print out the complete candidate list (20-50 whatever it is sit at) or we can print out a single subject data and the image print out.

For those that we have determined it is a 'viable candidate' for examination or further comparison we print out the data sheet of and the image sheet of that single individual. We do not print out the entire candidate list. So several of the issues that you stated would be taken care of because of the two sheets we do print out.

Also if no 'viable candidate' is observed then a notation is made on the latent envelope so it can be determined as to which system it is searched in. There are number markings above each latent print in the case packet to indicate which ones were searched and of course no markings means that it was not searched.

For our QA they will examine the packet, make further of subjects listed, but our QA process duties do not include going back and re-running any AFIS searches. Now if an AFIS Tech determines there is a viable candidate and it then is assigned to a LPE, then the LPE may run some additional latent prints or not-----it is entirely up to the LPE what will be run and what will not be run.

Since we do not count the results of an AFIS Inquiry as a 'hit' or 'no hit' but just a 'viable candidate' developed for further examination or 'no viable candidate' observed we do not attach a lot to system misses. The system will miss them----and yes a viable candidate might get past the reveiw process, but we do not attach any negative significance to it.

But I do enjoy these discussions on how different agencies work their work. Anybody else have comments or suggestions.

I would say my agency does it so there is a level of transparency in your casework. It is a document providing confirmation of what you said you did. It has potential to work as a quality assurance measure to prevent analysts from "dry-labing", saying they performed something when they never actually did. The list itself isn't going to be that helpful for a case reviewer or Defense, however if they wanted to take the time they could retrieve each specific finger compared to in the database to make sure your results are accurate.

If we want to be considered a scientific discipline then showing our work is instinctly part of that.

I am now 5 years removed from working at an agency with daily AFIS responsibilities, however retention of candidate lists was a subject of considerable debate when I was there. At the time I left, all lists containing a 'Hit' were retained (the complete list: 15, 20 or whatever number the examiner had chosen for that case), and a screenshot of the actual onscreen comparison of the 'Hit' was also retained. These 'Hit' records would only be required to be retained with the case file if the 'Hit' was confirmed as an identification by original hardcopy comparison.

When no 'Hit/ID' was made, no printout of candidate lists was required, however individual examiners were 'allowed' to maintain a personal file copy of those records if they so desired. The negative result candidate lists could not be included with the master case file. The AFIS vendor was not contractually required to maintain records of past individual candidate lists, and claimed they did not have the capability to produce those records.

The debate centered around an agency guideline requiring that all work done on a case be documented. How should an examiner testify in court if asked if he/she had done any other latent print comparisons in a case, besides with the defendent? If there were latents entered in AFIS for an extended period of time, there is a good chance that additional comparisons would have been done. If the examiner did the comparisons, shouldn't it be documented? And should the defendent not have the right to review those comparisons if they so desire? QA indicates that missed identifications may (will) occur. Yes, the system itself may miss scoring an individual high enough to be placed on a candidate list, even when they are in the database, but does that excuse human error as well? Transparency would seem to suggest those records should be avaiable.

DISCLAIMER: The preceding represents my personal recollection of practices at a single agency more that 5 years ago. While I believe it to be accurate, I have not reviewed notes or other references to verify my memory. I have no first-hand knowledge of what practices may be in place today.

Regards, RLT

RL Tavernaro wrote:And should the defendent not have the right to review those comparisons if they so desire? QA indicates that missed identifications may (will) occur.

An excellent point, Bob. But then, your responses are always well thought out and enlightening. This struck a chord with me for two reasons. First, as you know, I do a little defense work on the side. Second, I once failed to see an ident on a #8 AFIS generated candidate (not at the agency where I work full time). In that case, all candidate lists were reviewed by two CLPE's and the other examiner caught the ident I had missed.

It is safe to say that AFIS latent print operators occasionally fail to catch a correct identification somewhere on the candidate list and I believe the further down the list, the greater the chance it would be missed. It is also safe to say we go to trial often when a defendant's prints have not been identified in the case and we did an AFIS search without making any other identifications. Then how often do we go to trial in cases where we compared an AFIS search to a candidate list but failed to identify the real criminal's fingerprint and an innocent person is on trial?

Lest defense attorneys reading this get the wrong impression, I would guess the answer to the question in the last paragraph is one in many millions. Fingerprints are still the most reliable means of identifications, but even the best experts are still human.

Thanks, Bob. Great question!

P.S. -- I have voted "yes" to the question. We do retain all AFIS search lists and put them in our notes.

Pat Wertheim wrote:
Then how often do we go to trial in cases where we compared an AFIS search to a candidate list but failed to identify the real criminal's fingerprint and an innocent person is on trial?

Thank you for your response, Pat. And especially for picking up on my unspoken question. While it is my hope that your estimation of the rarity of such occurrence is accurate (& I tend to agree), that would be small consolation to the innocent individual facing charges without even the opportunity to find that potentially exonerating evidence.

Although a new AFIS search might be considered, I have been involved in a high profile homicide case in which just such a request was made. The agencies involved fought vigorously in opposition to repeat an AFIS search that had already been done. Even if a new search was done, the database changes over time, & someone who showed up originally at #14 on a list of 15, may have fallen out off the list today.
Regards, RLT

RL Tavernaro wrote: . . .someone who showed up originally at #14 on a list of 15, may have fallen out off the list today.

Just as likely the points entered on the original search would be entered slightly differently and the correct candidate would show up higher on the list. I know of at least one other private expert who has sought court orders to go into police agencies and search the latents himself on the police department's AFIS. Needless to say, he has not been well received. But for the most part, defense experts do not have access to AFIS. I am sure some enterprising defense attorney will eventually file a motion with regards to unequal resources being stacked in favor of the prosecution, thus denying the defendant a fair and impartial trial.

I have had personal experiences with re-coding a fingerprint and getting different (better) candidate responses. Case in point, just prior to upgrading from Motorola's (Printrak) AFIS 2000 to Motorola BIS, I searched a very decent quality latent print. No hits with the candidate list in AFIS 2000. I re-searched the print (using some automatic minutiae and some I encoded) on the new system and some of the old candidates that appeared on the old system but the top 10 candidates were completely different and the top candidate picked by the new system was the one!!! I know there were new algorithms and better matchers, etc. on the new system but I really don't have a problem re-searching a latent if I previously didn't identify any the first time. Afterall, as Pat said,

Just as likely the points entered on the original search would be entered slightly differently and the correct candidate would show up higher on the list.

That being said, we retain the candidates on the 'hits' and include them in the case file but all searches can be retrieved if necessary so I can really say that even if I don't print them off the day that I search it, I can still get the results. It's much the same reason why we keep mugshot line-ups and the candidates that were used to create the line-ups. If several witnesses pick the "wrong" guy (but all the same candidate), maybe "he" should be looked at as having some part of the investigation! (IMHO)

Hi All,

We also keep all the lists of candidates we search. I think we do it more to see when there is a miss (system or person) we can try to figure out why.

While we are on the topic of candidate lists, how many candidates do most of you search? My supervisor likes to see 30 candidates searched. Once, I actually got an ident at # 28. In my experience, if there is an ident, it comes out in the top 3 and usually it is first.

Also, how many searches do you do? We do 3 searches - statewide, county, and county and surrounding counties. Again, in my experience, if I am going to get an ident it come out first and on the first search.

Becky

First, a caveat: I am now 5 years removed from working at an agency with daily AFIS responsibilities.
That being said, the lowest candidate on a list I can recall making an ID on was #14, even though I frequently searched 30 candidates. Colleagues did have hits in the mid 20's. Most of my hits were in the top 5, with probably a majority as #1.

My recollection is that the minimum standard search was with 15 candidates, however the examiner had the option of searching more. The first couple of years I searched 30, however never hit below #14. I then switched to 15, except in special circumstances (poor quality of latent, few characteristics, etc.) when I would expand the list to 30.
Regards, RLT

Edit