| Author | Message | ||
     Haley Kuther |
Is it possible for you to alter your thumb print when you are cashing a check at a convience store or any place that does a thumb print? | ||
| Tim Seguin |
Thanks to all who answered this question. It was very helpful. | ||
| Graham Ford |
I doubt that he stats generated in trying to achieve the ideal, of an exact figure of comparisons undertaken will probably ever be achieved. There will always be variations on how we record our stats from one agency to another. The following is an example of one of my cases from back in the GB. 106 assosrted lifts and photographs that contained 180 separate finger & palm latents. 16 suspects nominated for comparison. 20 police officers nominated for elimination. 16 suspects + 20 police x 12 (10 digits & 2 palms) x 180 latents = 77,760 comparisons (max. Were all those comparisons done - no. As expected some were identified early in the comparison exercise. But how about the comparisons undertaken by the verifyer, are they recorded? When I was back in London 3 examiners were utilized (standard procedure) and all latents identified or not identified were compared by a 2nd examiner. A 3rd examiner was the verifyer, who if they selected, could have also compared all the non ident latents in addition to the identified latents. So what was the final figure of comparisons undertaken - who knows. Are we trying to over analyze the volume of what we do - well yes. I believe that Kasey is right, the knowledge, skills and experience we gain through our endeavors is a factor not only in how much we complete, but on how we go about completing the comparisons tasks. When you start out in this career you start comparisons by looking at the first latent against the first print card, starting at the right thumb and read the print card like a book. As you gain experience and confidence, you select a "finger choice" for the latent and just compare those to begin with, to short circuit the process, saving time and your eyes. Just in reading the book we are trying to skip to the exciting final chapter (get to the meat) and leave out all the preceding dross. Will latent stats ever be an exact science? Maybe we just need a Daubert challenge purely for our comparison stats. What more!!!!! | ||
| Steve |
Ernie, Regarding your double/triple check of cards, this reminds me of a discussion that Michele and I had a couple of days ago. She brought up the following scenario: If you've got a delta area that you aren't able to determine if it is from the palm or a finger you'll search all delta sources available for the subject. If an ID is made for example, to the carpal delta of the right hand, our statistics would indicate 2 comparison (one to each palm). Once ID'd, this would be determined to have been a palm print. Thus only 2 palms could have been compared. Yet there could have been a lot more comparisons done, to include all the deltas on each finger plus the interdigital deltas and the carpal deltas. These "two" comparisons could have taken a lot more time to complete than the "500" I mentioned in the earlier scenario. | ||
| Ernie Hamm |
The method of counting productivity figures by Mississippi (as stated by Kasey) has been used in variations for many years in other agencies. It is a simple formula, number of examined exhibits plus number of latent prints multiplied by record prints equals a productivity figure. True, it can be misleading, as stated in other posts. If you finalize the comparison with the first finger in the first card of a stack of “X” cards, you did not really make the reported comparisons. However, if it is a difficult latent that requires you to double/triple your comparisons through the stack of cards, you don’t get credit for the extra comparisons. It balances out. I kept many statistics during my tenure and there are several that are not routinely used, or at least acknowledged, that provide more meaningful input than ‘productivity’ figures. These involve the efficiency of the individual examiner and the agency. While the examiner may have had 100 examinations in a case, how many latent prints of value were developed (recovery rate) and how many were identified (identification rate) in a case? There are mechanisms to obtain and record this information, sometimes much to the chagrin of examiners, but they are not routinely used. If this information is developed, a supervisor can report that the section develops latent prints in X % of their cases and makes identifications/individualizations (I acknowledge a term preferred by some individuals, but reserve my use of a traditional term) in X % of cases. The examiner can also know their individual figures. This information is important to evaluate and support the efficiency, and proficiency, of an individual examiner and section. It is nice to report how much work you are doing, but how well are you doing it? I openly recorded this information for a number of years and I know there are others in supervisory roles that ‘unofficially’ documented these statistics, but the procedure is not commonly practiced. | ||
| Steve |
We call it "AFIS Math".  | ||
| Kasey Wertheim |
Steve, It almost seems like the mechanic shop. When you take your car in to get it fixed, the job is supposed to take X number of hours. If they get it done sooner, they still get credit for X hours. And guess what... you still get billed for X hours! The deal is the expertise... if they have knowledge that allows them to get a job done faster, but they had to pay for that knowledge (through training, experience, etc.) they are going to still charge you for the entire job, because whether it is knowledge or hard work, it still costs $... and they pass on that cost to the customer. So sure, we know that central pocket whorls are most likely on the ring finger, so we start there first. That's something we had to learn from training and/or experience, so we should get credit for the other 9 comparisons, right? Is there a term for this type of logic? ) -Kasey | ||
| Steve |
Kasey, I emailed this to Tim earlier, but it's in line with what you said: As an example, you have 10 prints of value for comparison and are given prints of 5 different suspects. We’ll just assume that they all are fingers. Three of them are simultaneous and appear to be from the right hand; index, middle, and ring fingers (based on their positions relative to each other). So you start your search based on pattern type, choosing the one determined to appear to come from the middle finger, based on your assumption. You come across the third card and make your first match. You subsequently individualize the other two prints to the index and ring fingers. So you’ve looked at three middle fingers until you made a match, then you made two more comparisons to finish the prints from that latent lift. Essentially that was only five comparisons. (You’ve initially excluded the other nine fingers in your initial search, but had no match been made you would have directly compared them.) The other seven turn out to be the right thumb of top card in your stack (which was the one you made your previous three ID’s to). Assuming you start at the right thumb and would have continued to focus on the one card, especially after having made an ID to it, you’ve now made a total of 12 comparisons. That’s the reality of what you did in this hypothetical. But in our statistic-keeping method, we would say that there were five suspects, multiplied by their 10 fingers, equals 50 known finger impressions. Then, those were compared against the 10 latents. The result is 500 comparisons… Funny how 12 = 500 some times, huh? | ||
| Kasey Wertheim |
Tim, The Mississippi Crime Laboratory collected information concerning the number of exhibits, number of items in each exhibit, number of latent prints of value found or developed on those items, the number of known prints (fp card = 1, and palm print card = 1). We multiplied the number of latents and the number of known prints to arrive at an estimated number of comparisons. Many argue this isn't accurate, because if you identify a print you don't compare it to the remainder of the known prints. Further, if you identify someone on an item, they are the first person you look at for the print next to it. But "comparisons" is a general indication of case difficulty, time spent, etc. So we still collected the information. FYI SWGFAST is looking into the definition of statistical categories and recommended practices for collecting a variety of case information. -Kasey | ||
| Tim Seguin |
Does any agency require their Latent Fingerprint Examiners to maintain records as to how many comparisons they do? If so, what information is gathered? I'm considering a similar situation at our agency, which is in the early stages of a forensic ID program. |
