Sufficiency Coin Analogy

[Boyd Baumgartner]

At the IAI this year, I heard an analogy regarding the reason there's no point standard to reach sufficiency. The analogy is that in the same way you can make change for a dollar (or any other denomination) in multiple ways: 10 dimes, 4 quarters, etc. Aside from the fact that it's a 'pat you on the head and give you a sucker you nice little juror' and in no way captures the actuality of valuing information in the impression, it always makes think of this:


http://www.hulu.com/watch/4258/saturday ... nge-bank-1

[Michele]

in no way captures the actuality of valuing information in the impression.

I’d disagree. I think it captures it better than saying ‘I’d have to see the print’.

[Boyd Baumgartner]

Why is 'I'd have to see the print' the alternative?

Once you bring up a value analogy as is the case with money, one had better be able to articulate a method, philosophy, etc on how who or how one sets the value of information in an impression. I also don't also don't find discussions on clarity or rarity particularly enlightening and actually beg more questions.

For instance, more clear information isn't always more valuable than less clear information. I've also yet to hear anyone articulate how they consider rarity in a way that then explains their ability to discern impressions run through AFIS when they've effectively diminshed the role of rarity because AFIS is returning 'equally' rare impressions.

[Pat]

Once you bring up a value analogy . . .

An analogy is a scheme of explaination that uses a particular trait or characteristic of a simple or familiar thing to help understand a more complex or unfamiliar thing. The problem is that the analogy usually only helps explain one or two narrowly examined traits. With over examination, the analogy falls apart because in other ways the simple thing is quite dissimilar from the complex one.

If the analogy of making change is used to explain value assessments of features in a fingerprint, I was tempted to pop off that in one case frequently discussed on this forum, the police agency in question short changed two innocent people. In another frequently discussed case, that police agency substituted some plug nickels and a couple of washers.

But, alas, there is no perfect analogy.

[latent_printcess]

An instructor once used this analogy:

"Let's say you own a black, 2006 Honda Accord EX sedan. You could take a number of Hondas of the exact same color, year, trim and model - remove the license plates - and you could still pick out your own car. It has scratches, dents, etc. which make it unique."

Again, not a perfect analogy (e.g., the exemplar is in your head), but I thought it was fairly descriptive...

[Tazman]

Don't forget the VIN.

[L.J.Steele]

"Let's say you own a black, 2006 Honda Accord EX sedan. You could take a number of Hondas of the exact same color, year, trim and model - remove the license plates - and you could still pick out your own car. It has scratches, dents, etc. which make it unique."

Only if you are the kind of person who never starts to open the door of the wrong similar-looking car in the parking lot (esp. at Christmas-time).

More seriously, this is a bad analogy for two reasons: (1) the mental mechanisms for recognizing objects in general is different from faces. As I recall Dr. Busey's research, experienced examiners use that different facial recognition area for print analysis. This is very much apples and oranges, if not apples and ocelots.

(2) This may cause problems for the case if identification, particularly identificaiton of an object (like a car), is otherwise an issue in the case. The analogy could get you caught up in disputes about misidentification, or be deemed a comment on the credibility of other evidence. Unlikely, yes, but not worth the risk.

Analogies are dangerous and can echo in weird ways in a case. Stick with the plain facts, in some cases you need more information than others because .... and talk about clarity, arrangement of features, etc.

[latent_printcess]

All excellent points!

[sandra wiese]

This same analogy was published in the JFI, vol. 55, no. 3, p. 291 from 2005 where Paul Ramsey, CLPE from Arlinton Police Dept. wrote:

How many times has a defense attorney asked the question, "How many points do you need to make an Identification?" I don't know about you, but for me it has been a lot, especially after the Daubert case. I would answer, "There is no set number", and then try to explain quantitative-qualitative friction ridge analysis. However, I always felt I was getting too complicated or technical for a layman jury. Then, while on my way to court last week, I realized there was a simple way to explain quantitative-qualitative analysis. I would say, "There is no set number", and then I would ask the judge if I might use an example to explain my answer. If he said, "Yes", I would say, "It is like coins in the pocket. If the defense attorney were to ask me for a dollar and all I had was change, I would pull a handfull of coins from my pocket and drop them one by one into his hand until he had a dollar. Now, how many coins did it take to make that dollar? The potential number of answers would be in the hundreds! It could be a hundred pennies, a single silver dollar, or any combination of coins in between. There is no set number! It all depends on the value of the coins used. The same is true in fingerprint identification except the potential number of answers is even greater! The features used to identify an impression all have different values based on their type and clarity. Until I see the exact known and unknown impressions being compared, I do not know how many features I will need to achieve the identification. It is that simple.

[Angie]

I heard the analogy of points described using the points at the end of the fingers, the end of the toes and the top of the head. If two people have all their fingers and all their toes, they each have the same 21 (counting the top of the head) points. Therefore, if all we took into consideration were points of similarity, these two people would be identical. This is supposed to show that it is not the points that matter but what is in between the points that really count.