SCRO

[Thomas Taylor]

http://news.bbc.co.uk/2/hi/uk_news/scotland/6512839.stm

[redlion62]

If it`s true that 6 scientists have lost their jobs over a mistake then it is a sad day for them and for their families.
Even sadder though is the damage done to their profession and to the reputations of the fellow practitioners they leave behind.
By continuing to refuse to accept that they even `might have been wrong` they almost brought their science to its knees. That arrogance has got them to where they now appear to be...unemployed.
For those within the profession who continued to criticise their stance this should not be a time for celebration.
There are still those within the organisation (SFS) who are on record as being loyal to the six and to their conclusion. They should now move on and accept their role in a new Fingerprint Service for Scotland.
There is still another hurdle to overcome...the Asbury Claim. The removal of these six will limit the damage this may cause if not decide its outcome already.
Let all of the Fingerprint Community now go forward with confidence in a future of truth and integrity.

[sharon cook]

I'm sad for group of people who were so pigheaded they could not admit they were wrong even when it was clearly demonstrated to them. And, yes, let's not forget that there were TWO mis-ID's in this case, not just one. I think the new bureau will go on stronger and with more credibilty now that the "tainted" ones are gone. Of course, this case and the Mayfield case in the USA will give fodder to defense attorneys for decades to come.

[David L. Grieve]

As one directly involved in this sorry situation, I have little personal satisfaction in the removal of the six other than my faith that justice does prevail if given enough time. I find that reaction somewhat strange, for I stood outside the court house during recesses within earshot of three of the people and bit my tongue at the comments made by them during Pat Wertheim's testimony. The conversations will never go down in history as professional and the comments were reprehensible. I wrote them off that day as being worthy of little compassion should their time come, and I think that is why.

I try to instill in my students that confidence and arrogance are separated by a thin line. The world has witnessed what can happen when that line is crossed. But beyond the examiners were those who ran SCRO and not only allowed this to happen but promoted the ugly smears and innuendoes that are still repeated, even on this site. They were not interested in justice nor in determining the truth, and such arrogance in management is more distasteful to me. Those people have moved on or are safely retired, and will face no consequences other than a few quickly forgotten footnotes in a committee report. I am more saddened that the ax falls only on six when so many are parents of the injustice. Had just one prudent manager called a halt to it all, this would never have happened.

I, too, wish the new service well, but part of me fears a systemic problem that has not been touched, let alone cleansed. I recall a class on ethics that stressed ethics are caught, not taught, and these six did not more than what was allowed or encouraged. Until a thorough inquiry seeks to determine the full cause of this travesty, the potential for another injustice remains. If I lived in Scotland, I would take no great comfort in their departure.

After Mayfield, the FBI reviewed operations to find out what could be done to ensure it never happened again. Everthing from training to supervision was examined and numerous recommendations were adopted. SCRO sacks six, hardly an exhaustive review of how something lingered and festered for nearly a decade.

[adroitcaledonian]

Smears and innuendo?
What was it you heard during the alleged recess conversation that alarmed you so?
Present specifics or are you not guilty of the same smear and innuendo?

One thing that I think we will all safely agree on, albeit for extremely different rationals, is that this case has smeared the credibility of fingerprint evidence.

If one accepts that the SCRO experts were wrong (which I absolutley do not for one fraction of a nanosecond) then it's damaged anyway.
But what if the experts concerned weren't wrong?

It's a strange system we have here in Scotland, the experts in unemployment can't speak in their own defence despite no wrong doing, Shirley -who appears to have misled at least one jury about her knowledge of her own expert- gets what amounts to more than many people earn in a lifetime for declining to pursue her case and is allowed to write a book about it.

David, you seem to feel that the truth hasn't came out, please tell us what the truth of this case is, or are you content to mire in smear and innuendo?

[H. B. James]

I was a late comer to this series of discussions because it seems it must have been going on for years. Isn't there a book coming out about it? It would be easier to read it in an organized fashion than to go back and try to find all the various things said about it here. I tried amazon.com but couldn't find anything. Is it out yet? What would the title be? Where can it be bought?

[Iain McKie]

Dear H B,

Thanks for your interest.

You will see from the above postings that the debate has been lively. Unfortunately anonymous contributors like ‘adroitcaledonian’ have chosen to smear respected experts like David Grieve but then they have little choice given that they have totally lost the debate. In this debate fingerprinting as a valuable forensic science has been the winner and those experts who would seek to undermine it have been the losers.

Full information on ‘The Price of Innocence’ that is being published next Tuesday 17 April can be found at www.shirleymckie.com

The best deal is via ‘Waterstones’ at

http://www.waterstones.com/waterstonesw ... ku=5604690

It can also be purchased via Amazon

http://www.amazon.co.uk/Shirley-McKie-I ... 136&sr=1-1

‘The Price of Innocence’ is also being featured shortly in a UK national newspaper. Further information will be supplied shortly.

A comprehensive online index to the book showing many of the associated reports and documents, and an online forum for comment about the book will soon be available through www.shirleymckie.com

Please use the forum to let me know what you think.

Best wishes,

Iain

[Daktari]

What a cheap stunt to plug a book!
Wonder if it will answer the question: Why did Shirley lie under oath at her Perjury trial? Somehow I don't think it will!

[Outsider]

The only way to put this matter to rest is to find a way that results in everybody - including the public - knowing and accepting the truth. The Scottish Parliament inquiry has achieved little and I doubt if a judicial inquiry will do much better. What is needed is a scientific inquiry.

Dr Dror’s research is directly applicable here. Nearly everyone who has looked at Y7 or QI2 has done so with far too much baggage in their head. We need a full scientific investigation into the McKie / Y7 and Marion Ross / QI2 matches with experimental design by an experienced researcher. The results will need to be interpreted by a research statistician. Critical review by other scientists will be a necessary step and only then will we be reasonably sure of what we know and what we do not know.

I imagine that the research method would involve sending the images to a large number of experienced fingerprint experts in different parts of the world along with a number of control (decoy) image pairs from known sources. As you probably know, I believe that there is an increased theoretical risk of error in the McKie/Y7 match compared with a normal crime-led case.

http://www.stevehornsc.pwp.blueyonder.co.uk/short.htm

We need to rule out the possibility of a random erroneous match so the controls must include a number of skilfully chosen “closest possible” matches from large databases with sources that are known to be absent. Maybe the opportunity could be taken to test Dr Dror’s findings by giving some of the experts background information about the images (of course, the rest should be working blind).

The objectives of the research should be:

a) To publish the results returned by the fingerprint experts (raw data)

b) To analyse variation in fingerprint analysis – decision, number of matching points, discrepancies etc., by region, experience, academic qualification, methodology, background knowledge, source type, or anything else. Tell us if any of this variation is “statistically significant” (i.e. more than the natural variation in the fingerprint analysis process).

c) Tell us how much natural variation in results is produced by the process of fingerprint analysis by experienced experts (we can expect some variation in the number of matching points but there should not be any cases of ‘identification’ and ‘exclusion’ for the same images – if there is, we need to investigate why).

d) To tell us how much certainty can be attached to the Y7 and QI2 identifications.

e) To tell us if the experiment gives any explanation for the differing views on these two matches (bias, incompetence, “honest mistake”, criminal intent, etc)

To those of you who say that all this is unnecessary because the truth is simply that the SCRO officers are incompetent - or criminal - then I have to remind you that the Scottish Parliament has had a year long investigation and has NOT come to this conclusion. Pauline McNeil, convenor of the Justice 1 committee said "Fingerprint analysis is not a science. It is subject to human error” and pointed out that the committee found it "staggering" that experienced experts could hold such widely divergent views.

http://www.theherald.co.uk/news/news/di ... 75.0.0.php

I think that pubic confidence in fingerprinting will be greatly enhanced if the profession is seen to be taking the lead in getting to a “generally excepted” truth about the McKie case by commissioning this research.

Finally, on a separate point, I would like to say that am very unhappy that the retiring SCRO experts have been forced to remain silent. I cannot see what that this achieves or who benefits from it.

[Thomas Taylor]

Pauline McNeil, convenor of the Justice 1 committee said "Fingerprint analysis is not a science. It is subject to human error”

I might suggest that every field of science, no matter how pure or exact, is subject to error when practiced by humans.

Outsider, there is a basic flaw in your analysis of this situation. You recommend letting non-fingerprint experts and statisticians determine the correct meaning of Y-7 and Q-12.

Let's look at an analagous hypothetical example. If two groups of mathematicians disagreed on the solution to a novel mathematical question or equation, should we find a group of people who know nothing about mathematics to solve the problem?

Even in a simple mathematical problem, human beings make mistakes. The commonly cited example is a checkbook register. If a bank account does not balance, would you take the register to a person who knows nothing about addition and subtraction to find the correct bank balance?

No, Outsider, the people to decide this issue should be other fingerprint experts. Do you really think our science is so subjective and inaccurate that it is impossible to determine a correct answer if we have previous knowledge of someone else's conclusion? If so, then perhaps you should take a fingerprint course or two. Or let me ask it this way -- How would you accept criticism of scientists who make mathematical errors by outsiders who had never studied mathematics? Should the critics not study a little math before they throw stones at mathematicians? And if a mathemetician (or two mathemeticians, or six) gets married to the idea that 2+2=5, should the science be banned? Or should the mathematicians who are wrong be sent to retirement?

I'm sorry, Outsider, but you really are outside of your field here. You are certainly entitled to your opinion, but it would behoove you to learn a little about the subject matter before becoming too vocal a critic.

[Outsider]

Thomas Taylor wrote:

Outsider, there is a basic flaw in your analysis of this situation. You recommend letting non-fingerprint experts and statisticians determine the correct meaning of Y-7 and Q-12.

In any science if there is disagreement which cannot be resolved by reasoning (and it should be evidence-based reasoning not just assertion), you design an experiment to demonstrate the truth. If there is variation or inconsistency in the experimental results you need a statistician to tell you what you know and what you do not know from the experiment. Medical doctors devise new treatments, researchers (who are not necessary medical doctors) design experiments to trial the new treatments and statisticians tell us if the treatments work. And the tests need to be blind because scientists have proved that every judgement is subjective.

If in a small 100% blind trial no fingerprint experts say that McKie is identified from Y7 (including experts who have the type of training used by SCRO) then the matter is finished and you don't need statisticians to interpret the results. You simply need to publish the results explaining how the test was carried out (taking special care to point out what you did to prevent bias) and then everyone MUST accept that the SCRO were wrong.

[Pat A. Wertheim]

Dear Mr. Horn,

As one of the parties in this mess, I agree with your suggestion that a blind test taken by proficient fingerprint experts would settle it. I doubt you will find very many experts who have not heard of it, but I believe you would find a reasonable number of good experts who have not themselves actually seen or examined the various latent and inked fingerprints at issue. If the prints could be submitted to such a group of experts in a blind testing situation in which they not only did not know who said what, but did not even know the prints were from Scotland, I believe their results would be fully reliable.

I believe the Justice 1 Committee was too political a group from which to expect a definitive ruling. But I disagree with you on the value of a Judicial Enquiry. I believe a Judicial Enquiry should be authorized and I believe the first task such an enquiry team should undertake would be to find the suitable number of proficient experts to do the blind test you suggest. Of course, the enquiry team would not find the experts themselves, but would go through an intermediary and would have the fingerprints submitted to the experts in a routine business way, depending on the agencies at which the secretly selected experts worked.

I would say that such a test need not be blind, because I firmly believe that any group of experts who are not emotionally involved in this situation would reach valid conclusions. Our science is not nearly so subjective as you might think, based on Dr. Dror's article. But, for sake of argument, I would agree that a blind test would carry more weight to those who do not understand fingerprints.

I am totally confident that such a test would confirm what I have said all along -- the fingerprint on the doorframe (Y-7) was NOT left there by Shirley McKie and the fingerprint on the Marks & Spencer sweets tin (QI-2) was NOT made by Marion Ross.

At that point, it would be up to the Judicial Enquiry team to determine what happened and what courses of action should be taken. I think there are a number of critical questions: 1) Were the misidentifications honest mistakes or was something else going on? 2) If there was something other than honest mistake, who knew what, and when? 3) Were there other mistakes in other cases? 4) Who killed Marion Ross? 5) Did the changes at SCRO properly address all of the concerns or are additional changes needed?

But the biggest question of all in this politically charged situation is how do we bring about this Judicial Enquiry in the first place?

[adroitcaledonian]

One would hope that after the Scottish elections if the SNP are in power, as looks likely, that they will honour their pledges over the years to hold a judicial enquiry.

Although I am concerned about the conduct of Alex Neil in all of this I am encouraged by the fact that the two SNP members of the Justice One committee didn't contest the committees unanimous report.

I am not a supporter of the SNP, and I agree that the whole matter cannot be left to politicians.

However Pat can you remind me of the politcal background of the co-author of the McKies fothcoming work of fiction? Can you remind me how this co-author came to be involved in the case?

[Michele]

Steve,

There’s a lot of misconceptions about science and statistics. If we put both subjects together the misconceptions rise exponentially.

Statistics don’t prove anything. They may add additional support or refute different ideas, but they don’t provide proof. Also, statistics can be done in a variety of ways, creating different answers. Whether or not they produce valuable information is dependent on if a particular method actually applies to the situation. Scientifically, this is determined by peer review.

If there’s variation in results then a statistician may be able to tell you how much variation to expect but they might not be able to tell you why this variation exists. This could be important because why the variation exists may change the likelihood of the variation happening to any particular person (or in any particular situation).

With regard to fingerprints, there probably haven’t been enough studies done for statisticians to determine an accepted rate of variation for statistics to be useful.

Blind testing isn’t a scientific requirement nor is it recommended in all situations. People who like to use blind testing as a safe guard against bias don’t fully understand scientific requirements. If they did understand scientific requirements they would realize that science has other methods to diminish bias that doesn’t eliminate the peer review process. Can you imagine how effective peer review journals would be if they stated:
“Don’t give any information about your research, it could bias the reader! Just state the problem and your conclusion, if others can reproduce the result then it’ll be accepted as scientifically valid”.

Or maybe a college physics test stating, “Don’t show your work! Points will be deducted for attempting to bias the grader”

If 100 examiners all independently excluded John Doe from leaving a latent at the scene of a crime, does that mean that John Doe didn’t leave the latent? Is the matter finished? Statistically you could say ‘Yes’ but scientifically the answer is ‘No’. Science continually leaves conclusions open to be re-evaluated for better conclusions. Although statistics is a science, a high statistical probability doesn’t give a conclusion scientific validity, it just gives support to the conclusion. What determines the scientific validity is the justification behind all the examiners conclusions. What I’m trying to get at is that conclusions from blind testing may give statistical support but only justified conclusions that are open to peer review give scientifically valid conclusions. Suppose 10 years from now a brand new examiner found that the latent was left by John Doe and showed everyone the justification behind it. Lets also suppose that once everyone saw the ID they agreed. Did examiner 101 bias everyone or did examiner 101 produce objective justification behind their conclusion? The justification will determine the validity of the conclusion.

This situation isn’t isolated to the Mckie case. I could list several cases where many skilled examiners erroneously excluded a latent and then later the latent was positively ID’d to that subject. If we assume that blind verification is the answer then many ID’s will never be reproduced (simply because someone can’t find it, not because it isn’t an ID). It’s not how many examiners find the answer that’s important, it’s the justification behind the conclusion.

[Outsider]

I agree with just about everything in Pat Werthiem’s post. A Judicial Enquiy that calls for a new apprasial of the identifications from a broad spectrum of fingerprint experts is the ideal solution. To Adroitcaledonian (and anyone else who is supportive of the SCRO experts) I ask this. If 20 experienced experts from different countries working blind return either “exclusion” or “inconclusive” for the Y7 to McKie match, will you then agree that McKie did not deposit Y7?

My call for some control comparisons from database searches would only be needed if there is a number of experts returning “identification”. I am worried about the theoretical heightened risk of a random erroneous match due to the fingerprints just being very very similar (compared with any one crime-led case) because a match that is not connected with an pre-existing crime is self-selecting from an unlimited pool of comparisions. With the controls in place we would be able to compare the results of the Y7 comparison with comparisons where the latents are very very similar to the inked prints from the database, but are known to not come from the same source. Maybe that could be done as a second stage if required. Statistiacal analyis would only be needed if there was argument over inconclusive results.

The place were I take issue with Pat is here:

I would say that such a test need not be blind, because I firmly believe that any group of experts who are not emotionally involved in this situation would reach valid conclusions. Our science is not nearly so subjective as you might think, based on Dr. Dror's article. But, for sake of argument, I would agree that a blind test would carry more weight to those who do not understand fingerprints.

I don’t think you can dismiss Dr Dror’s research so easily. In science research ALWAYS takes precidence over opinion, nomatter how much experience goes into that opinion. You can argue that the research was not carried out properly and you can publish your own research that shows different conclusions but a scientist can not just ignore research because it appears to contradict his or her experience. This is from Dr Dror’s latest research (he has carried out 2 studies into this matter).

The research reported in the present study replicated and expanded our previous findings and found that across eight comparisons made by each of the six participants, two thirds of the fingerprint experts made inconsistent decisions to those they had made in the past on the same pairs of prints. The findings of this study not only further substantiate the vulnerability of experts to contextual effects within a larger data set, but they further contribute to our understanding of this phenomenon.

Our data demonstrate that fingerprint experts were vulnerable to biasing information when they were presented within relatively routine day-to-day contexts, such as corroborative (or conf licting) evidence of confession to the crime. Thus, contextual information does not need to be extreme and unique to influence experts in their fingerprint examination and judgement.

The full paper is here http://users.ecs.soton.ac.uk/~id/JFI%20 ... 0error.pdf