Criminal Investigators

[Charles Parker]

I have been reading some lately how Criminal Investigators are forcing the poor forensic examiners into making choices or decisions based upon their investigation and not upon the evidence. In other words they are brow beating or interjecting bias into the forensic examination.

Now I am not going to state here that the situation described above never happens. I am sure it does on occasions. I am also not going to state that it happens quite often. I personally think that it happens very infrequently. But I also think that the factual situation is that no one can tell you with any degree of certainty how often it does happen. There are no studies or research that I can put my fingers on that will state how often an Investigator will interject bias or force their will on someone.

The only thing that has been offered are some anecdotal events related by some who really want it framed in a particular way. Since anecdotal events are all we have right now and so I wish to offer some of my own.

I have been involved in latent print examinations and crime scene investigations since 1979. In that time I can remember just two times where investigators really tried to project their investigative will on the forensic examinations and both times failed. If I state that 2 times against about 50,000 investigative requests I have seen or been a part of since 1979 that it would occur about 1 in every 25,000 investigative requests or 0.00004. I would say not often and in my PPOV quite rarely. Of course this is just anecdotal events that I have observed.

Of course we should define what leading information is and what is not. An investigative request could exist from just a name on a piece of paper or an investigator telling a forensic examiner in person who to compare. In my POV a name on a piece of paper is not very intimidating at all. Probably 60-70% of those named requests made by investigators are a negative match. Not something I would consider concrete.

Most investigators I know are very professional and only wish to solve the case correctly with the correct person. I have seen investigators not act because they wanted just a little more information to be sure with. Two investigators told me recently that they would rather let a guilty person go than convict an innocent person. I normally hear that from forensic people and not criminal investigators.

There are some bad ones though. Just like there are bad forensic examiners who dry lab, or overstate they qualifications or results. Just like there are bad attorneys who bolster their hours or cheat their clients. Just like there are bad clerks who tell you that your lottery ticket did not win anything and then you hear where they have moved to Jamaica. There are bad people in every discipline. I am not keen on blaming all criminal investigators for the bad actions of just a few.

My advice is use your own knowledge, reasoning, and experience in determining whether criminal investigators are trying to force you into coming to their point of view. Keep one eye open for those criminal investigators and those forensic examiners trying to sell you a line of bunk?

[Neville]

Since I joined a crime lab in 1980 I have seen it more often than that.

A long retired and now dead colleague often said, never tell them while they are hanging about, tell them to come back in an hour and you will most likely will have an answer. His thought was that they thought it took a lot more to make an ID and it kept them believing in white mans magic. I have noted that those who want to impress others of their ability are the ones to keep an eye on.

With the advent of triple checking it is very difficult to apply this kind of pressure to three separate experts. But as you say in the smaller labs it would be an issue. Examiners just need to develop a back bone very quickly. Sometimes it is just as satisfying to see their face when you say no it aint his.

[L.J.Steele]

I would hope that deliberate attempts to influence the forensics investigators are rare.

But the bias concerns also worry about much more subtle issues. Investigator comes in and excitedly drops off evidence, mentioning that an eyewitness picked out "Known Ned" at the scene and he's got a good feeling about a lift from the scene. The investigator isn't deliberately trying to bias the examiner. This isn't "intimidating" conduct. But it is potentially biasing conduct. The triple checks only work to the extent that the others are not exposed to the same information, or other biasing information, in the course of their work. I expect this is a particular problem in high-profile cases where the news is reporting every tidbit it can find about the suspects and the crime and it may be hard to avoid extraneous inforamtion.

[Amy Hart]

But you haven't looked at the opposite side of this subtle bias issue...

As far as I am concerned, anything that anyone says to me while dropping off evidence (or even discussing a major case) is taken with a grain of salt. For instance, when we do an AFIS case, we require that prints of people with legitimate access to the items be submitted, so we don't waste time searching victim prints. At least twice a week, a crime scene investigator or police officer will assure me that the elimination prints will be submitted later that day - the next day at the latest. About once a month, they come through with their promise. In major case meetings, detectives will put forth theories of the case that are pure fantasy. What I use that for is to determine what of the thousands of items they have collected might be useful and relevant to the case. Like, in the case of a husband murdering his wife...it makes no difference if his fingerprints are on the bedside table because it appears to have been thrown across the room - he lives there, and is allowed to touch his own furniture.

So, to summarize my long-winded explanation, I believe about half of what I hear, and that half involves things that I can see right in front of me. I am subtly biased against people, because I believe they're telling me what they think I want to hear in order to grease the wheels.

[Michele]

Carl Popper said that if you try to look for supporting data, you’ll always find it, that’s why it’s important to look for any information that might prove your hypothesis wrong (they use this technique on Myth Busters). If I try to disprove my idea, but can’t, then the only alternative is for me to accept my initial assumption.

Suppose I believe that examiners can’t be biased (or that it rarely happens), then I should look for any evidence that they can be. To me more specific, maybe I should first just consider myself (I can do testing on other examiners later).

a) A few years ago I was in court and had to take the suspect’s prints and compare them to the known prints I had with me. The setting for taking the prints was awful. The suspect’s skin was extremely dry and made the new prints unclear. The lighting was awful and I didn’t have the magnification tools with me that I normally use. The subject wasn’t contesting the identification, he admitted that he was the correct person, so I didn’t have to be as thorough as I normally would be. When I was making the ID, I didn’t do my normal process of trying to look for characteristics that didn’t match. I looked for a few points that matched and testified that the prints matched. Of course the prints did match, but the setting made me rush to judgment and not use the protocols that I knew were important. I justified reasons of why I didn’t need to be as thorough as I should have been (even skipping the verification). I think I was biased into believing that certain aspects of a good comparison weren’t needed when in reality, they are ALWAYS needed!

b) What about shopping for a verifier? Have I ever done this? Normally a verification should be an independent analysis, and I’d say it is when the person agrees with my conclusion but what if they don’t. I’m sure there’s been a time when my verifier didn’t agree with me and I didn’t think it was an ‘independent examination’… I thought this person was an idiot. I took the comparison to someone ‘qualified’ to ‘confirm’ my results.

c) Not long ago a single latent print was brought into our office and the detective said that the comparison needed to be done quickly because a suspect was in custody and if the print couldn’t be ID’d then the suspect would be released. The latent was marginal at best. To tell the truth, it probably wasn’t of value. If this had been another case, the print would have been deemed ‘no value’ quickly but I didn’t want to disappoint the detective. I think I had 5 or 6 people look at the print. Looking back on it, I wasn’t looking for people for someone to confirm that it wasn’t of value, I was looking for someone that had the ‘experience and ability’ to give the detective some good news (everyone loves to give out good news).

I think I have enough examples to know that I can be easily biased (even though these stories may not be real but I think they demonstrate real life situations).

Admitting that I can be biased doesn’t have to be a bad thing. If I acknowledge that bias can exist then I can implement procedures or QA measures to protect against it. In our office, we don’t allow comparisons to be done in court (or in the presents of the requestor). We also diminish verification shopping by listing all conclusions on the report, someone who doesn’t agree with the conclusion can’t just be ignored. We also state in our SOP’s that verification isn’t ‘confirmation’ of the results, it’s scrutiny of the entire case (even scrutinizing if enough documentation is done). We have other procedures that diminish bias too. Individually, they might not prevent bias from creeping in but when all the procedures are implemented, it makes for better conclusions.

[Dennis Degler]

Michelle,

Your examples, to me, do not show bias. The fact that, in the first instance, you didn't do a thorough analysis doesn't mean you were allowing bias to guide you. You made your decision based on what you observed. You didn't deceive yourself. You were just being sloppy. But the conclusion was a sound one.

In the second instance, you didn't allow any mindset to guide you to a preconceived notion about identity or suitability of the latent print either. It is obvious to me you were still looking at it with your "eyes open", recruiting five or six others to review the print as well. And in the end, I'm sure the detective received a reliable conclusion from your unit.

I wasn’t looking for people for someone to confirm that it wasn’t of value, I was looking for someone that had the ‘experience and ability’ to give the detective some good news

Michelle, If you had been looking for someone to confirm that it wasn't of value, might you have been acting with bias as well, but in the opposite direction? I'm still mulling over this subject of bias. It's kind of like having a bad temper. You might be able to insulate yourself most of the time from environments or situations that keep you from getting angry. But sooner or later, you will feel angry. But it doesn't mean that you're a violent person, or even that you use unkind words. Bias occurs in the mind, then may or may not spill out into our actions.

[Gerald Clough]

We talk a lot about bias, mostly in response to the issue being raised by theorists. It's a proper topic, since humans are subject to all sorts of subtle biases. But the original question here implies a rather unsubtle sort of "bias," a direct effort to guide the outcome - or at least a statement of hope on the part of an investigator that the outcome will result in success through the current direction of the investigation. But two points...

Because latent print units work in support of investigations, there will, from time to time, be situations where real benefits, maybe even something so important as preventing flight of a dangerous fugitive or even prevention of further death or injury, will prompt the investigator to ask for extraordinary measures. And even more commonly, a special effort may make it possible for the investigator to resolve the case by acting promptly, an example being given of getting an analysis while a suspect is in custody. (Having an previously untouchable suspect in custody on another matter can be the most fortuitous of events in an investigation, so long as you have what you need to break the suspect down while he's in hand.) I believe that at times like that, part of being expert professionals is being able to invoke uncommon but reliable procedures. It's not "bias" to step outside the mechanical standard process. Regardless of whether you think think this is ever proper, it's not bias, for instance, to throw a case up on a display for a "committee" examination because, if the suspect is the right suspect and can be ID'd very quickly, great harm can be prevented. Expert professionals know the hazards being introduced and should be able to deal with them. The ability to get the greatest reliable yield out of the evidence is part of expertise. Sometimes that means things become pretty dynamic and that unanticipated issues arise and have to be resolved quickly. I believe that when you signed your expertise on in service of critical missions, you signed on to deal with those things at need. The point is, there's a difference between mission pressure and bias. Mission pressure does indeed have to raise your awareness of bias, just as any stressor should cause you to become more aware of its effects on perception. But handling bias is just part of the expert task.

The second point, and this was touched in an earlier reply. The original question was about pressure from investigators. What we have to remember here is that, on this forum, we often share two characteristics that we can easily forget do not apply to large numbers of examiners. One is that we are here talking about it at all. We reading about the issues and thinking about them. A lot of us have picked apart what's known about some known errors until we were satisfied we had some decent idea what happened. The other is that most examiners who post here work in well-structured and fairly well insulated lab units. But there are a great many official examiners who entered practice with minimum training (some with less than minimum) and over many years of practice have never read anything else, never had any other training, never talk with other examiners, have no discussions. They do not recognize the name Mayfield, don't know who Cowans was, never heard of McKie or the SCRO, and have only the vaguest awareness of the NAS Report. Many of the older ones would take "ridgeology" to be a made up joke term, if they were to hear it. This is not just the occasional oddity. I know a number of them. (I painted that a bit bleak, but it's not all that rare. And I've become aware of practicing CLPE's who aren't much better off.)

Often, they are detectives in small agencies (and remember that "small agency," 12 officers and less, is the definition of U.S. law enforcement) and are doing examinations in their own investigations. To be fair, a good many are aware that they are not current and properly should not be publishing official conclusions. And many small-agency detective-examiners are very aware and keep up with the field and proceed with a high awareness of the hazards imposed by the local structure. The point here is that we should not view bias questions strictly from the point of view of the tuned-up, trained and retrained, supervised and annually tested, aware and thinking standards-compliant lab examiner and end up nodding to each other and agreeing that all this bias talk is mostly smoke and not something that should be considered any kind of big deal in fingerprint work. The real latent print examination landscape, at least in the U.S., is a lot more complex than the ideal lab unit model. I suspect that whatever we can learn about bias eventually goes toward making the discipline what we'd mostly like it to be.

[Ernie Hamm]

I would like to make some comments about all these “bias” confirmation decisions associated with “criminal investigators”. My first statement is that I was a ‘criminal investigator’ before I was a latent print examiner. In fact, during my 10 year assignment with the CID Laboratory, I was still considered a ‘criminal investigator’ with a military skill designation as laboratory examiner. Before my assignment and training at USACIL, I had 12 years as a law enforcement officer, military police and CID, which should help define my ‘bias’ to investigation of subjects and subsequent application to laboratory examinations and opinions of origin (a term that can be applied to various comparative forensic examinations) over a span of almost 30 years with the US Army and State of Florida.

I cannot accurately recall details in my numerous examinations over many years, but there are a few that could relate to ‘bias” or the lack thereof. I can recall comparing a latent to ‘known’ suspect and having a latent print (one case even involved a single latent in blood residue) and feeling in my “heart-of-hearts” that the latent was made by the subject, but I did not make the call. I did not succumb to ‘bias’.

I worked on several very hideous and disturbing crimes during my tenure and there were times that I REALLY REALLY wanted to link the suspect, established by other circumstances, to the physical evidence from the scene, but was unable. I had to rely to my examination results and, if they failed, so be it. However, there were times when I wondered if I could have done more or ‘fudged’ on some latents. I never did. I did not succumb to ‘bias’.

There were also occasions when I had to address other issues. There has been discussion about small departments and their latent examiners. I was giving a class at a department when the officers received word from their department examiner about an identification that would warrant their execution of an arrest and there was a lot of excitement about an impending arrest in an important case. I was asked (for some reason) to look at the comparison. I told them that they should regroup because it was not an identification and there was no basis for an arrest on that evidence.

There was another incident in when a visiting “examiner” (from a small department) requested our laboratory prepare a court chart for him to present in a Federal trial (Federal trials in Jacksonville extend to many states west from Florida). While preparing the chart, there were discrepancies noted between the latent and record. When asked, the ‘examiner’ explained that the identification was ‘close’ and that was acceptable in his department. The Federal Prosecutor was notified about our concerns and the ‘examiner’ was sent packing back to his department without testimony.

I am not taking the high road on these issues. I am no different from hundreds of examiners in the field. All this talk about bias in examinations and opinions is damaging to the many, many conscientious examiners going about their duties in very ethical manners. I post this because I do not appreciate the continued attack on the influence supposedly applied by the law enforcement personnel to laboratory examiners to provide the ‘right’ opinion. I was both, law enforcement officer and forensic examiner, and I used the qualities of both to provide service to the community.

The issue is not about training or standards, but ethics. An examiner must accept that the discipline standard is "NO MIS-IDENTIFICATIONS". That will not be tolerated. You transgress and you are OUT. That is a necessary part of the training.

I am no longer a forensic practitioner and it may be a vastly under simplified solution, but if you do like an opinion of origin, find your own “expert” (because they are out there in one form or another) and go from there. Let the game begin!!

[Gerald Clough]

All this talk about bias in examinations and opinions is damaging to the many, many conscientious examiners going about their duties in very ethical manners.

I can't really argue against the truth of that. But I've come to see the "bias" topic as maybe the best example of a peculiar kind of conflict that comes into being. Our method is largely one of human perception, which we know is complex and requires considerable discipline to apply reliably. Part of that discipline, part of the expertise, is handling potential biases. That we think about them is good, since they're unavoidable potentials in any practical process. They're part of what we do just as distortion and optics and image reproduction are parts. As potential factors in what we do, we're really bound to want to know as much as we can about all of them.

On the other hand, we don't work in a vacuum. We're not doing examinations as an academic exercise. Since we operate to the end of fact-finding that's frequently resolved in actual litigation or pre-litigation negotiation, the questions that prompt study of our processes naturally are seized upon by litigators. Additionally, we draw the attention of the more academic sciences, because (1) there are areas where other sciences deal with similar questions, and (2) the nature of academic science is that it operates to different standards and to different ends, and they're allergic to what they see as misuse of the term "science" in "forensic science." So most any question that seems to relate to what we do is going to be, at some point, presented as critical of our products.

It is indeed a distasteful fact that much of what we ask ourselves in our efforts to become better and some of the issues honestly raised by others will be badly used, sometimes by those who are entirely adversarial and prepared to spin anything to their position and sometimes by those with genuine concerns. We're not alone in this. All of us who are or have been investigators have, more than once, been subtly or not so subtly accused of deliberate or careless bias. It would surprise me to hear from a active detective who hasn't had it proposed to them in court that they "just decided who you wanted to charge with this and ignored anything that directed the investigation elsewhere." Surely there were cases and times when some investigators were guilty of that. But most of us did it properly and could demonstrate that we had done it properly.

Talk of bias is not going away. Is it damaging? Of course, in a sense it is, because, again, we're not in a vacuum, and only rarely are our issues properly represented in the popular press or even in the science press. But, where it counts, I don't think it damages competent and ethical examiners, exactly because we do talk about it. Our reliability is based on an abundance of caution. Because we have no sharply defined thresholds, it must be that way. I think potential bias is like that. Yes, we know that human perception and interpretation is subject to various biases. And the ones we think most about are not gross attempts to influence. And those aren't the ones most outside the discipline, at least those influence how our products are used, are talking about. The point is that we do talk about them and think about them. And, being aware of the potential, we impose on ourselves a discipline of demanding a high standard of confidence. We can't say, "Oh, there's really nothing to worry about. It's not an issue." If it's an issue for someone who matters, it's our issue, too. And we know how to address it according to what's known today.

But we had better know all we can about the issue. Otherwise, we can't effectively put anything proposed as potentially corrupting our conclusions into the proper perspective. Or maybe, just maybe, we ignore something something that really deserves attention.

[atorres]

I would like to begin with saying that one of the only reasons that I am posting in regards to this topic is the comment that Dennis Degler said about:

“Bias occurs in the mind, then may or may not spill out into our actions.”

I found this comment really intriguing. To me it is common sense, but it was nice to hear/read it put in a concise manner. Whether we want to admit it, as humans we have been presented with biases from the time we started learning. We are presented with the same and new biases everyday (mostly). These biases form in the mind and there they stay. Like Dennis said, with most things that occur in the mind, we have the potential to act upon them. Also, what I believe Dennis was saying is that we cannot shut out all these biases, they will always be there.

I think that the most important thing when discussing bias is:

1. The recognition or awareness that bias exists. It exists in more than one form and we as humans can chose whether or not we act upon these biases.

Examiners are human and therefore are subject to the same bias that everyday people are. It’s the awareness of certain biases that make us experts at what we do. If we are aware of these biases, then we can put certain measures in place and/or do casework in a manner that will prevent bias (or prevent the least amount of bias) from affecting the outcome of the case. In the end it is up to the examiner to choose whether or not they will be influenced by bias. To say that it will never happen to them, well... all I can say is we are human!

I was doing some looking around and came across the “Bias Blind Spot” which basically is a person says that they know that bias exists (mostly in others), but they don’t understand how those biases influence their own judgements that they make. There was an article that I came across titled: “The Bias Blind Sport: Perceptions of Bias in Self Versus Others” by Emily Pronin, Daniel y. Lin and Lee Ross. I found this article very interesting.

The last thing that I wanted to say is, I am fairly new to the latent print field and am currently in training. I am realizing that bias (just the ones relevant to our field) comes in so many more forms than I thought. As a trainee, I feel that I have to be more aware of these biases, so as to not let them affect what I do in casework in the future. In my old job, I worked a lot closer to the detectives/officers than I do now and I have to say that the amount of bias has been cut down, but not completely eliminated. I can obviously keep talking about this subject (and other fingerprint subjects), but I am sure there are more interesting things you could be reading. So, to end, I must say that personally I believe that bias is out there and will always be a part of the job, but it is our awareness of it that will hopefully not let it influence our decisions at the end of the day!

Anne

[raybie]

If I allow someone who is not a trained expert in my field to influence my findings the results are not true. That is not bias it's perjury.

[Graham F]

Raybie

Well the issue here is that you have allowed this to occur. “I allow someone who is not a trained expert…. to influence my findings.”
The other person was not a trained expert, but you have allowed them to influence you. Shame on you. If it is your evidence, you created the perjury not the other party.

[Gerald Clough]

There are all sorts of people who are "not trained experts" who have various interests in cases. And it's so rare that someone is deliberately trying to instigate outright fraud, that it's something of wasted effort to talk about it much. Biases are subtle. The things we could reasonably be concerned about are anything that affects our idealized standards. We are sometimes working at the edge of our abilities. I think that's not unlikely to happen in a major deal when everyone is going all out. Chiefs and sheriffs obviously want results. They focus on their investigators who want to resolve the case. There are more remote people that the people we work for worry about. I think it's fair to consider those things in a case like the Madrid bombing. Hot-button terrorism case where the examiners were, I think we would all agree, working with a most difficult latent. When you're working on a case, especially a difficult one, all sorts of methods and techniques are on the table. There's a lot of intense thinking. A lot of concentration. A lot of decision making. There may not be overt time pressure, but in any case like that, the overall effect feels like the same kind of pressure.

You have to step back from the aftermath of this kind of error case and ask yourself how decisions were made and influenced. I think about that one a lot. Since I in no way even suspect that there was any kind of effort to just pin the tail on the convenient donkey, I have to ask how, with such a difficult latent, readily available techniques, techniques rather beyond what is commonly done in many cases, weren't invoked. I mean, the OIG managed it and produced a graphic demonstration. I further in no way believe the caliber of experts involved wouldn't have recognized and forthrightly reported a non-identification after looking at those results. There are varied and complex reasons, perhaps, why someone involved didn't up the level of examination, didn't raise their index of suspicion or at least their openness to suspicion, but I can imagine no other primary reason for the error but the array of pressure biases that affected the process. Even that is complex. High experience and status tends to make you more confident in your abilities. That's natural and expected. And if something is subtly affecting your thresholds and other decision making, altering the point at which you become confident in the identification, that confidence is all the more solid for your proven abilities.

In what we're talking about, no one "allows" influence. That's the point. If someone is aware of the influence, they belief they are proof against it. But knowing in a theoretical way that such influences can work doesn't automatically result in recognition of the effects at any specific moment in an examination. Any number of examples may be given showing that people persist in exhibiting the effects of biases, even when they are fully aware that such biases are known to effect people in general. There is some evidence that being subject to some biases, particularly biases producing factually unjustifiable cautious optimism and confidence, offer strategic advantages and that if an activity is observed, the number of people seen to be affected by them will grow over time toward some point of equilibrium at which those people are in the majority. (The highly over-confident are frequent losers and get weeded out.) Over-confidence bias is arguably significantly more powerful than confirmation bias. And, if being subject to bias offers advantages, those advantages are most readily realized only if we are biased toward not knowing we are biased, the "bias blind spot."

As humans, we are programmed to seek heuristic short cuts. And we are subprogrammed to seek shorter cuts under more intense pressure. That's a very natural thing. We wouldn't have lasted long by utilizing the most certain and therefore longer judgment processes. Nor would we be very effective if we always had to carefully calculate the degree to which we were being urged to get quickly to a workable result. That would take even longer than the method of highest certainty. The bias blind spot, too, is an advantage. We are exquisitely designed to be good guessers. A "good guess" doesn't mean the absolutely most certainly accurate guess. It means the guess that makes us most successful, often the guess that makes us winners in the long run, partly on account of how our guesses affect the competition's response. In a rough analogy, a good poke player doesn't win by betting accurately on the likelihood that his hand is better than the others at the table. He wins by guessing bets that will affect his opponents' play to his advantage.

Think about how we measure our reliability. Since we don't have a proven valid measure of absolute accuracy, we validate our practice as never having been shown to be inaccurate. That's how we describe how well we guess. We form our process of concluding according to how often we win, meaning here that we want to win every time. In our practice, we don't have even a long process that leads to absolute proven accuracy. We operate what is effectively a short cut alternative to whatever that imagined long, known-accuracy process might be. Seems to work. It stands up to our standard of producing results almost never shown to be inaccurate. (Can't say never when we are talking about practice in general, because we were just talking about an error result from operating the process under actual conditions. Saying, "We're always right when we do it right," is circular and meaningless.) But it's well to remember that it is indeed a kind of short cut and that we are already into that kind of decision making using a mind that is, happily for our survival, specifically evolved with bias inputs, and that part of the effective design is that we tend not to recognize the bias while it's operating. It's well to remember, too, that our practice has evolved with those biases in operation. The evidence tells me that the practice is sufficiently robust that it operates with the desired effect under conditions of "normal" biases. But, because bias is something else we can't measure, and apparently we can't be expected to recognize at the moment we might want to measure it, we can only crank up our suspicion of its influence as the task becomes more demanding.