UW School of Law > News > Multimedia Gallery > Eyewitness Identification

UW School of Law Transcript

Eyewitness Identification Science and Law Enforcement Practice

November 22, 2013

Paula Wissel

Good morning. I am Paula Wissel a law and justice reporter at KPLU radio. I want to welcome you to Eyewitness Identification Science and Law Enforcement Practice. I was very honored that Lara Zarowsky asked me to moderate this. I cannot imagine anything more important in the justice system than this. I was just thinking about the power of eyewitness testimony. Listening this morning on KPLU NPR's story, of course it is the 50th anniversary of the John F. Kennedy assassination and two reporters were just describing with that day was like, what they saw, what they did and just the overwhelming power of listening to that. And also I have been on a jury. It is just very powerful to hear someone say, “what happened, that's what I saw". So it is so important to make sure that that is accurate.
I was also honored to moderate a panel in Tacoma. And Alan Northrup was on the panel who has many of you know is one of the people at the Innocence Project, worked very hard to have exonerated and it was eyewitness testimony that put him away for over 17 years. And I so often, after that panel, that just struck me so, I think of him often and just think how terrible to be put away knowing that you didn't do it but someone identified you as having done it. So to me this is just a very, very important thing in the justice system and I know you'll agree.
We have wonderful panelists and we're going to ask that you hold your questions until the panel discussion, the Roundtable which will be at about 1145. With the exception of chief William Brooks who has to catch a plane so when he does his presentation you can ask questions then. So we will begin with the innocence perspective with Lara Zarowsky and Rebecca Brown. The role of eyewitness misidentification and wrongful convictions and methods used to implement eyewitness identification reform nationwide. Rebecca Brown engages in state-level policy work at the Innocence Project in New York where she has served since 2005 is the director of state policy reform. And Lara Zarowsky serves as policy director for the Innocence Project Northwest and is a law lecturer here at the University of Washington school of law so let's welcome them.

Lara Zarowsky

Thank you all for being here and thank you Paula for agreeing to moderate for us this is been great so far. And good morning to all of you and thank you for being here. My name is Lara Zarowsky I am the policy director with the Innocence Project Northwest which is based here at the University Of Washington School Of Law. And soon you will be hearing from a group of practitioners and experts from the criminal justice system from all across the country. And we will actually be showing you a video clip of a case of mistaken eyewitness identification that was the case out of North Carolina. But I would like to open this program today by emphasizing to you the importance and the relevance of the information that you're going to hear today with regard to our own justice system here in the state of Washington.
While states that have attempted to address the problem of eyewitness misidentification's have done so in different ways the fact remains that the principles of memory and recall and the reliability of procedures used to gather that memory evidence is not subject to state boundaries. Through the work of the IPNW and the requests for assistance that we receive we have an opportunity to see many examples of the way that eyewitness evidence is gathered in this state. Although it's unlikely that in every request for assistance we receive is the case of actual innocence there can be no denying at this point that some of them are. And I don't believe that anyone once the wrong person to be investigated and prosecuted and imprisoned for something that he or she didn't do while the actual perpetrator goes free. But the past year for decades of research have demonstrated that some practices traditionally employed for gathering such evidence can be problematic and may increase the likelihood that an innocent person is identified. So although total certainty, really no procedure that we would put in place could ever be 100% accurate or be completely reliable research suggests that there is a tremendous amount of improvement they can be made in the way that this evidence is gathered in order to improve its quality overall. So in other words we can do better and we should.
Before turning things over, I would just like to talk a little bit here in addition emphasizing that the risk of eyewitness misidentification is just as present here as it is in any other jurisdiction. It is also important to realize that this isn't really new information. We have learned much, no question, from the DNA exonerations at the other exonerations that have happened since 1989 when DNA was first used to exonerate an innocent person in an American courtroom; and also from the research on memory that has really been done in earnest since the 1970s. But the prevalence of eyewitness misidentifications in connection with wrongful convictions was first reported by Edwin Borchard in 1932 when he published this seminal work on wrongful convictions in the United States. And notably 37 of the 65 cases that he highlighted-- and from the foreword you learn that there are actually many more samples that he could have chosen from-- but of the 65 that he chose, 37 involved at least to some degree a mistaken eyewitness identification.
So before turning things over to Rebecca Brown who is here from the Innocence Project which is based in New York City I would like to take a moment to describe the work that we do here in Washington state. The Innocence Project Northwest in 1997 it was established as a volunteer attorney organization and operated as a volunteer attorney organization until 2002 when we were first offered as a clinical law offering to law students here at the University Of Washington School Of Law. The students, the volunteer organizations, all of the work that we work on is exclusively post-conviction claims of actual innocence in Washington state. There are DNA cases and non-DNA cases but only within the state of Washington. And that was our model with our director Jaclyn McMurtry really shouldering the burden of all of the cases with her roving band of students for each year’s clinic for many years until in 2008 we saw the addition of our first full-time staff attorney and in 2010 I returned to the Innocence Project Northwest to establish the policy department. So of course the policy work differs in substantial ways from the litigation work. The litigation work is cases of viable claims of actual innocence that are worked through the court system. But the policy work is really about identifying the causes of wrongful conviction that we have seen in Washington state and promoting changes in our system that will reduce the likelihood that innocent people will be convicted in the future, going forward.
Of course as you might imagine the policy work that I take on as the policy director of the IPNW is really driven by the cases. If you look at the exoneration cases that we have here in Washington state you will see that it's kind of a smaller scale version of what we see on the national picture. There is absolute consistency in the causes of wrongful conviction Washington state in relation to what we see from the national picture. We have cases that have involved incentivized informant testimony leading to wrongful conviction, false testimony, bad lawyering on both sides-- both defense and prosecution and of course mistaken eyewitness identification.
In the case that Paula just mentioned to us-- the case of Larry Davis and Alan Northrup is a very strong example of what can happen in these cases and what did happen in our state. The Larry Davis and Alan Northrup contacted us at the Innocence Project Northwest in 2002. They had both been convicted in 1993 based on a mistaken eyewitness identification of the victim. What is probably most notable about the identifications, at least for those of us who are not steeped in the actual literature, what stands out most right from the beginning is that the sheer number of identification procedures that were done in this case. There were three procedures starting with a composite sketch than a photo lay down and then a live lineup. And you might conclude after some of the testimony you here later on today about how that could be problematic in leading to the wrong person. So Larry Davis and Alan Northrup served over 17 years in prison before in 2010 DNA test evidence excluded both of them from all samples of evidence and actually identified the profiles of two unknown males and those profiles are currently residing CODIS.
With that I would like to turn things over to Rebecca Brown who will be speaking to you about the role of mistaken eyewitness identifications and exonerations and reform efforts on the national level.

Rebecca Brown

Good morning my name is Rebecca Brown and I am the director of state policy reform at the national Innocence Project. Just to kind of explain who we are vis-à-vis the Innocence Project Northwest we are a national group. We are based in New York City. In terms of policy work we work side-by-side with our state partners. So, for instance I have the great pleasure of working with Lara Zarowsky quite frequently on different innocence related policy reforms.
The Innocence Project, we are a national litigation organization. We also take on claims of wrongful conviction. Unlike the Innocence Project Northwest, we only take on cases where DNA could prove guilt or innocence whereas the Innocence Project Northwest also looks at claims of innocence involving other forms of evidence that could prove innocence. And then we also seek to reform the criminal justice system as described by Lara and our policy work. There are 65 projects around the world now most of them here in the states. Some states have multiple projects for and since Texas has four different Innocence Projects. And then other projects will cover multiple states like the Rocky Mountain Innocence Project covers Utah, Nevada, and Wyoming.
And just looking at the national picture-- and again I am just going to focus on a smaller universe of wrongful conviction cases mainly DNA cases-- but there are 311 DNA exonerations across the country now. You can see that the rate of revelation of wrongful convictions using DNA is really growing. That is because we have more lawyers dedicated to litigating these claims. And we are also seeing that a lot of these statutes around the country that permit access to postconviction DNA testing are loosening in terms of who they are permitting to seek testing. For instance, in the past many state statutes limited testing to people who did not plead guilty. Certain states are now widening the net and allowing people who played guilty for instance to seek testing and they are improving their laws so that all deserving applicants can get tested.
On average our clients spent 13 1/2 years behind bars for crimes they did not commit. Notable is the fact that 18 of them spent at least some of their sentence on death row. As you might expect, because these are cases involving biological evidence, the lion share of these cases involve rape, murders, or a combination of the two. What often gets lost in the mix particularly when the media is focusing on a wrongful conviction case, and of course the terrible harm done to the innocent person and his loved ones, is also the fact that a wrongful conviction really hurts public safety because every time the wrong person is identified and placed in person the real perpetrator is at large in a position to continue to commit additional crimes. And we have begun to collect just some data relating to this at the national project. We have identified 152, we have identified 132 real perpetrators connected to 152 cases of the 311 DNA exonerations and that is because some of our cases involve multiple exonerees where only one perpetrator was responsible. What we have found is that wall the innocent were behind bars in these 152 cases, the real perpetrators went on to commit 74 rapes, 33 murders, and 30 additional serious violent crimes. And I want to just stress that these are convictions only. This does not represent the scope of the criminal activity that these people were probably engaged in. So obviously getting the right person helps all of us.
We dissect all of our cases to look at the major contributing causes of wrongful conviction. I I noted before that DNA cases are just the tip of the wrongful conviction iceberg. Criminalists estimate that DNA is probative or tells us something about innocence and well over 10% of cases. That number will I think grow more and more particularly now that touch DNA and other forms of DNA testing methods and advancements are taking place. But we do know that DNA is really still a small universe of wrongful conviction cases. Also, we have a lot of difficulty finding old evidence. I think a lot of that is because first governments can be forgiven for not anticipating the probative value of DNA. They didn't know that DNA was going to be able to tell us so much. But we are also now still in a position where it is incredibly difficult for us in many of our cases to locate the biological evidence to test. And this is also something that cold case detectives are struggling with as well in their old cases. And actually, we think a lot about innocence claims is a nether form of cold case. This is another issue that kind of holds us up in terms of proving innocence on the DNA front.
And finally as I mentioned before post-conviction DNA testing is not always permitted. Some state statutes are incredibly limited. They might say that you have to apply for testing within a certain time after your conviction or they may say that if you confess to the crime you can't seek testing even though we now know that many of our cases involve false confessions.
Here are the major contributing causes of wrongful conviction: informant testimony or sometimes jailhouse snitches as folks have heard of, they are present in about 1/5 of our cases. False confessions as I mentioned more than a quarter of our cases involve them; this is of the 311 DNA exonerations. On validated or improper forensic science was a contributing factor in about half of our cases. And the single largest cause of wrongful conviction is eyewitness misidentification present in about three quarters of our cases. That number is 226 of the 311. What is interesting as well is that archival studies which really look at real cases, old real cases find that eyewitnesses just get it wrong about a third of the time. This makes sense right question mark memory fades. Memory is fallible. And we are all human. So a lot of this has very little to do necessarily with what anyone does right or wrong. It is everyone trying their best but still getting wrong.
Brandon Garrett is a professor out of the University of Virginia and he took a very robust look at the nation's first 230 DNA exonerations at the time that he did this inquiry that was about how many DNA exonerations there were. What I thought was fascinating about some of his research was that in nearly 40% of the cases he looked at the innocent person was misidentified by more than two people, two or more people. So this wasn't just one person getting it wrong in a case it was several. The good news is that over the past 30+ years there has been an emerging body of scientific research that tells us how we can make simple improvements at least two police protocols to try to improve the accuracy of eyewitness evidence. You're going here a lot more about that I know from Dr. Ross and from chief Brooks who will describe really the details of what those reforms look like. The single most important reform from our perspective is the blind administration of lineups meaning that the person administering the lineup does not know who the suspect is. And we are seeing growing support for this reform around the country. The entire states of Connecticut, New Jersey, North Carolina, Rhode Island, Texas, and Ohio are now using a blind administrator in their lineups. Many large cities around the country from Baltimore to Denver are also implementing the blind procedure. And the number of states have also made changes to release their training. We think that is a great first step. What we have also found a disconnect at times between what is trained in and what the local agency actually ends up adopting as their policy. So I raise this only to say we really support and we think it's wonderful when training programs embrace scientifically supported best practices but we also think it's critically important that the local agency adopted for use in their own departments.
And I'm just going to touch on some fairly recent developments around the country, some court action in New Jersey and Oregon. And some model policies and directives that have been issued by major policing organizations. So the first major sort of the valuation of eyewitness evidence that took place recently was in the state of New Jersey. And they were looking at the issue of whether the test that was used by the courts to assess the accuracy or reliability of eyewitness evidence was out of date. This is because every state has adopted what the Supreme Court adopted about 30 years ago in the Manson decision. It was a balancing test to kind of assess how accurate the identification was. For instance one of the factors they looked at was how confident the eyewitness was in the identification that he or she made. And we now know that science tells us that there is little correlation between accuracy and confidence. So the court knowing this in New Jersey said let's re-examine this test and see if it still makes sense the present day. They appoint a special master. They listened to many, many witnesses. There were 360 exhibits that were submitted to the court. Most of those were scientific research, psychology experiments and the like so they were really doing this robust analysis of all the existing science out there to determine whether that balancing test still made sense. And they determined that the test was confounded. Memory does not work like a videotape. Contamination is quite easy because memory is so fallible. And that lineups or other identification procedures are really just more like memory experiments. So what we really need to do is figure out how we can make sure those memory experiments are as clean as possible. They issued a very long opinion and I think the key point here which is really interesting is that all the scientific data and research on this issue has really emerged since the initial balancing test was determined, the Manson test. And what they did was endorse a new legal remedy. They did not, for instance look at the issue of whether or not to suppress the identification if it didn't follow best practices but instead they said that there should be very carefully tailored jury instructions that addressed any deviation from best practices so that the factfinders have a strong understanding of what it meant when for instance a blind administrator was not used, or for instance, law enforcement didn't take a statement from an eyewitness about how confident they were in their identification. So these are very specifically tailored instructions and the Supreme Court rules committee then came up with a set of jury instructions for use in New Jersey. And if you want to read the science legal findings I have included links in your literature and it's really interesting. It cites to a lot of the science without there.
Just a few years later the Oregon court took up the same issue--the High Court in Oregon. They even went further than New Jersey. They actually took that traditional balancing test and they flipped it. They said that the burden now should be on the state to prove that the eyewitness evidence, they are in a position to demonstrate that it is accurate and reliable before it is to be submitted to the court. And so they really shifted that burden which was a huge, huge change in Oregon. I note these two cases only to say these are two state high courts really now beginning to weigh in on all of this emerging research and taking notice of it. So we can see that this helps factfinders. It helps public safety. All of these changes are important all of us but it also aids law enforcement. While the benefits are undeniable to the innocent, we should think to about the fact that it helps law enforcement to not be misled early in their criminal investigations by focusing on the wrong person which simple changes to police protocol can enable.
The other thing that we have heard a lot from law enforcement that have implemented these changes is that they're no longer subjected to defense challenges in court because they are just doing it perfectly. Their identification procedures are unassailable in the courtroom. Some national policing organizations have also weighed in. The commission on the accreditation of law enforcement agencies adopted directives back in 2008 which endorsed some of the reforms that we seek. Just a couple of years later the international Association of Chiefs of police issued what I think is a phenomenal model policy on how to do good eyewitness identification protocols at the policing level. It is just excellent. I encourage folks to have a look at that.
So what are we seeking here: the same thing that we seek in any state which is uniform implementation of best practices at the agency level. We of course want training to be a big part of this because law enforcement you know often in the face of change feels that their integrity is being attacked by these reforms. It is far from the case. This has nothing to do with law enforcement integrity. It has just everything to do with best capturing the memory of the eyewitness and so that training is critical in terms of enabling law enforcement to understand the need for the change. Reform does not have to take place to legislation. We have worked on this issue in legislatures we have also worked collaboratively with law enforcement through attorneys general, through divisions of criminal Justice services, departments of public safety. There are all different ways of affecting change but what is critical is that that change be done uniformly across the state so that there aren't different forms of justice depending on which county or district you're in. It's often very helpful to have a statewide model policy that all of agencies can look to. Really a roadmap for implementation to assure that it is Matt and that compliance is taking place on the local level.
I am happy to answer any questions later in the program and this is my contact information. You can also feel free to contact me and thank you so much for your time.
Let me just, I can just actually, we have a great case study that we wanted to share with you. It was done on 60 min. This was a few years back. This was the North Carolina case that were just describe to you before. I think what's really not worth the about this case example is the impact that had on all members of the system and all the people who were involved in this case from the crime victim to law enforcement.
25:21

Paula Wissel

So next we have Dr. Stephen Ross talking about the science behind eyewitness memory and behind reforms commonly implemented and to improve the quality of that and Dr. Ross is an assistant professor of psychology within the interdisciplinary arts and sciences program at the University of Washington in Tacoma.

Dr. Stephen J. Ross

It is good to see so many people out here on a Friday morning. I just wanted to spend some time with you today talking about the science behind these reforms. Rebecca and Lara mentioned that there are some common reforms that are put out there and these reforms are quite comprehensive that are there but across all these reforms there is about four reforms that are considered core reforms. They seem to be a part of every single recommendation package that is put forth. So I want to spend some time today to address what is the scientific bases for these reforms? We are not going to touch on all reforms that are out there in all of the scientific research that can be used to help us understand eyewitness memory and identification. We are going to talk a little bit about these four reforms.
So the purpose of my talk today really is to first give you an overview of basic memory processes. I am not going to kill you too much with theory and data on how these decisions about memory are made but it is important to have a general understanding of what contributes to and identification decision in order to really appreciate how these procedures that are used in collecting this eyewitness evidence can really influence the decisions that are made by the witnesses. So we are going to spend a brief amount of time going over that. And then after we get that basic understanding there then we will spend some time talking about what is the scientific research that has been conducted on like I mentioned these most commonly recommended reforms.
So the four reforms that we are going to spend some time talking about here are like I mentioned commonly included in most reform packages. These are related to ensuring that the lineups are adequately constructed. By lineups we are referring also to photo spreads, both live presentation of a group of individuals but also photo spreads which are the most common way in which this evidence is collected nowadays throughout the US. We are also going to discuss that appropriate instructions should be provided to witnesses before the engage in any identification procedure. We will look at some of the research on that. The third reform which Rebecca mentioned is one that the Innocence Project feels is probably the most important reform from their perspective is that the individual administering the identification procedure should not know who the suspect is in that lineup and that is something that we call blind administration and I will present some of the research on that. And then lastly that witnesses should be asked to provide a confidence statement immediately after their identification decision and I will talk about some of that research.
Time permitting, I also want to spend a little bit of time talking about sequential presentation. And this is a pure is a maybe. Sequential presentation is an area of hot debate with the scientific community and in the criminal justice community as well. If we have time I would like to discuss with you what the current research is on sequential presentation in comparison to simultaneous presentation and more specifically or more importantly from your perspective is to understand what is the debate in the scientific community. Why is there not consensus in the scientific community about and acknowledged benefit in engaging in sequential presentation?
As both Rebecca and Lara mentioned there are wide recommendations out there as far as from different organizations from Korea from my ACP, the national Institute of Justice put out their recommendations back in the 90s. All of these recommendations are based in part, not entirely, but in part on research from the scientific community. So we're going to spend some time talking about that today but I wanted to just give you exposure to this text for those of you that want more information on the science behind these reforms. This is the text that was just published this year. Brian Cutler is the editor of this body. In this whole body although it's entitled Reform of Identification Procedures, it's not addressing policy issues. It’s not addressing practice issues. It's based solely on summarizing what is the research on these recommended procedures? It is an excellent read. It is accessible, which is nice. It's not too technical. So if you're looking for a desk reference that you would like to have I would recommend at going and taking a look at this text here.
But first let's start with the first purpose of my talk today, it's to give you a basic understanding of what factors contribute cognitively speaking to and identification decision. Really when an individual is making an identification, this is an eyewitness identification. This is any type of identification task when you recognize something. There's two components that will influence that decision: one is going to be the strength of the memory and the other is going to be a decision threshold. And really what I mean by this is that there's two steps to this process. The first step is that you have to determine whether the stimulus that you have whether it's an individual and a lineup or a photograph of somebody how strong of a match is there between that imagines a memory that you have? That is the first step. You ascertain, you come up with some estimate of the strength of that match. Then, you compare that strength and determine whether is that strength strong enough? Is that match strong enough for me to be willing to say yes I recognize this object of this person? So there's two components to an identification. The relevance of this is that some things may affect the person's first step, that is assessing the strength of that match to memory. And factors that will affect that will actually result in being better memory tests. They will actually make it so that you are a better, you do a better job of recognizing. Factors that affect the second component, that is they alter how willing you are to choose may not make your decision any more likely to be accurate it just may make your decision more likely to be yes. And this is important because are going to look later on at some procedures and were going to look at these procedures that identify is this affecting the probative value of the evidence that we are collecting or is this procedure affecting just the willingness of the individual to make a decision?
Now the other relevance to this is that when we have a good memory for somebody it really doesn't matter much what procedures are used to test that memory. So if I ask you if this is your mother or if this is your uncle Richie or if this is the president it doesn't really matter much how I asked that question of you. You are going to be a will to make that decision accurately that know that's not your mother that's not your uncle Richie but yes that's the president regardless of if it's a biased lineup or if it's a more fair lineup. You are going to be able to make a decision regardless if I tell you hey he is their pick them out versus hey he might not be there. When your memory is strong you're going to be able to make these decisions fairly well in the influence of the procedures that are used to collect that information isn't going to be as great. However, when your memory is weaker all of a sudden now we have decreased the likelihood of having a strong match to memory. We decrease the likelihood of having a strong memory response in that first component of the decision. Now all of a sudden factors that influence our willingness to choose are now going to likely be represented in our decision choices. So for example if you look at this you might not have any idea of who the suspect is because the lineup is fair. However if I give you a bad lineup now all of a sudden it becomes obvious that might affect your willingness choose. So that is the relevance of what we are talking about here. We want to be sensitive to this is you look at researching you read all this research and you start to understand or you start try to identify why are we getting some variability in results? Well this variability might be related to the strength of the memory. Some factors might show a result only when the memory strength is weaker.
So, researchers have specifically been interested in identifying how these identification procedures influence these two components, that is what if any of the aspects of the identification procedure influence that first step of determining the strength of the match because those will actually result, we identify procedure a procedure that can make us better at doing that first step, we can actually get better evidence. It can increase the probative value of the identification evidence. But then the other side of the coin though is what if any procedures influence that second component of altering the witness’s willingness to make an identification? Factors that influence that may not increase the probative value of the evidence but just might increase the likelihood of getting evidence at all. Now the key thing with that to it and as we look at this is a fact that the influences latter point when Isaiah might not increase the probative value, it also might not decrease the probative value. It might have no effect whatsoever on how likely the ID is to be accurate. It just might make it more likely for you to get an ID. And we will look at that in a minute.
So I want to go through each of these four components with you and look at some of the basic research behind these reforms. And the first we will look at here is the recommendation that we should construct fair lineup is. Now in order to understand why we are concerned with constructing for lineups it is important just to look at what is the reason for having a lineup to begin with. The basic premise of a lineup is that that lineup provides some sort of protection to the suspect. We acknowledge that sometimes the suspect is innocent and so we want to have a procedure in which we have some protection to an innocent suspect from a witness that just feels compelled to identify somebody. So the whole purpose of the lineup is to provide protection. And in order to do that we must ensure that the lineup is constructed in such a way that the suspects doesn't unduly stand out.

Audience (36:07)

Dr. Ross I'm sure you'll go back to this but is there some way forensically for us to assess the strength of memory?

Dr. Ross

The strength of memory?

Audience

The strength of that memory in the choosing criteria where it is critical what sort of things do we look for and assess the strength of that memory in the first place?

Dr. Ross

As far as in individual decision?

Audience

Yeah and a particular case of the particular witness or in the forensics sense

Dr. Ross

You are going to be looking for factors that might contribute, other factors might be relevant or contribute to the strength of memory so objective estimates of how long the person had to view, the familiarity with the individual, those other factors that have been in the literature coined estimator variables those are factors that you might want to look at to get some sort of estimate of their strength of how good of a view they had of the individual. You might not want to rely on witness reports depending upon the procedures that we use as all get to a little bit later on of the factors that could contribute to the reports but projected estimates are things that you could look at to try to get an estimate of the strength to memory there.
37:12
Now when we talk about a suspect not standing out, really what we want to focus on here really what we are relying on is that there has to be some level of similarity between the fillers in the lineup-- those known innocents that we put in there to offer the protection and the suspect. In practice there is basically two approaches that are put forth to construct lineups. I'll go through these kind of quickly. The first one is the intuitive approach. It's the approach that is most commonly used in surveys of law enforcement policies and law enforcement practice we find that it's used in 80 to 85% in all of lineup construction techniques. They basically have the individual in custody or you have a photograph of the individual. You take a look at the individual and you pick out people that you think in some way look similar to that individual. And this is the match to suspect approach. You select fillers based on their physical similarity. The other approach that's out there is something called matched a description. In this we are not necessarily interested especially at the early stage of what this suspect looks like rather you take the description of the perpetrator evidence provided by the witness. You take that description and say to describe the perpetrator as 25 to 30-year-old white male, short brown hair, glasses and scruffy facial hair. Well now I go and I find people that have those characteristics: 25 to 30 years of age, white male, short brown hair, glasses and scruffy facial hair and I just use them. So I'm not really basing the people's inclusion in the lineup upon any direct assessment of similarity to the suspect. I'm doing it based upon the description. And research has examined which of these tends to produce more fair lineups and also which of these are going to be related to the quality of the evidence that is collected if you use that lineup in an identification procedure.
In initial research on the match to suspect approach actually found that although we think that that's an effective way to construct the most fair lineup but we actually find is that by selecting people that is somewhere physically similar to the suspect, we can actually counterintuitively create a lineup in which that suspect stands out. I will go into all the cognitive explanations as to why that can occur but basically the premise is now that suspect is the only person the lineup was guaranteed to have something in common with every single other person in that lineup and because of that now they might stand out. So first research suggested that because they stand out you may actually get an increase in suspect identifications regardless of whether they are actually guilty or not. So then researchers started to look at this match to description approach and they found that initially this is a much better approach to use. And the reality is this is a much better approach to use but there is one majorly significant caveat here. That is a relies completely upon the quality of the description that you have from the witness. And witness descriptions are notoriously vague. When I worked with the police at El Paso Texas which if you don't know it's right on the border with Juarez Mexico and we used to joke that if we could find not one 18 to 25-year-old Hispanic male with a shaved head and goatee would stop every robbery in El Paso because that is the quality that you would get. And at that point in time the descriptions are too vague to produce a good lineup. So the recommendations that's out there right now is use matched a description if you have a good enough description and if the description is too vague you're going to have to rely on match to suspect.
Both of these approaches though again the basic premise is that by using one of these two approaches we are going to construct a lineup that has the proper amount of similarity between the filler and the suspect. And so there has been research that has looked at how does suspect filler similarity influence witness decisions? And recently there has been a couple of meta-analyses that have looked at this. And a meta-analysis for those that are familiar in this basically it's a way in which we can statistically summarize the results from a lot of studies that address the same issue. So think of it in its most basic sense as a way for us to average the results across a lot of studies and see really what seems to be the effect that we are observing here? And what these meta-analyses have indicated is that high similarity lineups produce fewer correct IDs and fewer false IDs compared with low similarity lineups. So what these are resulting in it is that high similarity lineups are getting people to choose less but interestingly, the value of the IDs that we're getting our increased. There is more probative value to IDs that you get from a high similarity lineup even though you're getting fewer IDs, those IDs have more value. You can trust the validity of those IDs more. So that is why it is still recommended that you use high similarity lineups to a certain extent. Now you don't want to construct a lineup in which there is identical twins in a lineup but in practice that is extremely unlikely to begin with. So high similarity lineups are what the research is indicating. While they produce fewer identifications the probative value of those identifications is greater.
The next recommendation that is commonly out there is related to instructions that are provided to the witness before the identification procedure. Quote we tend to know here is that witnesses approach this identification procedure just like Jennifer Thompson. As you heard her now report what she went into that procedure with, "I'm getting get the guy. I'm going to be able to identify him." And witnesses tend to approach the lineup task with the assumption that there is a suspect somewhere in this lineup. Why else would they be showing me a lineup? Why would they be wasting their time and my time if they didn't have a suspect. So they approach it with the assumption that the suspect is in there and because it's not blatantly obvious to them they often aren't aware that pays saying not there is an okay response. I don't have to identify. Saying not there or I'm not sure unless they're told that they don't tend to keep that in mind is that really it. Based upon that the commonly recommended instructions suggest to use what we call the may not instruction: keep in mind the perpetrator may or may not be in the lineup. Which if you recall back to the short clip that we just saw that was an instruction that was provided Jennifer Thompson. It is also recommended that you need to inform the witness that is just as important to eliminate innocent suspects from suspicion is that is to identify guilty parties. It also is to ensure that they realize that the investigation is going to continue regardless of whatever decision they make. These are the commonly recommended instructions. Some of these instructions are included because of the research that has been done in the scientific community. Again I'm going to report here a lot of meta-analyses because they do a good job of summarizing the body of work that is done. And basically what we find here is that unbiased instructions decrease the number of identifications that we get. We give people unbiased instructions they are producing both fewer correct and fewer false identifications. So instructions are manipulating the second component of the identification decision. They are not affecting the person's ability to make a match to memory. They are affecting the person's willingness to choose. These meta-analyses indicated that there is no difference in the quality of evidence is collected; we are just getting less of that evidence. And by definition when we get fewer ideas we are going to be getting fewer false identifications also some fewer corrective certifications. Studies specifically looking at only the may or may not be instruction so unbiased instructions could be a lot of things. Some studies have looked at biased instructions like hey pick up a guy. And that has been included in the first set of studies. The second studies have looked just only at using the may or may not be instruction or not using it. And what they found there is that they produce fewer identifications again. But the probative value the IDs might be greater. So you get fewer IDs but it might be better evidence that you are collecting by using this instruction.
Another important thing to consider though this is what recent researchers have started to look at is well what is said to the witness before they even get these set of instructions question mark and this came actually anecdotally from some researchers working on the case in which the investigator called the witness on the phone and said hey we've got a suspect would do you want to come down and view a lineup? And then they get into the room they say oh yeah by the way the person may or may not be in the lineup. The question is well is that instruction even effective in output of time because they came into the station with the idea that they had a suspect in place? It really might be intuitive here leading the witness to believe that you have a suspect before you give the instruction means the instruction is completely ineffective. The results of this study were that it was identical to if you had never even provided the instruction to begin with. So does important to take into account, the research is saying it's important to take into account not just instruction but also when the instruction is given what is set before the instruction is provided.
So those of the first two core procedures: constructing lineups and what the research has to say about that and then also the instructions that are provided. I want to move to the third procedure now and this is blind administration. What we mean by blind administration is basically just does the administrator have knowledge of who the suspect is are not? In a blind lineup the administrator does not know who the suspect is. And the basis for this research-- it's interesting about this recommendation, this recommendation was one of the first proposed recommendations for reforming the system back in the 80s and it was part of national suit of justice working group recommendations the command the late 90s but at the time there was absolutely no research looking at blind administration and eyewitness context. The impetus for this conclusion comes from basic psychological research. And really it comes from all the way back to a horse. Maybe you are asking what does a horse have to do with ID procedures? The horses Clever Hans those of you who took an intro to psych undergrad course may remember the story of Clever Hans. Clever Hans was a horse at the turn of the 20th century and he was clever. He was incredible. Not only could he understand what people asked him what he could do math. If you ask Hans what's 3+2? He would stomp is hoof five times. If you said what's 9-7 he would stomp his hoof to times. And his trainer used to take him around he was almost like a sideshow he mesmerized everybody. And there was a psychologist at the time that sat there and said what's going on here question mark is a horse really this clever? And he found that Hans was clever but Hans wasn't as clever as people were giving him credit for. What Hans was doing is Hans was able to pick up on these imperceptible cues from the person who is asking the question and knew the answer. Because what would happen is as Hans started stomping his hoof as he started to approach the right number of stomps people would tense up a little bit. They would look to see is he going to get it right or not question mark and he would pick up on this and he would stop because he learned that if I stop and people do these behaviors people will applaud they say yay and I get care. And Pfungst was the researcher that demonstrated this showed that if he had people ask Hans a question that they themselves did not know the answer to Hans wouldn't get it right. If they had Hans, ask cons a question which you couldn't see people he wouldn't get it right. And this was the first study to demonstrate that the expectations of the person asking the questions although unintentional the suggestion wasn't that this guy was trying to pull the wool over the eyes of the people what they were finding was that these people didn't realize they were giving off these cues. But just that knowledge can be passed along and people can pick up on these cues. Obviously there is a lot more basis to this than just an anecdotal study of course. Blind administration has become standard practice in all scientific research. It is become standard practice in all scientific research that if I am doing a study, I set it up so that I don't know what condition is being administered to a participant so that I can't inadvertently queue them to what I think should be the correct response. Research has consistently shown the possibility for these effects and other real world consequences-- I mentioned lab research, elementary education classes. Again, those who took intro to psych might remember the Pygmalion in the classroom studies in which if the teacher was told randomly that some random student was on the verge of an intellectual growth that student actually developed faster than the other students in the classroom. You actually see it in the 80s and 90s they did some work looking at it in courtrooms finding the judges expectations about the guilt of the defendant can actually sometimes influence individual jurors values or expectations of the guilt of the defendant. Just through the way that the judges providing the judicial instructions before the jury deliberates. And so we see these effects are present in other areas and that was the basis for inclusion in these initial reform recommendations. But since that time we have said well we can't just base it in other contexts. Maybe the context doesn't apply. So researchers have begun and I say begun, in the last 15 years or so, started to look at blind administration in ID procedures and what we tend to find here is that when the administrator knows who the suspect is witnesses are more likely to identify the individual, regardless of the individual’s guilt or innocence. This is not suggesting that these administrators are intentionally biasing. That is one of the criticisms that come with this reform is that there is some perception that the investigators are doing this intentionally. That is not what this is suggesting here. It is suggesting that there is just some way just like with Hans it wasn't intentional, but some cues are being passed.
A closer look at the studies shows that there are some mixed results though. Some studies find evidence, other studies don't. So what seems to be-- again this is recent research that is coming out-- it seems that witnesses may be more likely to be influenced by this non-blind administrator in situations in which their threshold is already lowered to begin with. So we talked about bias lineups. They produce a lower threshold now there is more opportunity for the witness to be influenced by a non-blind administrator. We will talk about simultaneous lineups little bit later on and in that that's where you tend to see that evidence that in these situations this is when they are likely to be influenced.
Now an important thing to consider here is if the influence that is going on is it perceptible? Can we look at it after the fact and say okay there is something going on here where although it is not intentional we can see that influence is present. And the research that looked at this has identified that witnesses who were influenced didn't report observing anything from the administrator different than those that weren't. And if you have a neutral third-party observer president likes a defense counsel at the ID procedure who is tasked with keeping and I out for any bias they didn't observe it. And that gets to this issue of it's not necessarily intentional. It's these imperceptible things. Another study that was just published your recently demonstrates that again this isn't putting the onus on the administrator. It's not the administrator’s fault that this is occurring. Rather it is something about the witness. Just like it wasn't the handlers of Clever Hans that were doing it, It was in what Clever Hans picked up himself. And this research is demonstrating that when his expectations are very important. In this one study that has been done on this they found that all they did was let witnesses believe that the administrator knew who the suspect was and those witnesses that believed that were more likely to be influenced by the administrator and those that didn't believe that. And anecdotally Lara and I have talked about this before. She has an anecdotal explanation of this or example of this in which one of her friends was a victim or a witness I don't remember if it was a witness or victim but was asked to come in and do an ID procedure. And she's sitting there looking at the six-pack and she observes that the detective they're working with her just sitting on the other side. He's not involved in it at all he is just sitting there observing, is tapping the table. And he has three fingers out tapping the table. And so she reported afterword that I kept thinking is he telling me to look at number three? Is he telling me to look at number three? And she didn't identify anybody. She said that it's none of them in their but speaking with the detective afterword asked him about it and he said what are you talking about it wasn't doing anything. And apparently that is just something that he does when he's killing time. The witness expected and perceived influence that wasn't even there. It wasn't intentional. It was just something that the witness perceived. So that is the status of the research behind that procedure there. So so far we have talked about constructing lineups, instructions, blind administration.
The last of the core reforms that I'm going to talk about with you is looking at the collection of immediate confidence statements after the person makes an identification decision. Now the recommendation here is that it should be collected after any decision whether it's a not there or a positive identification you should collect the confidence of that witness at that point of time. And the research behind this is based into areas: one as Rebecca mentioned the Manson criteria for assessing the validity of eyewitness evidence at the time of the pretrial hearing is to-- one of the five criteria is the confidence of the witness. You should rely on the confidence of the witness when assessing the validity of the ID however plenty of research shows that both memory and confidence can be changed. It can be changed rather easily for a variety of reasons. But one of the reasons they can be changed is by something called posted notification feedback. Posted identification feedback can be something as explicit as good she got the guy after the person makes the identification or something as subtle as you've been a great witness. Any type of that feedback has been shown that it can actually influence three of those five Manson criteria. They can influence witness confidence and make them more confident in their decision it can also influence their subjective assessments of how good of a view they had, how much attention they paid and how in the degree of view they had which are two of the other five criteria in Manson. So the basis for this recommendation is simply that subtle feedback can get people to alter their retrospective reports and the confidence they have in their decision if that feedback is provided before they provide a confidence statement now we have no uncontaminated record of what their confidence really was and they made that decision. Interestingly we find that this effect is even greater when the witness is inaccurate. And this is one of those important distinctions here. A correct witness that gets confirming feedback doesn't have a huge effect on their memory or on their confidence. However, an inaccurate witness who gets confirming feedback that has a greater effect. In the explanation for this and I put it up your for your reference but I'm not expecting you to remember selective queue integration framework, but basically what this suggests is that when we assess our confidence there is three stages to this. First we do an assessment stage where we look for internal cues, that is the strength of our memory. If our memory isn't that strong or if we have questions about the strength of our memory than we my search for external cues. And then once we identify those external cues now I'm going to assess the credibility of those cues. How much can I trust these cues? And really what this means for an inaccurate witness who has a weak memory they are going to be more likely to use the credible external cues, that is the administrators feedback as a way to assess their confidence.
So to summarize here the research on these four core procedures is found that three of these procedures influence that second component of the decision task. That is, they will have an effect on the decision threshold or their witness’s willingness to shoot each. That is how you construct the lineup, the instructions you provide in whether the administrator knows who the suspect is. Two, maybe two of these procedures and is it to maybe because instruction it all depends on the instruction that's provided will actually influence the quality of evidence that you collect. Highly similar lineups give you better evidence. You can trust it. If you get an ID from high similarity lineup it is more likely to be inaccurate identification than an ID from a low similarity lineup. The same thing as the Steadly meta-analysis suggests if the may or may not be instructions provided that that can be up slightly higher probative value the ID received from that time of identification procedure that in one in which that instruction is not provided. Also with immediate confidence statements in the key thing with this is it doesn't affect the witness’s decision, the confidence statement aspect because the decision has already been made but the importance of the immediate confidence statement is because many jurisdictions, many states so rely on the Manson criteria. If we don't get an immediate confidence statement post identification feedback can actually influence three of those five criteria making it an invalid test during
Lastly, I want to spend a little bit of time talking about the Sim-Seq issue. This is not included as one of the core procedures primarily because there is a lot of debate over this procedure--over whether we should use sequential, whether we should use simultaneous and as such some jurisdictions have implemented it. Chief Brooks will talk about it. They have implemented it in Norwood. Other jurisdictions are considering changes but don't necessarily want to include sequential in their package. And to give an idea here of what we mean by sequential: there is a traditional sequential lineup that was developed in the laboratory. This sequential lineup involves presenting each individual in the lineup wanted the time. The witness has no idea how many photographs here she will see. The witness is asked to make a decision about each individual independently and they must make that decision before they move onto the next one and importantly for the laboratory-based studies the lineup terminates as soon as they make a positive identification. So if they identify the second person in the sequence they don't see any more pass a second person. The reason I say that this is a tradition of the laboratory is that in practice the sequential presentation has been modified. Most jurisdictions that are implementing sequential presentation will allow witnesses to see the remainder of the lineup after they have made a positive identification. They may allow multiple laps through the lineup. It will be documented of course whereas the traditional you just go through once if you don't identify anybody it's done. Where is most jurisdictions have now allowed for multiple laps and that distinction is important because that is part of the controversy which I'll talk about here in a minute. This obviously contrasts with the SIM presentation, the simultaneous, in which you see everybody at one time. You know how many are going to see and just like with the Jennifer Thompson and the photo spread that she saw you know I'll be seeing six these are the six and I'm just asked are any of these the person that could be the defense?
The current debate is about whether sequential presentation produces superiority or simply a shift. And to put it in the context of what I presented earlier the question is does sequential presentation of fact the first stage of the recognition process or the second stage? Does it affect your ability to search your memory and make a match or does is somehow just affect your decision or your willingness to make a decision? The peer Yorty would suggest that the individual witness has the ability, a better ability to correctly identify the guilty party when their president and all slow correctly reject a lineup that contains an innocent individual. That is the claim of superiority. The claim of shift would simply be that sequential just increases a person's unwillingness to identify anybody. It makes them more conservative. That is the debate that is present in the literature. Early research suggested that it was superiority. It suggested is so much that it was dubbed the sequential superiority effect so much so that it was led to the perspective of being reported in the media, being reported in experimental studies that sequential presentation has little or no cost and we are not talking financial cost here we are talking cost to the memory test, little or no cost to the accuracy. However recent research suggests that that superiority might not be robust. It might only be present in a very confined set of parameters primarily when the lineup is biased and when the suspect actually is placed lay in the sequence. If the suspect is placed early in the sequence you don't demonstrate the effect however if the suspect is placed lay in the sequence then all of a sudden the effect emergence. The other perspective is that it is not affecting the person's ability to discriminate or to test their memory rather it is just affect their willingness to choose. In recent research has found this. It is found evidence that the sequential presentation is just producing a conservative shift in their criteria. That is, when you give sequential presentation you're going to get fewer false IDs but you're also going to get fewer correct IDs. So some have expressed this as, and this is where some of the debate in the literature is that the initial presentation of this aggression is a presentation the initial description of this having little to no cost may not be accurate, that there are costs to it but the key thing is just because there is cost doesn't mean it shouldn't be implemented. That is policy decision. Just like when we showed with instructions, instructions have costs. They got people conservative instructions made people unwilling to choose they are going to be less likely to correctly ID. They're going to be less likely to falsely ID. It doesn't mean you shouldn't use it. That is a policy decision though. But that is where the debate has been in the literature of it well are we misrepresenting what the sequential presentation is? So the current issues with this is that does sequential actually have a benefit probative value of the ID and that isn't clear yet in the literature; and because it's not clear in the literature that is where you make your debate over this. Also for what conditions influence the benefit? As I mentioned some research is suggesting that is only beneficial in some situations. There is no difference in other situations between simultaneous and sequential. In that same research has also demonstrated that in some situations simultaneous is better.
Lastly, and this is where I mention what the traditional sequential is earlier, the question arises of well, what has been evaluated in all the laboratory research is different than what is in practice? We don't know how those practical changes affect the witness’s decisions and as such we can't speak to that. Like I said that doesn't mean policy should be enacted but it is a science can't really speak to what those changes mean that to the witness’s decision. We can't inform that decision at that time because the research hasn't been done to say how does this change in practice actually affect the witness’s decisions? And so that is why you will hear that there is uncertainty or debate over sequential presentation because the knowledge base just isn't there for how it is actually being implemented at this point in time. Apparently that is the last that I have there.

Part 2

Paula Wissel

Okay were going to start the second part of our eyewitness identification science and law enforcement practice. We have with us Chief William G Brooks III who is the Chief of the Norwood Massachusetts Police Department. In his 36 years he has served as a patrolman, Sgt., Detective Sergeant and Deputy Chief. He is going to talk about how procedures to improve the quality of eyewitness evidence have been implemented and have impacted a small law enforcement agency. And one more thing because he does have to catch a plane we will be opening up to questions after his presentation.

Chief William G Brooks III

Don't clap. You don't know if you are going to like it yet. So let me just tell you a little bit about where I come from I guess in that I spent a lot of time as a detective, Detective supervisor in my career, 14 years as a detective sergeant. And then when I was a deputy chief at about 2005 the eyewitness identification protocols from the in IJ report were just kind of circulating in Massachusetts and I try to keep up with this stuff so I can't really explain why they were out there for five years and I hadn't really heard about it or seen it. And our DA announced that they were going to do some training on these new protocols so I took a look at what they said. They talked about showing photographs one at a time is that of simultaneously and I kind of understood why, I thought we did, why we showed them simultaneously. And then I read about blind administration. I will was a deputy chief at the time and I went in to see my chief and I said you know you're not going to believe this at the crucial point of the investigation when we are going to show a photo array you are supposed to hand your case over to some guy who doesn't even know what's going on. He supposed to show the photo array. This is crazy. We’re not going to do this. And my chief at the time was a great guy and he said well whatever you want. I said okay fine I get another training but were not to make this change. Okay fine. I put my foot in my mouth before. So I went to the training and the instructor who is actually on the in IJ technical working group was explaining about showing photo arrays sequentially in the way that he was doing it which I'll show you the ultimate of how we do it but the way he was showing it is he was showing the photos up in front of them like this in the witness would say no and he put it down. And as I'm watching him something you to myself I know that that's the suspect. The next one is a suspect, do I (gestures). I became very uncomfortable that I didn't trust myself. I kind of giveback boss have begun a minute I kind of get this. I thought we were going to do this. Well I think we should. So the series of events really was that I rewrote our department policy and I did it by going back and forth with our DA's office sent in the end they used it as a model policy in our county and then it went around the state and so forth and I started teaching it. In about 2007 we have a state program for new detectives called detectives basic training which is a transition course. So I wrote a day on eyewitness identification and we do Dr. Ross's presentation was marvelous obviously outstanding I have learned so much in listening to him. But we really cover nothing but eyewitness identification for really the better part of a day and now I do it as a standalone for detectives and I've trained the trainers for other states who wanted you training in this area. So I guess it was sort of dismissing Dr. Ross mention this the so-called estimate of variables which are the things that are beyond our control.
Simultaneous eyewitness identification: I think most times eyewitness identification is difficult for witnesses for reasons having nothing to do with police procedure protocols. It is dark. The exposure is short. It's a stranger identification. Maybe there is a weapon in play which focuses the witness on the weapon. So sometimes identifications are just difficult but from a policeman's perspective it's important to do everything that we can. In other words, if a witness gets it wrong in the wrong guy goes to prison it's not because we didn't take every step we could to make sure that our procedures were fair and modern and progressive. In other words, if the wrong person goes to prison it is really important that as tragic as that is all by itself it's important to me that it not be our fault. So when we make decisions about should we go with sequential or simultaneous, should we adopt these various parts of reforms? I have always been fairly aggressive in that area about if we can do something-- and this comes up when we talk about sequential versus simultaneous-- for me it's about minimizing risk. I understand the metadata that's out there and I understand that there's conflicting studies and I've seen all of that. But I know that relative judgment is real. I know that people really do run comparisons when they can see the photos all at the same time and I'll talk about that later. If we can minimize that risk then we will minimize that risk by only showing photos one at a time. So for me as a policeman that is what drives a lot of it.
Recently I served on the state committee for supreme judicial Court which is our state Supreme Court. And it was a study committee on eyewitness identification. I was the only cop and I was the only non-attorney so some days I felt like a cat in a room full of rocking chairs but that's okay we got through it. The four core that Dr. Ross talked about you will see repeated in the SJC study. You can find the online. It is about 180 pages or so. The police protocols start on page 85. So I am going to talk about please protocols really from the police perspective. What Dr. Ross is talked about you will see repeated.
The protocols as listed in the SJC report start with that every police department must have a written policy. In Massachusetts we are in pretty good shape. We have 351 cities and towns, state police, and a transit police. We recently did some work through the New England innocence project and the Massachusetts police Association working together. We actually got her hands on 253 written policies. Some of those towns are so small they just have achieved and nothing else they don't really have much policy. We took those 253 and we have them assessed by the New England innocence project and they found about 85% of them met the reforms. We do have written policy throughout Massachusetts. There is training going on. Pretty much everybody is on board. Now at this SJC report has come out and we have worked with her back on some pending legislation which will mirror the recommendations of the SJC report and we will finally mandate that any outlying departments that do have policies but on a policy in this area that they get on board very soon number one obviously every police department should have a written policy.
Number two, it advises our offices to separate witnesses when they are interviewing them upon initial arrival at a crime scene. The written policy should explain the police should not use leading questions. Sometimes this can be difficult but this is really a training issue more than a policy issue. When you are interviewing a witness at a bank robbery, for instance, I always hate it when this happens, you are trying to talk to a witness and somebody knocks on the door, a cop sticks his head in and says we got a witness who saw a red sports car leaving the scene ask the witness if they saw a red sports car. Okay I'll do that I'll get right back to you and please stop knocking at the door I'm trying to do an interview. And then you say to the witness do you have a sense to how he left? And that the non-leading question. You don't ask them about the red sports car because half of them will say that they saw because they think that they were supposed to see it.
Always get a description before in eyewitness identification procedure. That happens a 9% of the time anyway but when the subject is stopped directly after the commission of a crime sometimes the police attempted to do a show up before they really recorded what the description is. And you want to get this description down so it can be compared to what your suspect looks like.
Read to witnesses from cards and forms. I will go over the Massachusetts instructions and just a minute but Dr. Ross talked about it. And I think this will come up again. A lot of police departments have photo array forms where the instructions are on the form. I will show you the video in a minute of our department showing a photo array. And it's pretty easy to have a form where you set the witness stand and say the person may or may not be here we will continue to investigate the offense. It is just as important to clear the innocent as convict the guilty. And there's about 11 or 12 instructions. What doesn't happen smoothly I have found is on show ups. A police officer in the field doing a show up right after a crime has been committed, how do you get the officer to give correct instructions first on second of all honey memorialize exactly the instructions that he gave? In Massachusetts we have a rule of criminal procedure around discovery that says in any eyewitness identification case the police must turn over to the defense or the state must turn over to the defense words spoken to the eyewitness and by the eyewitness during the identification procedure. I think that probably exists in a lot of states. So you look for a way to make sure that everyone of your officers carrying those instructions.
Now I mentioned that I've done some train the trainer in other states. I went to Connecticut did a couple of days of train the trainer for their instructors. In fact yesterday got an e-mail from one of those instructors who was going over some things and filling me in on a very enthusiastic female detective I want to say from New Haven Connecticut. The state of Connecticut had show up instruction cards made and laminated, 15,000 of them. Every policeman in Connecticut carries one in his pocket. That is huge for making sure that every policeman who is going to a show up starts and says to the witness now look Ramona show you some people. We don't say a person we say some people we have the plans in their mind that there may be more than one. We may not be going to end here. And then they go through the instructions. My department does that. Massachusetts police departments have been doing that for a number of years. You actually hear the officer on the air just before show up. One officer will call and say can you bring the bank teller up here. And they say yeah Roger okay. They come back on the air and if they were right around the corner I'm going to instruct the witness I will be along in 2 min. And the near silence, they go off the air. What they're doing is that they have stopped and they said to the witness of a looker going to show you some people. The person you see may or may not. These people you see may or may not have been involved in the offense. It's just as important to clear the innocent as it is to convict the guilty. And we kind of run down the instructions. Now Connecticut uses laminated cards in my department did for a period of time. My predecessor, I've just been she for a year and a half, actually had to put it in our department notebooks. These are just blank memo books, notebooks that we issued to officers. It's got a patch on the front. On the inside cover it's got the Miranda warnings. Well on the back cover it says you are going to be asked to view some people and here are show up instructions. So every one of my officers carries in his side pocket in this notebook, our department notebook and on the back cover are our show up instructions. My department also issues a calendar and this is quite a happenstance, it is a calendar where officers can put their court appearances and their overtime and that sort of thing. There is a little insert the goes in them, they can mark their vacation days and it's got radio call signs of the various members of the department, we put show up instructions. The inside cover is Miranda you flip the page and it show up instructions. So part of it is figuring out how you can deploy these instructions so you make sure that every member of your department has them in the field.
Another is to report every identification attempt. There was a time when I think the please only reported on those identification procedures where they got hit. And they didn't report on the ones where they didn't. And there's two sides of this. If your witness looks at a photo array and says I don't see the guy. And then you develop a second suspect in you show a photo array and they say I don't see the guy. And then the third one the make identification it's actually helpful that you have reported on those two previous attempts. By the same token if you show an array and the suspect is in there, they don't make an identification everybody's got a right to know that. So you write a report on every identification procedure.
There is a series of best practices particularly around show ups in the SJC report that they must be done within two hours. That is pretty much a standard I think in most states. Certainly I don't think that there are many states that allow them beyond two hours.
Prevent witnesses from overhearing radio transmission. As a training issue we tell the officer to turn the radio down if they have got a witness and they're on their way to a show up.
Minimize suggestiveness. In the report it talks about never doing it show up when the suspect is sitting in the back of a cruiser or sitting in a cell. In one of her SJC cases it actually says kind of everybody knows the seating plan in a police car. I think that they were a little bit more eloquent but that's basically what they said. The good guys sit in the front the bad guys sit in the back and that's really not a mystery.
Turn suspects of the cuffs don't show. That is pretty easily done. I have seen some states where they say you don't want to do show up with the cost suspect. Sometimes they are cuffed. If they are cuffed properly, they are cuffed behind their back and you can just have them lien against the car and then minimize the number of officers who were there and then do your show up.
I don't know why this is in this particular area but I guess I had no place else to put it. Composite sketches and mug shots are disfavored. We don't show much on books or digital mug files. Nobody working for me has done a composite in the last eight years. We don't do them. They are more trouble than they're worth. They do nothing but skew the memory as far as I'm concerned. I always hated doing them anyway. I used to be a identikit operator. I came back from the school. I said I'm very rarely going to use that. I never liked them. Our department doesn't do them. We have a very restrictive policy that is they get permission from a supervisor they can do an artist sketch if it's a trained artist and their very specific criteria but I can tell you that we turn down those requests more often than not. We really do very few, we almost never do sketches, we never do composites.
Avoid successive identification attempts. He saw this in the Jennifer Thompson piece. I always sensed that it was a problem anyway before I got into this stuff. I remember detective coming in to see me one day and he said I'm going to show a photo array to this witness and issues and pick him to me do a lineup? Now this is before I had any training in this stuff and I said no. If you want to do a lineup will do a lineup. The only person that is in the photo array that you're going to have show up in the lineup is going to be your suspect. And that's where they're going to get the recognition. There was a case not too long ago where there was a bank robbery in a town not too far from me and they do show up. They stopped the guy, bad guy, stopped him on the street one town over 45 min. after the robbery and they took the teller out in the teller sat in the back of the car. So the cruiser pulled up and the guy is standing out on the curb. The witness had been properly instructed and everything in the cruiser pulled up in the witnesses sitting in the back and she goes in, first bad sign. I tell officers to drive away. If you're not getting a visceral response it's not going well anyway. But the officer did sit there and she went hm, boy, hm, boy. I don’t know. A short time later I was talking to one of the detectives he says do you think we should do a photo array? And I said no. He says but I think shall pick him. I said I know shall pick them. What I don't know is whether he robbed the bank. I understand that sounds close but there's a difference. And that's the way to view it. You got to be careful though successive identification procedures. Once you done one any given identification and in some cases where you don't and you do successive attempt there's going to be recognition. And Dr. Ross we call him in Boston wicked smart, he probably understands how the mind actually works in these areas. I just know that when a witness sees the suspect the second time plus they're able to kind of associate them with the event that's being investigated that there's more of a likelihood that you're going to get an identification the second round and it's really some to be careful of. So I tell officers that have a class you put all your effort into your best identification attempt. In that bank robbery, your best shot at getting identification was not show up. If she says no or doesn't make identification, he might be the bank robber but you're all done with the identification part of the case.
Best practices for raise in lineups. This seems like a simple one but I actually, and I wrote a lot of the protocol stuff that was in the report, use a current photo of the suspect. This just has to do the diligence on the part of the detective. You don't just pull a file photo and thrown in a photo array and I've seen this happen. And the officer says to she couldn't pick them. Another officer says he didn't look like that. He's been on heroin for six months. He weighs 30 pounds less. He's lost a lot of weight. Last time I saw him he had shaved his head. Well I'll show another photo array. Well how many times are you going to show the photo until the witness gets it right? No, you put all of your effort into the one showing and use a current photograph.
Fillers should fit the description of the offender and not look like the suspect. This is somewhat contrary to what Dr. Ross had talked about but close. For me as a policeman it's really about the balance. Start with the offender description and then get your array of milk for outliers. Does your suspect stand out? Do any of the fillers stand out? It should be good but not too good. You don't want everybody to look like the suspect. You're not trying to confuse the witness that you are going to give them a selection you wanted to be a fair selection. And when you're all done building it show it to another detective. It's like showing somebody else your search warrant before going into the court just to make sure you haven't missed anything-- that nobody stands out. So for me it's more of a balance. You want to test the witness not trick the witness. There is a 5+1 rule in Massachusetts at least five fillers, now we just react or model policy actually within the last couple of weeks as a result of this is JC report and added a few things to it. The model policy calls for seven fillers. It's a huge bump from the required five but we thought it was something to do. Most police departments in Massachusetts show photo arrays of eight anyway. But the key is only one suspect. And I will tell you that there was a time in my career where you would run a composite the newspaper and six people would call and say hey you know who that looks like? And you take all of those photos and you put them in an array, we always believed that if the witness said that's the guy well that was the guy. The problem with an all suspect as an array is that you don't leave any chance to detect an error-- the anybody the witness picked your middle unlock up. They're all bad guys. It's not like there's really known innocents. Your fillers should be all known innocents, not suspected of involvement in the crime. If you get six phone calls saying this might be your guy you really want to do six photo arrays each one with one of those suspects but that's really the fairest way to do it. That wouldn't necessarily be the best way to go.
Blind administration is required. They talk about that separately. I'm also going to talk about sequential showing. Anybody who has worked for me since 2005 has shown every single array double-blind sequential. It is not an implementation problem for us. It’s not that expensive. We have gotten no kickback from the detectives working the cases. Just kind of empirically we hear back even from judges who sat on this committee with me that the, it's a more polished procedure. In our detectives don't get cross-examined on the stand about why they did what they did. If you show photos simultaneously your detectives was simply here isn't it true that the research runs against this? Isn't it true that the US Department of Justice and the innocence project recommend you do it sequentially? Once you go sequentially you don't get those questions. You're on board with the research. Only the last thing that the defense attorney wants to hear is the detectives say why we only use research-based identification procedures recommended by the Department of Justice and the innocence project. And they are all designed to minimize the likelihood of an innocent person going to prison. I don't know how many you're going to put the defense bar but you don't want that question coming back from the stands. And even judges who sat on this committee with me told me you know we don't hear cross-examination about procedure because really across Massachusetts the stuff has all been adopted.
And another thing I'll mention, this is in the SJC study, is no more than a second lap--I kind of struggle with this policy-wise and trying to figure it out. You got a witness, you've got all through the array and the witnesses can I see number four again? Or can I see the array again? It doesn't matter, the answer is the same. I can show you the entire array one more time. See don't tell them going in that you can see it twice but after they have completed the array and they asked the question you told him and you put them on notice. You can see it will more time. So onwards this is it. And then you completely show the array of second time. So that's how we handle lapse.
We take an immediate statement of certainty and you'll see this on the video where when the witness stops at number three and says that's the guy. The officer says immediately without using a numerical scale how certain are you. And we use that instruction without using a numeric scale because I really don't want, you know let's face it nine on a scale of 1 to 10 give me reasonable doubt. And the numbers I don't think tell us as much. I really want the witness to say yeah that really looks like the guy. I'm pretty sure that's him. And the days of the positive ID or the not positive ID are kind of out. It is what it is. You look at what they say. You assess what they say. And your investigation proceeds from there. The problem is, and you'll see this in the instruction sheet. We added an instruction after conferring with Gary Wells. And you've heard his name a couple of times. If you stop and array at the second photo and she says that's him and you say well how certain are you on certain okay that's it. Well you’ve shown an array of two photos. So how do you get over that being unduly suggestive? Yet if you just kind of March on what is the witness think? Well I must've been wrong. I should pick someone else now. So how do you handle that? Will we added to the instructions: I am required to show you the entire series. So that when they make a mid-array identification we take an immediate statement of certainty: without using a numeric scale how certain are you? I'm positive that's the guy. And our officers write a demo the now we videotaped to see that the minute as well. And then after that statement of certainty is taken the officer reminds them nonmember I'm required to show you the entire series. The witness says okay right. And we show them the rest of the array. Now number one we ensure that all of the photos are shown. Number two know what if the witnesses oh my god, look at number six. I thought it was number two but it might be number six. And detectives say will that's bad. We have that's bad but is it? Were they sure about to question mark as to the suspect? You feel one way if two is a suspect you fill a different way if six is a suspect. It is with this. Really it is what it is. Don't put in your report I got a positive ID. I don't want to see this. The witness looked at the photographs and this is what she told me. And then you assess where we go from there? Where do we take the case?
Shuffle between witnesses was the other. If you have got to bank tellers and you show the array one way to this teller that you show it in a different order to this teller. And that's really for our benefit. It so that we can demonstrate that there is no collusion between the witnesses. Witness number one didn't tell witness a virtue it's number three. You now text: It's number three. Could that happen? Yeah it could happen. We don't think it does much but now demonstrated that it could not have.
Submit the array instruction form with the report. And, whenever practical, video or audiotape a photo array or line up. There is a great deal of debate in the city committee about that in the end I bought into it and I will show you just how we do that.
Some of the recommendations: that the court take judicial notice of legislative facts, in other words, that they take judicial notice of certain science: that weapon focuses real, that people have a harder time identifying people of other races. The exposure duration is an issue and that sort of thing. A lot of it has to do with estimated variables that some of them have to do with police procedure. Failure of the police in a substantial way to follow the protocols requires a hearing and there also jury instructions the lineup with that. So if the show up was done three hours later or the officer did not use blind administration there is a jury instruction that says there is basically a right way to do this and the police did not follow their protocols and you should way that. The admissibility of certainty statements: there was great debate about that. In the end of certainty statements are allowed if the SJC follows these recommendations the only certainty statement that would be allowed as the one made to the officer at the point of identification because you do see a witness looking at a photo array those I'd have to say it's number three I think he looks pretty good. And on the stand they say I'll never forget that face. There is an escalation certainty or there can be. I think the most accurate certainty statement is the one that occurs right at the point of the photo array so that will come in really is a form of hearsay although it's going to be the only certainty allowed. Witnesses will no longer be allowed on the stand to talk about how certain they are that they made the right selection. We are only going to look at the statement taken in the field at the time which is a little different. Around courtroom testimony it's basically the same thing. The purpose of that is that witnesses will be able to make an identification from the stand by the judge will not allow examination or cross-examination around their certainty.
I think you already know this: show ups are allowed for two reasons. Number one they allow the efficient capture of the subject and number two they do allow us to eliminate innocent people and it lets the innocent be on their way. Not a good time guys. Not a good time to do is show up.
So what's the procedure for doing it show up? Number one you turn your radio down. Number two use your notebook or calendar book but you find your show up instructions and revenge of the witness. The preference is to transport the witness to the subject not the other way around. It's allowable in Massachusetts but there is a strong preference for transporting witnesses not suspects. Manage multiple witnesses by putting them in separate radio cars and ensuring that they don't confer with each other. Note the witness’s reaction and use that certainty statement without using a numerical scale how certain are you?
In preparing a photo array I already talk to a little bit about the 5+1 rule in the Massachusetts model policy that went on to the Chiefs just last week that calls for 7+1. Some of this I think is repetitious about the suspect photo.
Mitigating unusual features: sometimes officers run into this and it's more of a training issue. It is addressed in our policy but if there is an unusual feature on the suspect our rule of thumb don't mitigate the feature of the suspect, mitigate it on the fillers. The question the detective doesn't want is detective isn't it true that before you showed a photo of my client to the witness you Dr. the photo? Well there's a good reason. “Yes or no detective? Did you Dr. the photo?” You don't want to do anything to the suspect's photograph. You want to do something with the fillers. Now this is kind of a well-known photo array graphic. I'm not sure where the case came from but the police are provided with the description of the cross eyed black man. So they went through their files looking for fillers of cross eyed African-American males and found just a few. And then put this array together, very nicely done. Only one of these people really is cross eyed they photo shot the eyes on the others. They did not touch the suspect, only adjusted the fillers. And created I think very fair display. Could there be a motion to suppress? Perhaps, but I think the state would win it. I think that this kind of diligence and effort on the part of the police is helpful.
This next one that you're going to see is actually a true case. It's in the town where he used to work before he became chief a year and a half ago—Wellesley, where Wellesley college is located. The story is that a woman came home, there was a car parked in her driveway. There was a man walking away from her front door. She said who are you? He said I'm here for the roof estimate. She said I don't want to roof done. Who are you? He gave an address. It was a false address. She said that's not the addressee or. He said I'm sorry must be mistaken. He got in his car and drove away. She jotted down all six digits of the license plate. Now you and I both know that every time you get every digit of the license plate the car is stolen. Nobody always gets all of the digits of a plate that can actually be tied back to the subject. So, sure enough this is a stolen Jeep. We contacted the entering agency and had them amend that the car was used in a housebreak call for prints and so forth. And the NCIC entry two days later they arrest two guys in the Jeep. Now the story is what had they really done when they were in her driveway in the stolen car? Well when she went to let herself then she went up to the front door and the glasses broken right beside the doorknob and they were in the process of breaking into her house. She was asked for a description she said kind of a light-skinned black male. He had a white stripe of hair (gestures). And I just wanted to see what this guy looks like because I was really interested in (gestures), what that meant. Sure enough sure enough when these two guys were arrested in a stolen Jeep-- this is four photos from the eight photo array-- I want you to tell me which one is a suspect. Here it's getting harder. I don't know, any idea? There actually was two. I thought the detective did a pretty good job photo shopping the hair to create a white tufted chair in the filler photographs to present a fair photo array. It was number two. She did pick number two. There was no motion to suppress. He pled out and did time. I think if he had not done that the array would have been suggestive.
This is the photo array instruction form. I'm not going to go through all of it but there is that number three: please look at all of them I am required to show you the entire series. Number four: make a decision about each before moving on. And then the “may or may not instruction which is the biggest of them all-- I think most people agree. Number five: it's just as important to clear the innocent. This is important. Don't just do blind administration. Make sure that the witness knows it. The way we do it, it's pretty obvious that that's going on but we have added to the instructions of blind administration is in use here. In a photo array but not to show up that the suspects may not look exactly like they did on the day of the offense regardless of whether you select someone we will continue to investigate. We're going to ask certainty: number 10. 11: please don't ask the officer questions when you are done. 12: regardless of whether you select someone please don't discuss this with other witnesses and do you have any questions? So essentially 12 instructions and the question: do you have any questions before we began? That is our instruction form.
So let me just talk a little bit about simultaneous versus sequential. I have seen simultaneous referred to as the traditional way. Dr. Ross actually mentioned it. It really wasn't. The traditional way was that-- if any of you have uncles or grandfathers who were in law enforcement, they would carry a stack of photos in their soup pocket and they would pull them out and give them to the witness and the witness would kind of flipped through them. It was a sequential photo array if you think of it. We went to simultaneous. It didn't start off as the traditional way. Why did we go to simultaneous? Two reasons: number one is that as the witness was flipping through the photographs if they stopped the second one it might be suggestive because they only looked at two photos. We talked about that earlier. And then some guy somewhere came up with these folders with the little squares cut out and I just imagine he's worth millions because just do the math-- there is 18,000 police departments in the country. Everybody bought a couple of cases of them. They went right in your case file. They were excellent. So those are essentially the reasons why we went to simultaneous. The reason I mentioned that is it had nothing to do with accuracy. The police never selected simultaneous photo arrays because we thought they were more accurate. They were four other disconnected reasons. Simultaneous arrays cannot be more accurate than sequential. First of all, I know that relative judgment is real. I know that people run comparisons when they see all the photos at once and I can prove it. If you've ever seen this happen, if this were a police audience I would ask erasure hands. You have certainly seen it before. When you lay all of the photographs down loose on the table, you may be eight of them out or the six or whatever you're using and you say the witness take a look. The person may or may not be here. Do you see anyone? More than half of the time which is the witness begin to do?

Audience

Sort them.

Chief William G Brooks III

Thank you. Slide the photos away. None on never struck me. But it does now. What is that? What's with moving the fillers? Fillers don't count. Fillers have nothing to do with it. Fillers are simply put in there and so that this procedure won't be suggestive. They've got nothing to do with the case. Why would a witness move fillers? There is only one explanation. What is it? They are running comparisons. They're pulling out the ones that don't look good enough. Now, if the offender is in the array you will always get more hits with the simultaneous showing. Because witnesses who do the comparison, essentially the offender looks most like himself. And they will use that comparison to find the guy area well if he's the offender and you say that cops they'll go okay was wrong with that? Well what's wrong with that is, and they would understand it, what's wrong with that is we don't get to know that the offender is in there. The purpose of the array is because we don't know if the suspect was the offender. We’re trying to determine that. And when your suspect is not the offender but looks the most like the description that's when they get to him using this process. And sometimes it's a confluence of factors. It's not just simultaneous versus sequential, okay? You get a 7-Eleven robbery. The subject is in the store for 1 min. After the robbery you take that witness down to the police station, you spend an hour building a composite. When you're all done what is the most stable, memorable image of that night's events? The way the offender looks or the way the composite came out? Now advertise the composite and get suspects who look like the composite. You get a suspect…(inaudible 35:02) He looks like the composite, he looks like the description. Put him in an array. Now bring in a witness who thinks that the suspect must be there. Who are they going to pay? That's who they're going to pick. And if you put all of the photos out at once it allows them to find their way to the suspect using relative judgment, using comparison. So for us to move to sequential arrays was simply risk-management to simply remove the risk that that could happen. But it's more difficult to run a comparison where people only see one photo at a time until you know. And you put it down. And you show them another one. They say I don't think that's him. We put that down and you pick up another one. For me that's what it's all about. Yes, when you look at the metadata you see that there are higher accurate pics with a simultaneous array but that can happen when witnesses are using relative judgment and making those comparisons. So my point is, from my perspective there's no good reason to hang on to simultaneous arrays. We discarded them in 2005 and everything we show now is a double-blind sequential, all the time.
There is two ways to do what is call blind administration. One is a double-blind system and that's a winner shown in the video where you actually use a second officer. The other is a so-called folder shuffle and I do have that video although I don't think I'm going to have time to show it. But the good news is if you go to the Norwood police YouTube channel I have put those videos up. Actually, somebody smarter than me has put the videos up but there, there. You also see some really cool videos about our new charger police cars and watch that to. But the stuff I'm going to show you here is also on the Norwood police YouTube channel. The folder shuffle basically is, and we made that actually with the New England innocence project helping us, it's basically put every single photo in a separate folder and then you shuffle the folders. So, if another officer in the room knows which one is the suspect, he can't see when me witnesses looking at that photograph. The witness opens up the folder, looks at it, closes it and says that's not him. And then you go to the next one. You could really teach yourself to do the folder shuffle method just by watching this video. It's an instructional video and you can check it out on our channel. But we use the double-blind system where we use a second officer. All of our photo arrays are shown by uniformed patrol officers. The detectives use uniformed patrol officers to show their arrays. Rebecca has heard me tell the story. We were going out to do one of these talks in another state. I forget where. A couple of months ago, and it actually was the night before I was going to leave and I hear over the station intercom. A detective says: any officer available to show a photo array call extension 192. And it struck me. It's that easy. He never repeated it. Somebody picked up the phone and called 192. And when you see the video you're going to see how easy it is for a patrol officer to show a photo array. Because they don't of the suspect is and if they do we just picked another patrol officer. It protects the innocent from inadvertent cues as Dr. Ross said. It takes away a defense strategy. They can't ask us about, did you smile? When you got to number three did you smile? Did you say thank you when she handed that photo back? Did say thank you when they had to back any of the others? Did you say, good? Number one now we videotaped this. Number two the guy showing the photos has no idea who the suspect is. So it really takes away the defense strategy. And if you get the chance to introduce it, it has a positive impact on the jury.
I think I said enough about statements of certainty. So this is about 5 min. and what you're going to see here you're going to see it all pulled together. You're going to see a plainclothes officer walked into the room with a patrol officer sitting there. The detective, we were very concerned about witness rapport. The witness who has been dealing with the detective, the detective who's been dealing with the witness all the time will actually do the instructing. The detective only leaves the room and the photos are in play. You see it's a very simple exercise for the patrol officer to do. He's going to show the photo array and take a statement of certainty and then call the detective backend. And then we let the witness go. We don't even discuss in front of the witness. We try to completely sanitize the system.
39:16

Speaker one

Mrs. Kady this is Officer Fundora and he is going to administer a set of photos to you after reconstructions to you. And let me remind you that everything that's taking place right here is being audio and videotaped. It's on the record. I'm going to read a set of instructions and officer Fundora is going to show you a photo array. You are being asked to view a set of photographs. You will be viewing the photographs one at a time and in random order. Please take a look at all of them. Officer Fundora is required to show you the entire series. Please make a decision about each photograph before moving onto the next one. The person you saw may or may not be in this set of photographs of you are about to view. You should remember that it is just as important to clear innocent persons from suspicion as to identify the guilty. The officer showing the photographs does not know whether any of the people in the array are the person that you saw. The individuals in the photographs may not appear exactly as they did on the date of the incident because features such as head and facial hair are subject to change. Regardless of whether or not you if you select someone, the procedure requires the officer stay in your own words how certain you are. If you do select a photograph please do not ask the officer questions about the person that you have selected as no information can be shared with you at this stage in the investigation. And regardless of whether you select a photograph please do not discuss this procedure with any other witness in the case or the media. Do you understand everything before we begin?

Mrs. Kady

Yes.

Speaker one

Okay if you understand that I'm going to ask you to please sign that top line with today's date of July 3. Now I'm going to leave the room and officer fun Dora is going to show you the photographs and he will let me know when you guys are done at the end.

Officer Fundora

Are you ready to begin? I would pass this to you one at a time.

Mrs. Kady

No. No. No. No. Yes this is him.

Officer fun Dora

Okay without using a numeric scale tell me how certain you are.

Mrs. Kady

I'm positively certain. I would never forget him.

Officer Fundora

Could you please initial at the bottom. Just remember I have to continue showing you the rest of these photos okay?

Mrs. Kady

Okay. No. No. No.

Officer Fundora

Thank you for your help in going to walk you outside.

Mrs. Kady

Thank you.

Speaker one

How did you make out?

Officer Fundora

All right she picked one off a positive look sir.

Chief William G Brooks III

So the point here, he's just going to report on how they made out with that array. We do that outside of the presence of the witness. The witness really has been sent home we will get back with her later. It is time now for us to kind of assess how that went. When we videotaped like this and we just started, when we got to this point in the SJC project that this is going to be one of the recommendations I just told the guys that from now on just start videotaping. Now of course is part of our policy and other departments in the state will do it. It's the same as videotaping suspect interviews. The way we do it in Massachusetts, we have a one party consent rule in our wiretap law. Just means that no recording can be secrets of the detectives simply flips a switch and he walks in the room and says everything here is being recorded. Thanks for coming in and he goes right into the interview. If the person stops and says I won't be taped then at least we have the refusal on tape. If they don't react and it is in secret and we have recorded the interview. We use the same process and the showing of a photo array. We no longer see detectives being cross-examined on the stand about what happened in the interview room. Judges essentially start every hearing by saying okay has everybody seen the movie? Counselor what's your point. And that's how motions to suppress on interviews are done. We think the same thing will happen with the showing a photo arrays.
I've covered that.
Unconscious transference is that issue that I talked about before that comes from successive identification attempts. It can also happen when you show, which is why we don't show mug files we show somebody kind of 100 random photographs and a mug book, first of all that's going to do nothing but scramble memory just in my own humble opinion. I don't know I think that, I just think that. But then what if one of those people who is in a mug shot becomes a suspect and shows up in a photo array and the person picks them did they pick them because they recognize them as the offender or because they saw them yesterday in the mug book? So again this risk is created by successive identification attempts, which includes the showing of mug files.
Response latency just talks about the issue that quick identifications are likely more reliable and I know there has been some studies on that. Dr. Ross could probably tell you more about it but I think if you ask people she don't you want a witness who really takes their time, thinks and looks to the photographs? I think you heard Jennifer Thompson, I took my time. Really, it's the jump out identification that can be more reliable. And when you say you get a sense for that. This runs literally about 2 min.

Speaker 1

Once we figured out what the science was telling us we built our practice around the science. Speaker 2: This is the new way lineups are done by the Dallas PD-- one of the few departments in the country to videotape the process. This woman is trying to pick out a suspect in a homicide case.

Speaker 3

First of all, lineup administrator and my job is to show up for all the lineups.

Speaker 2

Lineups are now administered by an officer who knows nothing about the case and who doesn't know which photo is a suspect. Also new, the officer explains the suspect may not be in the lineup at all in the investigation will proceed whether or not there is an ID.

Speaker 3

You can pick it up if you want to.

Speaker 2

And as you can see the photos are now shown one at a time.

Speaker 4

That's him!

Chief William G Brooks III

All together do you think that's him? We do have some concerns. I'm going to really skip through this stuff because I'm running out of time-- on how police officers make identifications from surveillance media. About their is a restriction there that's in our policy: they cannot use their cruiser laptops for eyewitness identification. It's just too easy for them to pull up a driver’s license photo and say to the witness is a him? So we don't permit that. That's essentially a photo array with only one photo. You can do lineups that are double-blind (that's another famous line of photo) you can do lineups that are double-blind sequential if you go to the Norwood police YouTube channel you will also see an instructional video on how to do a double-blind sequential live lineup. I just give you this slide to demonstrate how it can be done. This is set up for it. Because it sequential people don't line up in the lineup are many more. Police departments that have an interview room which typically has an observation room that backs up to it can do a double-blind sequential lineup. You see the leftist the witness room. There is a police officer, assistant DA, an attorney there. And on the right side is just a police officer. The room on the left is the witness room. The room on the right is the participant room. You essentially load the witness room by putting the witness in there with the officer and the prosecutor then shut the door. And the six participants are lined up in the hallway not in the room. They step in one at a time, step up to the glass turn to the left, turn to the right then leave the room. Then number two comes in. So that's how you do a live lineup that's double-blind sequential. Logistically there's a little something to it. That's why we have made an instructional video. If you're interested you can take a look at that.
We don't show mug books but no presentation would be complete without the Nick Nolte mug shot. We don't do composites. I think Artie talked about that. So just for implementation the way to do this is I do believe strongly as Rebecca does in the model policy that can be standard police policy throughout any particular state that everybody's doing it the same way. We have done it through our Massachusetts police Association. The second thing you do after your Department has implemented the model policy is deploy the forms and cards to make sure that everybody is carrying them and has them and so forth. And lastly, provide training. Massachusetts police academies now offer half a day of eyewitness identification to all recruits. Periodically cover it in service training or in roll call training as a reminder. Within six months of promotion to detective a new detective gets a standalone course on eyewitness identification.
If there is anything that I can do for you or if you have questions, my phone number and my e-mail are there. Feel free to reach out to me. I hear from people of been in these audiences all of the time and I am happy to answer questions or chat with you. Or help you out in any way that I can. I do have the issue with the flight and I will be here for another about 20 min. or so which you will be able to talk to me then. If you questions about my presentation I would be happy to answer them now.
I know this is being videotaped so I'll repeat the question. The question was in the videotape the officer at the end says thank you very much for your help and when you can't think the witnesses of that kind of an indicator that they got the right person? Number one it's difficult to tell an officer don't be nice. Don't be thankful. But seriously it doesn't have any impact if the witness knows that the officer doesn't know whether she's picked the right guy or not. So once you have a blind administrator in the witness knows that and you say thank you very much for your help well this guy doesn't know what he's talking about here is of the suspect. He doesn't know who I picked you could pick the filler. You could've picked the suspect. You don't know. So I think it minimizes the effect of kind of a thank you at the end when you know that the officer showing the photos doesn't know what he's talking about.

Question (51:32)

I guess… Thank you for your time is polite and thank you for your help could… Reinforce someone's…
Yeah could've the officer knew what he was talking about but the general school of thought is that if you're using a blind administrator you minimize the effect of any of that. That's just my view.
Yes ma'am.

Question (50:51)

(Inaudible)

Chief William G Brooks III

The question was what about in the photographs, the fact that people are dressed differently? You would want to be a little bit careful. Actually the person she picked, I think it's Lindsay Lohan's father I think that's why he was on celebrity mug shots but he was in a shirt and tie and everybody else's casual. I suppose that something you would be careful of. But there will of thumb is to make sure that nobody unduly stands out. If my suspect was in the shirt and tie would make sure that the others were dressed similarly. But I think that beyond that as long as they are in casual wear or everybody's innocent I think that's probably fine. And the gentleman behind you had a question.

Question (52:40)

Yes so you're talking about the guy gives the instruction where they say the person may or may not be in this photo lineup they may or may not be in this array. Do you ever or how often do you give a person, a witness, a photo array or line up where the suspect is not in it?

Chief William G Brooks III

That actually came up in the SJC committee. Somebody was advocating that as part of every investigation the police can just call the witness then and give them an array with no suspects and it. I guess I get why. But that would be really weird. It's pretty obvious the officer have been pretty good about everything that's happened so far and the idea of kinder showing a dummy array would eventually get out that in every case you're going to be asked to come in and review a photo array. And these people take time off of work. So we kicked it around I understand the reason behind that. First of all of the offender may or may not be there. The purpose of showing the array is that we don't know who the offender is. Just because a person is our suspect doesn't mean that he did it. The person really may or may not be there and we think that that's a sufficient instruction. And that's the recommended instructions so that is what will follow. Another question that's for me?

Question (53:55)

Do you have an opinion on or any procedures relating to the use of voice identification in lineups?

Chief William G Brooks III

The question was do I have an opinion about voice identification lineups? Yes I did some voice identification lineups when I was at Texas. One around a guy who was a serial masked bank robber and the best part of, it went up to a mess appeals court, the best part about it is he didn't show up and then tried to testify on his own behalf that he was cooperative with the investigation. You've got the right to remain silent you don't have the right to not sure fair lineup. So actually that came in against him and the bank teller actually pick him from the voice lineup. Different states have different guidelines. Massachusetts is particularly strict. If you have done a visual lineup and you're going to follow it with the voice lineup you have to instruct the person that the members of the people they saw have nothing to do with the numbers they are about to hear. The second requirement is that they cannot speak the words spoken during the crime they have to speak something innocuous. Typically, a couple of paragraphs is cut out of that day's newspaper. And the witness cannot see the person speaking. So like on television where it's number three step forward give me the money you dirt bag. It's not really done that way. You do a visual lineup first and then you tell the witness they're going to see different people with different numbers. Here are different people with different numbers. You shuffle the order and then he just to the voice identification part. It's only recommended when the witnesses blind, when the subject is masked, when the light is so low or when there is something unusual about the person's voice. These are the standards in Massachusetts. Yes, sir.

Question (55:33)

Yeah you indicated earlier in your talk that in a photo array the photographs should be similar but not too similar. Now aside from the identical twin problem that I think is kind of obvious what's the problem with greater than ordinary similarity? Another words at some point if there actually identical photographs are really not learning anything, it's essentially show up but the fact that they are more more similar it would seem to me would require greater and greater discrimination from the witness in terms of their identification decision and therefore would render their decision assuming they identify someone more reliable and given the enormous weight the juries place on this kind of evidence that seems like a desirable lineup.

Chief William G Brooks III

Yeah the question was so what is the problem with having an array where everybody's too close? I really only seen that happen a few times when it was that bad that you really would have a hard time distinguishing among the people. One case there was sex assault a kidnap sex assault and I was looking at the photo array, the subject was a 50-year-old thin male with a shaved head. And honest to god I knew who the suspect was and I was having trouble picking him out. Everybody just looked really looked the same. I think you can get it so it so close that you’re not really testing the witness if they can pick the guy. You're closer to tricking been testing I think. So I think it's a judgment call. It doesn't happen all that often that the array is that close but I have seen it happen although not that often. And I think just because you have other speakers here that I should probably stop.

Paula Wissel

So now were going to hear from David Angel who is the assistant district attorney in charge of conviction integrity in Santa Clara County California.

David Angel

So you know after a while you start speaking to groups more and if you haven't done a lot let me assure you speaking in our right before lunch is the most coveted spot so thank you. Actually last time I spoke, I'll tell you this, I was the first person to speak with but will good so people will be fresh. And I was up and northern, well I live in Northern California, this is much further north, a small rural area. I was the first speaker but before introduced and they talked about somehow the topic came up for the first person who wishes introducing the event was about how they felt in the area of the state should secede from the rest of the state because they were tired of experts coming in from other parts telling them what to do. And with that here's David Angel from Santa Clara County. So I'll tell you what I told them I was invited. I did not come saying all have to do this.
So I have been a prosecutor in generally for 19 years. I'm a prosecutor in one of the things I'm in charge of now is conviction integrity but that's not been my whole career. I started trying cases like everyone else and I've done a pretty wide array now. There's definitely colleagues of mine who have more hard-core trial experience and I do but I have done my fair share. Right now we do our kind of our writs and appeals are law and motion. I'm in charge of our North and South County units, are training among our grading committee and I do officer involved cases as well is conviction integrity. It's not the only thing I do.
Before I start, and I hope not to speak for too long but let me just tell you something that happened earlier in my career with this it is influenced me a lot. And it involved a case I was reviewing and this was to DAs ago. So under our pretty conservative law and order DA who asked me to look into this. So this woman, she was asleep in her home and her toddler was in the next room and she woke up. It's just like every person's nightmare. She woke up to a man who had broken in and who had a knife and all she could think about was just getting through the night and protecting her child. And she succeeded by both of those things. But he did rape her before that happened. And the next he was leaving and what I can only imagine was just like a last gesture of cruelty he just grabbed some new tracks off her night stand you know things that would be valuable to her but I can't imagine had any real worth. And there wasn't a whole lot of other evidence aside from her identification. They did find these knickknacks in this person's home and he was subsequently convicted. Well years later actually I think it was a referral from the Northern California innocence project, and they said that this is a mistaken identification. If you look at the data, you know the information that I had, well you know she did have a pretty long time to look at him because this didn't happen right away but the lighting was pretty poor. There was only some streetlight through the window. And it was cross racial. And he had a knife so there's that weapon focus and if you do a lot of that work that's kind of what people look at. The way I conceptualize it intellectually is you think of how we evolved, if you're being chased down by a will for a lion or something we evolved to escape. So your adrenaline floods your body and you evolved to escape. You didn't evolve to say well I think it was a wolf with under that white patch above his year that's not really all that relevant. Again, this is a lawyer talking not a scientist but that's why kind of makes sense to me that we didn't initially evolved to have perfect recall for recognition of a specific face because you want to evolve to know that lions are dangerous and wolves are dangerous not that particularly one is the one that took a chunk out of you. And he actually had a not crazy story as to why the nick-knacks were in the house. He said well look here's the deal there is this guy and he identified this guy when he is the sort of homeless transient ne'er-do-well but my mom was this really big hearted woman and she would let him stay there sometimes and sometimes give the sandwiches and he gave her this stuff just as a gift to say thank you and she took it that's where it was found there. When I investigated I found that guy was in a local jail around that time. So this wasn't the famous owner was John but almost last in. This was an actual person. So I thought wow because this is a pretty big deal. And then we searched an amazingly we found biological evidence which is why is a prosecutor I think it's so important to keep this evidence of its biological really forever because we never know if the case is going to get reversed by the Ninth Circuit 15 years later so I was want to be able to retry the case and they always want to be able to look at that evidence again if some new issue is raised. But we did have that evidence in the lab looked through it and they could actually detect biological material on the sheet and then ran it and it came back and they confirmed that this guy was in fact the rapist. So this is this guy.
This is one of the first cases I worked on. And then I sometimes ask people for multiple-choice which I want to now but you could raise your hand. Hey multiple-choice: this guy in the paper with some laudatory article about the prosecutor who prosecuted this the first time around, did I get invited onto Oprah to talk about this case I found? And the correct answer, where their symposiums where I was interviewed? No. What happened was actually nothing and nobody knows about this case. So you know you hear about these things when something goes wrong you tend not to hear about them when they go right. And the way press works there is this sensationalism to it. And I can guarantee you in conviction integrity overwhelmingly the cases I review the person is guilty and we either find more evidence of it and the bulk of my work in conviction integrity is defending the prosecution's we have. My intent is to get a fair shake and we do. And we have a lot of procedures to do that but I can tell you that the bulk of my work and I think the chief pointed this out: we adopt this procedure and it solidifies the good work that we do. It will also help us to detect, help us to detect when a mistake was made. But the payoff actually comes much more frequently.
Now I just point out, this is Santa Clara County in California because I don't know if any of you know where it is. And the reason I point this out is not so much because I want you if you decide to move their, come work for office although that would be cool. But to give you an idea that, I just want to tell you about our experience in certain lineup procedures and I mention this because we actually aren't very different profile than what you just heard. We are a fairly big County-- 1.3 mi.² but we also have 1.8 million people so it's a fair size. I mean a huge diversity in our population centers. Certainly racially, ethnically, class, education and so on but what's really important for this analysis is we've got, San Jose is a pretty big city with 1 million people but we've got Morgan Hill one of the areas I'm also responsible for that's only 39,000 people so it's a pretty small area. And we have unincorporated areas that are even smaller still. We have a very diverse economy. I point this out not so much for a geography test but to show that we really have very different agencies of very different issues. On top of all the normal things we have Silicon Valley. So we've got Apple and Google and I think Cisco and HP they're all there. And police departments range from Morgan Hill’s got 36 sworn officers. San Jose PD has over 1000 sworn officers. And the Sheriff has I could break it up but about 1800 sworn and unsworn. And I just point this out because what we've done can work in our small agencies and it can work in a very large agencies. So it doesn't mean you have to do it just means they can work that way. And we have about 180 prosecutors in our office. We do about 35,000 cases a year. I hasten to add these are not 35,000 homicides some 35,000 rape. It's just 35,000 cases altogether. It's like a pyramid you know most of them are not that serious and then suddenly we do have some serious cases to.
So let me just quickly tell you what we do, do. The first is all of our law enforcement agencies to have their own protocol. We don't have a hierarchy whereby the DA can just say hey this is what everyone is going to do. So everyone has their own protocol and we’re happy with that because we want each agency to have some flexibility. Enough you have 36 officers you might want to do something differently than if you have 1800 officers. But, they are consistent within the department. So the words if Gilroy PD is running a lineup every officer within Gilroy is going to do that the same way. So you don't have this problem as a prosecutor where you don't have any idea really how this lineup was done until you get the person in court. And I should add more or less right you never going to hundred percent. There's always going to be some errors in the process but that's just you know. You know I'm sorry I don't really know the composition of this room but if I'm with fellow prosecutors or with a peace officer and mean it's just the way it goes. Mistakes are made. That's just always going to happen.
But while I said they'll all have their own protocol they all share certain characteristics. We have guidelines for fillers that were talked about. My experience is the same thing as the Chief, you can't make it entirely built on the description because you don't often have a good enough description. And you also run into a problem, I mean first of all most of the time the description of the suspect should in fact match. So were talking about the minority of cases where for some reason they don't. But if for some reason they don't and the witness said oh it's an African-American man your suspect is Latino you don't want five African-Americans and one Latino. So you need to use some common sense on it. And I don't know whether entire research will guide us in a better way to do it but we kind of try to do that meld. Match it to the description but if there someone this is going to make it suggestive don't do it that way.
Taking confidence statements: admonitions and what that means is I think you've heard it a few times but really it's just important to tell the person hey the perpetrator may or may not be in this lineup. It's just as important to clear innocent person. And were going to keep investigating the matter what happens. I can't answer the question recent hey why don't you sometimes just to dummy lineups where nobody's in it so that people don't know the suspect has to be there. But the issue isn't that the suspect is in there. We assume they think the suspect is in their but what is important to tell them is that the actual perpetrator might not be there. I mean that's the thing that you're testing. So I don't see a problem with having to do fake ones. Again maybe research little something different. But the important thing is we’re telling them the guy that did the crime may not be there and it's just as important to clear someone is not.
Now we have done double-blind for quite a while now and we do use the sequential method. And in the new thing we've added we now videotaped them I think in last couple of years. And I really want to, videotaping is just so powerful so if you're a law enforcement agency or prosecutor many of these I've used win-win and this is a win-win right. The defense bar well it's kind of funny though say well you as a citizen I want you to do but it's really not good for my case if you tape it which is true. And we found this exact thing happened. We started audio videotaping just about every interview we do and it was just the greatest thing ever since we started doing it because the amount of time we spend in court arguing over whether Miranda was given are not exactly 0 now and it was in every preliminary ever did ever was the allegation that was a given or it was lied to or there was a promise or there is a threat. Now just as in, because yes roll the tape and you can see would happen. And in the very rare case where there is a problem you know, I’ve never seen threats or anything like that. But I have seen maybe there's a discussion beforehand with a police officer and usually unintentionally maybe they say something that seems like it's a promise but we just all know what it is and it's all there. So becomes transparent. Just makes our lives a lot easier. And so we've started videotaping all lineups.
I viewed one videotape of a lineup but of a show up and it transforms what would've been a useless identification and to the best piece of evidence because what happened is if you just read the report it would have said: broad victim to look at lineup. She said she couldn't identify anyone. She said she couldn't identify anyone. Then I told her to try and she identified the suspect. Now that would've been the worst identification the world. We had a videotape which was in the squad car. He drives her there in the videotape is focused on her in the passenger seat. And it's almost funny because the cop knows he's being videotaped but he could care less. It was clear he could care less about this case which was not great in fact for the image that we wanted to show. But it was really good like he wouldn't feed her anything. So he was like can you see anyone? And she was like this (gestures) she's not looking. She says no I can recognize anyone. Well you can see she's not looking at all because she's terrified. And then he's like walking you see anyone, kind of irritated. She says again. And then he turns to her and he says while you got a look. Can you identify anyone? And she looks up and she starts, I mean her face goes white. She begins to shake and she says that's him. And you look at this and you're like, oh my god that's him and I want to do something about it. So these videotaping transformed what would be bad evidence and to really good evidence. And if it's bad evidence we can see it then we all know it's bad evidence. So I really say it's a win-win and I had a when possible. We always put this in all of our things is always going to be some reason if you can't do it. Now fortunately I just think we have great police departments in our County. And so if you tell the officer's this is what you should do and this is what's better than us at they're going to do unless they're really a reason to do it.
Now the one thing I did put into the policy though is if you don't do it to either get the permission from your commanding officer on duty and you have to put a note into your report why you didn't do it so the DA can call your Lieut. and ask. And lo and behold that becomes a whole lot more of a pain than actually doing it so we have very, very high compliance. Mainly because if you tell folks hey this is a better way to do it you get by and do it.
Very briefly our history on it: by 2002 we had consistent policies of the department concerning fillers, admonitions, and confidence statements. That was pretty much already happening. Then from 2002 we switched double-blind sequential. So we've been doing this for over a decade now. And then what we added in 2011 is we kept everything we had before but we just added the videotaping.
So what have our results but? We have nearly 100% compliance. I don't want to say never. Usually when there's a noncompliance either there's very good reason for it. You know someone refused to be videotaped. Or just have an error in judgment. So for example we found some of the once a regarding double-blind but detective would remain in the room while it was happening. I don't think that invalidated the entire result but will you can see it wasn't a great idea. So all that took was, especially now that someone videoing saying hey not a good idea what you leave the room. The costs have been negligible. I want to say zero but especially double-blind that was very controversial when we started. Law enforcement was really concerned that they were going to be able to do it they were required at their going to require a lot more effort. We came with all different things. There is this folder method so that she could actually have the detective do it. Bottom line we never really needed to do anything and we been doing it over a decade. And again I don't know if law enforcement, how much is in the room but we haven't had a single request for additional funds to do this in over a decade. What you got a no that that means it really is not costing anything. And the reason is because you know being a police officer is a professional job. You're not like just producing widgets on a factory line. So almost by definition the detective who is working the case is the most busy person in the department at that moment so if you give him or her a moment to not run the lineup they're not going to shoot a few rounds of golf. They're doing something else related to that case now. But similarly that means that you're going to have moments where before you go on patrol or you're working some paper Kasich you don't have to do it just the second. So there just seems to be even in our smaller departments always so many complain.
We thought about training non-sworn in officers to do it but again it just has not proven necessary. Because of these are just doesn't take a lot of time.
Now we are nonacademic institution but we have had no discernible drop in ID weights. What I mean by this is according to all the states that I've read you probably wouldn't notice it because all these are marginal, marginal changes just trying to reduce the risks. But we have tried to adopt the philosophy of like a if we have an issue were trying to improve let's try something with the best evidence we have. My DA's expression is always give it six months and then see what happens. So you know if after six months or while detectives in the field are saying hey this isn't working and here are the reasons or I'm just not getting good ideas we would've stopped doing it. But we haven't. We've been doing it over a decade. Nobody is noticing a reduction in overall drop.
What about trials? We've never had any evidence excluded of and ID since adopting these new protocols. We don't really have challenges to them anymore. I mean what he really been a challenge? And we've greatly reduce the use of experts. I room are specially at the beginning I would get called because you know they're going to call a defense expert in on an eyewitness of unification and usually this would be at a pretrial will liken immediately pretrial hearing to see what evidence is admissible. So they bring their expert on and I'd say why don't you make one of your questions to that expert: if you're going to design a lineup protocol how would you do it? And almost invariably they were like oh I do it exactly how you guys do it. Well no defense attorney wants their own expert saying this is the best method you could use. And it's great for us because even when they do use, defense attorneys do use defense experts and there are legitimate reasons for an expert still to testify about these estimator values. There's nothing on my screen here but that's okay. So they'll talk about the things that I talked about before oh the lighting was poor, it was cross racial, weapon focus but then usually the follow-up is okay expert are you saying no identification is ever good? And they're always invariably, I plainsman expert, he'll say oh no sometimes they can be good. Well are there things you can do to make them better? Yes there are. Can you list them question written invariably they're just checking offer protocol. So it's a real win-win for us and what I mean answer prosecutors it's good, for police officers. We’re spending way less time coming into court being second guessed. Since I've gone double-blind just eliminated the sort of attacks on their integrity which I think most, I'm speaking out of turn here but I don't know, I think most cops will say they don't care I think you work hard. You risk yourself for the public and you get on the stand and somebody is trying to make you a liar, you framed someone. I think you care. And it just hasn’t happened. It doesn't happen for Miranda anymore, doesn't happen for lineups. And to say nothing of the fact that just makes my trial go a whole lot smoother.
Reversals: we have not had a single reversal since adopting the proposal, let me say based on eyewitness identification. We have had occasional reversals for other things. It's just become a non-issue at trial or on appeal. That doesn't mean ID issues aren't relevant sometimes you have an ID case.
So why did we choose to do this? The first is as I said my responsibility now is conviction integrity. Obviously every prosecutor has that responsibility. Every cop there's a responsibility but we wanted our office to have somebody who wasn't necessarily in the mix of day-to-day trials to have that is their responsibility. And I'm telling you part of conviction integrity, the reason we call a conviction integrity and not the innocence project is we've you integrity also I mean it means strength rate-- like structural integrity, coherence. We believe this at the great strength to all of our trials and convictions by reducing the risk of an error happening it makes people more confident in what we do. So it's strengthened our convictions and we have had five exonerations in our County. I was counting in my head last night I think we've had three or four of those five were based on that: identifications that were unintentional. I mean the witness just made a mistake. I worked on most of those. I think it was conveyed better by the officer in that clip. You don't become a prosecutor, you don't become a cop because you want that to happen so we felt really strongly: I want to strengthen our convictions that are good at July really want to reduce the risks of wrongful conviction. I say reduce the risk because I think we have to shift to a mindset that you're never going to eliminate this. This is a human system. So what can we do to make this little bit better?
So the other reason why we focus on eyewitness identification? Because it's forensically weak. It's a weak source of identification. Anecdotally I can say when I would speak to police officers I really respected you know detectives who would have done this a long time and I was surprised at first where they said oh I try never to get an identification on my case. And I said well what you mean? And they said well it's just not that evidence so if I can build my case up in every other way I'd rather never have an ID at all. I think most cops who have got experience know that. It's just hard to identify people so you really don't want to get your case on it. I can guarantee you I don't know a single prosecutor who has ever said oh I really want to try one ID case. It's just forensically weak. Yeah so you never like that. Okay so here's this thing that conforms that anecdotal evidence but that's beyond anecdotal. I think this is based on hard-won experience of police officers and prosecutors that eyewitness identification-- and I said weak I didn't say, I mean we prosecute cases with eyewitness ID all the time just because everyone in this room you've identify people I miss on we can't do it but it's something we should be cautious about. Now recently I read, I can't room if it was a meta-study, but bottom line these archival studies were they when the police departments. They didn't do ours but they wanted to others. And they just looked at all of the IDs they had done over a period of time. What they discovered was that I think comparing all the data people were successfully able to identify someone. They made an identification about 70% of the cases. In the 30% that didn't is presumably because conditions were such that they can identify someone. 70% of the cases they identified someone and about one third of that 70% they picked out fillers. This is just an archival looking at the lineups that were already done. And I think this is like eight studies, erroneous numbers are rough, that have been found the same thing. This comports with trial experience and most cops that I know, their life experience. But what that tells me is eyewitness identification is just kind of week so if there's something that I want to do with that it's I want people to be more conservative in choosing someone. I guess that basically means in one third of the time you're getting a pic archival he they're just guessing. They have to be guessing because they picked the filler.
So I don't know if I'm making this point well but it's not like eyewitness identification is the only thing we look at. We have lots of types of evidence. But now we know and I think this is covered here earlier, we know that eyewitness identification is closely correlated of with wrongful conviction. So if there's a wrongful conviction, which does not happen a lot, but if it does happen eyewitness identification is one of the leading causes. Number two it's not like I'm looking at eyewitness identification-- let's say this archival study data have come out differently and they said oh no, no, no you show people a lineup they're only making a choice about 30% of the time. Meaning they're pretty careful not to guess. But when they make a choice were finding that 90% of the time they're not picking the filler there picking actual suspect. Well that would lead me to think that people are pretty conservative naturally in picking someone so maybe I don't encourage it. I look at this and I think wow this is just not good data. This is not really solid evidence. When Prof. Ross is talking before I is a policy matter I would rather shifted so folks are a little bit more conservative when they pick an ID. Now that's not just to protect the innocent although I hasten that, of course that's the prime motivation but the editors if somebody picks the wrong person it really messes up any further investigation. Because now, I don't care if you found the actual suspect on video, I can guarantee you if they picked up the wrong person that will be front and center saying hey jury is this not reasonable doubt? They already picked out someone else. And why are they here? There's going to be all sorts of other stuff. You can never get that witness to identify anyone again even if they could. So to me again this comes down to like well given that there's a high rate of inaccuracy in our traditional method of doing it I would rather industries the accuracy, make people more cautious. I'd rather than just say they don't know then have sort of unreliable data. And then there's a prosecutor I guess is a more personal note that obviously I have been doing this so long, I think cops are a lot better at this than I am, I honestly don't know when people are lying to me. I have come to that conclusion because some people are just really liars and some people are just really nervous when talking to me. I may not seem very authoritative but in circumstances that can be. So I just don't know when people are lying. I did a while in our drug treatment court and people would swear on a stack of Bibles that they hadn't used and I realized like I might as well flip a coin between either on it I don't think they were using and they do a confirmation test and they had. And I had someone else who's telling on stories and sweating and swearing they didn't use lo and behold they hadn't used. And that's when I realized I can't really tell. And when I'm looking at that too is I'm really frightened of evidence that I feel like I have very little ability to tell if it's truthful or not. And so if I get an ID I just rather they said I don't know then say their sure and I can't really tell if they're sure because there sure if there sure because they made a mistake. Much less, sometimes they’re sure because they're lying, but you know.
So I think I've gone over that that I just want to be more conservative and more accurate in our results. And so why do we do it? I do think the data clearly shows, I guess some folks think there is a debate between sequential or simultaneous. To be honest what I would say to that is first of all folks should just with the data into they want to do. Second there's a lot of different things you can do. If you want to start off on all of these other like the court for I think that's a great idea. I would say this though, eyewitness identification is not the only protocol that we implement and in general... So when it comes to this I want to be consistent with how we judge other things that we implement. And in general my view is hey we want to be very aggressive and very innovative as it were when it comes to how we investigate cases and all manner of other stuff. It is the prosecutor when it comes to evidence that I present as a jury I want to be very conservative. And so when we're talking about science we always try to take the most conservative approach that the most scientists would say that it's kind of like the most common not the kind of more outlier result. That's usually the defense expert who's got the one study and we've got like 50. So on the whole I try to be conservative and if I'm going to error on a side I'm going to err on the side of being more cautious and not putting the evidence in then being unless cautious, because it's not a civil case, right? And I am not the defense attorney I want to be really cautious.
The second thing is we have a lot of other forensic evidence that we use is we want…you know we have a lot of other forensic evidence that we use whether it's fingerprints or arson. I'm sure a lot of you have heard about abusive head trauma that sometimes called shaken baby syndrome. If I were going to wait until there was 100% consensus from every scientist we just wouldn't introduce any of that evidence but we do. Because we just try to do our best. We just choose what we think is the best evidence for the time. And if we are wrong you know will change it. That's kind of like that same six-month role. I think this allows us to be more aggressive in our prosecutions because when we have ID evidence it just goes right in. We just have not had problems with it. It makes are convictions a lot more stable than they would be otherwise. And in public affairs, and I mean this in the most positive sense, I didn't put public relations just like to people like you a lot? I mean engaging in the politics of the community. At least in Santa Clara County in California the people who pay me, basically hired me in my colleagues to kind of look out for them when a crime has been committed. They expect us to do the best we can to adopt policies that reduce the risk of a mistake. And all I can say is this has appeared to have been very comforting and seems to have met with a lot of approval from our community.
Experience: no significant costs, very high compliance, no detectable drop in identifications. We believe with an increase in accuracy we have a rapid reduction in the use of experts at trial which means our trials are faster our police officers spend more time on the street, less time in court. We have reduced the time of trials and litigation and we've not had a single reversal. So experience is been pretty positive. So like I said I'm truly not coming in to tell anyone what to do but we have had a very positive experience with it. I think I left my contact information so if this is something that you are interested in please fill free to contact me. Alright that's it.
1:30:00

Paula Wissel

So, now moving on to our roundtable discussion. We have some people who have been up already and some who haven't. We have, I'll wait till they get seated and introduce them. Okay we have detective Nathan Janes case the Seattle Police Department's violent crimes training coordinator. We also have of course David Angel again and you know him. And we have Travis Stearns who is the deputy director of the Washington defenders Association and Dr. Ross. and then Mark Larsen who is the prosecutor with the King County prosecutor's office the chief deputy in the criminal division. And then we have Lara Zarowsky again. We are going to start by having them speak a few minutes to the question: how should Washington state move forward to improve the quality of eyewitness evidence and will start with you Nathan.

Nathan Janes

My name is Nathan Janes I'm with the Seattle police homicide unit as well as a training coordinator. We need to develop comprehensive protocols which we don't have currently that we have been meeting on but we need to have copies of protocols. With that we need to have truly comprehensive training on that. I mean education, you've got to tell the cops why you want us to do something. Give us the background as to why. Convince us that it's a good idea and you give us the science were going to go okay that sounds good to us. If you just come up and say these are the protocols for you, have fun, the cops criminal I don't think so, okay? For comprehensive, the bigger picture is not just: these are procedures for photomontage identification, for live lineups. This is how to do street show ups. But a bigger picture of where does it fit into your investigation? Why are you doing them in this point in your investigation? Should you be doing them at all? You have DNA. You have the suspect on video, right? Has he confessed to all of his friends and neighbors and you? Then why are you doing this at all? Why are you doing the photo lineup that all? Versus that's the only thing you have to do. You have nothing but a live eyewitness identification: the witness saying that's the person or not. I would be concerned if that was the only evidence I had against someone. I don't think we should be going forward if that's the only evidence against someone. So you need to look at the big picture of how the whole investigation is going and when do you do these things. It's easy right now to do a photo lineup. Let's get the pictures, get them organized and show it. So sometimes that's the first thing the detectives to but it doesn't mean it should be the first thing they do. They need to understand how it fits into the whole big scheme of things. Make sense? That's what I have.
1:33:47

Paula Wissel

So you see there is no procedure in place with the Seattle Police Department?

Nathan Janes

We have a lot of protocols but there is no comprehensive protocols in Washington State right now.

Paula Wissel

David.

David Angel

You know, since folks just heard me talk when I pass and folks who haven't spoken as much can weigh in.

Paula Wissel

Okay let's move on to Travis Stearn.

Travis Stearn

Sure, so let me just start by saying that I think that one of the things that needs to be done and it's really great that were doing this today is that we all need more training. There is a great report from the Veritas Project from the Santa Clara innocence project that talks about the need for training in many, many areas and this has to be one of them. And it's really great especially that we are all sitting here some law enforcement and prosecutors and some of us defenders I think actually talking about the same thing. So that really needs to be the first step. For us in our state I think the next step needs to be to look at those uniform procedures. I was reading as part of my preparation for this the national survey of eyewitness identification procedures in law enforcement which is the DOJ report prepared by the police executive research forum and they said that the 1999 recommendations which included developing policies, providing standardized instructions, one suspect per lineup, minimum of five fillers, standardized photographs lineup members, coordinating witness statements and preparing lineup reports still have not been adopted by many police agencies. And it's kind of surprising that when we are talking about putting in the big four and where we should be getting to a place like Santa Clara there were not even there yet from the 1999 protocols. And we really should be. I don't think that there is any excuse for that. What that report summarizes in its executive summary is that it is critically important that law enforcement agencies review their eyewitness identification procedures, practices and trainings to ensure that they are in line with the 1999 NIJ guide. So get there is a minimum. Those of us who are in the court rooms and who work either in trial courts or in appellate courts need to start I think should embrace what Justice Wiggins said in state versus Allen. I know you don't have state versus Allen or we didn't hear about that this morning when we talked about the case law. Tut Washington looked at it. Washington did not do the comprehensive report that the other states have done but Justice Wiggins in his dissent in that case where they don't say that an identification procedure was appropriate in that case says, "the most important lesson of this case is that every member of this court would support giving a cross racial identification and instruction in an appropriate case but that we differ on what constitutes an appropriate case." It's really important to bear that in mind that our court is also telling us that we need to analyze this. And those of us who are lawyers and who are analyzing them after they have come out of the police departments need to work on that. We need to come up with those instructions and to embrace the use of experts and so we have the protocols and until we're at a place where none of our prosecutions get challenged like they do in Santa Clara. Which are really great for defenders but really great for all of us. And finally we need to think about postconviction reform. It starts solid conviction integrity units. It's great when you hear about them happening. I know that Santa Clara is one of them, San Bernardino, Dallas Texas, Brooklyn. Manhattan has just I think adopted a unit not so long ago. And that needs to be happening everywhere. But what needs to be happening with those is that it needs to be the solid partnership that we started with just like training. We all need to be working together on postconviction relief. We have a forensics commission in this state that includes after a big fight the inclusion of defenders on it. And that needs to be happening in all of this reform. It needs to be a joint project that we are all working on together. Prosecutors need to embrace that their convictions can be, and again, I go back to the beginning of the Veritas report. One of the problems is that prosecutors sometimes don't want to believe that they were wrong just like in the picking cotton case with Jennifer. And that is a change to. We need to be working on that and established these changes. So I think that's 3 min.

Paula Wissel

Dr. Ross had a move forward in Washington?

Dr. Ross

I think there's two things that I want to say for this one that's kind of that the meta-level, more of an abstract discussion of where science fits into this and then limits more specific. As far as it goes with the first of where science fits into this I want to just clarify and make sure that we view this as science can inform practice but science should not and frankly we cannot say to do we can provide you with information on what affects these have on witness decisions but just because this has an effect there are many other facets that need to be considered on deciding what to inform, excuse me what to reform and how exactly do it in practice. So I just want to clarify that and make sure that we don't just look to science for the answer. It can inform the decision but it cannot make the decision for us. The second thing and I think it's something that is dramatically overlooked in almost all of the reforms that are out there is perhaps the greatest way that we can improve the quality of the eyewitness evidence we collect is to focus on interviewing of witnesses not identification procedures. There are dramatically better ways to interview witnesses that have been shown consistently, reliably not just here in the states but throughout the world that will improve the amount of quality evidence that you can get from witnesses. It's often overlooked. It requires training but it can be done and it's a good way that we can look at here because that's not just going to benefit the ID case. It will benefit every single case in which you have to collect information from a compliance individual. And I think that's an area that we shouldn't overlook as we are focusing on these issues. Today was all about identification but there is a lot more to it than just the identification procedure.

Paula Wissel

Mark Larsen you are with the King County prosecutor's office how do you think we move forward in this area?

Mark Larsen

Thank you and I think we are. I think there is good news here. I think there is a lot of momentum right now to do this. I actually worked on the NIJ technical working group back in 1997 it was convened and wrote the guide that came out of that as sort of a beginning point of a lot of this discussion. At the time we dated there was a very large group, I was actually part of the preplanning group. Even at the time we did I'm rethinking law enforcement is an incredibly, this isn't a bad thing, but it's very parochial in the way it is administered in the United States. It's not like England. It's not like some other places where we have national police forces. It's parochial down to Milton, Seattle, to every small berg, town, Hamlet. So effectuating change in practices and procedures through the publication of a book out of the Beltway you just never struck me as the thing that was going to strike gold and make this change. And in the PERF study that you mentioned earlier Travis I think was sadly not that surprising when I read it. And I thought well 15 years later, gee that book didn't revolutionize practice in the thousands and thousands of police departments that deal with this. So I think we still have a challenge ahead of us. What we have seen around the country though is that if you don't embrace this legislators and courts I think I left to fill that gap. And I think they have done it poorly and places. They have done it well in places. But when you see the two other people you're not always going to get exactly what might be the most workable. So in Santa Clara they have gone ahead of this and have been able to adapt some of these reforms. In King County we have a model policy. It isn't trained to though and it doesn't have a lot of the things that could really make it seek all the way down to all levels of law enforcement. Currently there is an effort involving the innocence project. Washington Association of prosecuting attorneys of which I am a representative, WASPC which is the Washington Association of police Chiefs and sheriffs and who am I missing? Seattle police is at the table. But we were sitting down right now and I think there is a growing consensus that we need to grab a hold of this. We want to bring these reforms to every police department, every show up, every line up. What you do I think is something that we are still coming to our consensus on. I would say this though that the court for, I am good with all of those. I think those are things that we can do. I am very, very concerned though about a lot of these sort of what I think has become some of the noise around these various procedures in one aspect or another the amount of conflict and debate around the sequential or simultaneous to me has been unfortunate in a lot of respects because it is massed other reforms that people do easily agree on which really should roll forward. So I am hopeful that we can sort of find at least those things where there is clear consensus and make sure that is and everywhere in every case that every time. The final thing I want to say is about prosecutors. With all due respect Jackie if she's still here McMurtrie? Jackie I love the work that you guys do I have enormous respect for it. It's important to continue. But I will say that any good thoughtful working prosecutor every single day has a chance to exonerate people who might otherwise be wrongfully convicted by accurately and fairly evaluating the evidence that they have before them when they make the decision in the first place to charge or to not charge. So that is something that I talk a lot about in the institution I work in. That is our function. It is apart from the police function. We have an independent role not just advocate for cases but to also make sure that we are convinced not just that somebody pointed at the picture but what's the strength of that evidence? How do we understand that in the context, Nate as you say of all the rest of it. How do we make those determinations? And I think in the final instance it is to be conservative in those cases where there may not be corroboration, those rare that terrify US the prosecutor where all you have is somebody saying that's the guy with the kind of shaking finger and there's nothing else, those cases should put fear into good, honest prosecutors as we decide whether we should or should not charge those and I think that's increasingly part of our culture and what we want to sort of talk about in practice. So the final thing I want to say that is this: our history with experts in the social science has been built almost exclusively on cross-examining people in the courtroom. That's not a great way to get to know people. That's not a great way to kind of just appreciate the things that they can bring to bear in the things that we encounter like eyewitness evidence and the science that surrounds it. So I am really, really pleased that we are trying to build some different relationships of people like Dr. Ross and others to say hey we all want to meet you on a witness stand. We would really like to have you inform us, talk to us, help us understand how we do our jobs well. How we evaluate the strength of eyewitness evidence in the cases that are coming to us. Putting aside a cookbook of these certain procedures, you can do all of those procedures absolutely correctly and you'll still get errors. You can do them all wrong and people will still be right in terms of picking out the right person. So procedures I think we get lost but if we do it the right way, right? That's going to be yet will have number errors. That's not really it. Beyond that it's important to do it the right way and the best way but it doesn't solve the fundamental problem which is we are going to still have people pointing at a picture but are going to be wrong. How do we understand when and how that happens? What circumstances promote that? How do we evaluated as prosecutors? I am excited about the idea of having a new partnership with social science where we can really take advantage of that apart from just cross-examining people.

Paula Wissel

Thank you, Lara?

Low Zarowsky

Thank you so much in this is the unenviable position at the end of the panel where everyone has said all of the intelligent and enlightening things to say. I would like to pick up on a couple of themes that I am hearing here that I support 100% in terms of how Washington state can move forward to improve the quality of evidence that we have in our courts. Number one appears to be that I'm hearing a little bit from everyone is training. We need defense attorneys, prosecutors and especially law enforcement to be trained in the science behind these procedures and behind eyewitness memory and general just so that we can understand a little bit more. So that those other practitioners gathering this evidence in dealing with that in court have a greater understanding of how reliable or not reliable that evidence actually is. I am very heartened by the progress we have been able to make recently. A few of us on this panel have been working together closely for about three years now trying to sort of figure out how to move forward and build a coalition. So to the extent that this work is happening with the prosecutors Association and with the sheriffs and police chiefs and our smaller working group I would really like to see those efforts coordinated so that were having the same conversation and contrived to move forward as efficiently as possible particularly given that there seems to be so much support for these core four reforms that have been mentioned throughout. There are a few that have sort of some controversy attached to them but to the extent that we can move forward with that which is not controversial but could make a tremendous difference in the quality of the evidence. I think we really have to do that. And the final thing that I would say is if you are looking for the Veritas initiative report that Travis was talking about earlier do not look for Veritas project because that has something to do with aliens. I'm not talking about foreign nationals I mean extraterrestrials. So that's just a final note. Thank you.

Paula Wissel

Use a different search term.
So now we're opening it up to you and Jackie McMurtrie has the microphone there.

Question 1:47:53

Thank you. Now we all know that not all exonerations can come through DNA evidence. There is a lot of other types of evidence. We also know that at times some of the best investigative resources are with the police departments. So my question is this: I am a practicing attorney and I have to deal with postconviction time bars and other concerns where it is not always possible to go into the court system or the burden of evidence is very high. Now I heard about Santa Clara counties conviction integrity unit so my question is and I guess this would first be to Mr. Larsen but in general it is not possible or not a good idea whether through the attorney general's office or the particular counties to consider having conviction integrity unit's to locate cases that may not be addressable through the court system but that upon further examination would show that the individual’s at least entitled to a new trial? Thank you.

Paula Wissel

Mr. Larsen

Mr. Larsen

What I would say Mike we have done that in an ad hoc way at times and we don't have a set unit to do that. I look at my friend Jackie go to and say you know I think attitudinally we have worked very, very hard to try and keep an open mind and try to work with the claims in cases that they have brought to us and try to analyze those answered open up files in our minds to look at cases where there may be claims of innocence. So I think attitudinally we try to be in the right place to do that but we don't have a dedicated resource and maybe that is something we should look at.

Jackie McMurtrie

And I second that. I agree Mark. Your office has been very cooperative with the cases that we bring to your office.

Paula Wissel

Another question?

Question 1:49:50

I have a two part question. You have discussed a lot of losing forward. My first question is why was so far behind? Second relates to the fact that within the past month in the case that I am part of we had Dr. Reece Berg testify in a case and Dr. Reece Berg was the expert and Lawson out of Oregon that here in Seattle he witnessed the worst lineup identification he has ever seen in any case at any time. What do we have to do to get some training right now for the people who are conducting these lineups and identification procedures even if we can't get uniformity across other agencies?

Paula Wissel

Detective Janes, since you are with the Seattle Police Department.

Detective Janes

Right, um, well we have been working on protocols that were going to be for our department and then recommend them for the state as a whole. That's what we have been working on. I don't know who did the lineup you're talking about. I can tell you that I run our detective school and I have done that for several years and everybody has been trained in pretty much what we have been talking about when I have trained them. I don't know, I haven't trained all of the detectives in our unit and I don't train much outside of our department right now, so I don't know who is trained or if they were trained. It's pretty easy to get trained.

Paula Wissel

Doesn't seem like it has been a priority? Has it been a priority in the Seattle Police Department?

Detective Janes

No. But with Chief Pugel that is changing though. So it has become more of a priority now.

Paula Wissel

Is it more of a priority with the Department of Justice be involved?

Detective Janes

No, no they are not talking about this at all not at the DOJ.

David Angel

I don't want to jump in front of the moving bullet but I know a little bit about that case, not a lot. I think what I want to say though is that my experiences that ID procedures, there are model policies in King County, we've adapted, the Chiefs have adapted. But having applicability down, you know the chief spoke earlier talked about a policy, forms and then training behind it. It really takes a coordinated effort to sort of get that to penetrate all the way through department because the truth is most police officers collect this type of evidence. It's not like CSI where a small group of people can be trained to do it. The other thing that happens, I’ll say, and this is a conversation I've had with so many police officers they will tell you what I have never had my evidence suppressed. I have never lost a 36. I must be doing a fine. And my point is to say you know that's not really the standard. That something wasn't impermissibly suggestive shouldn't be the highest standard we are sort of shooting for as we gather the evidence in the first instance. So I think there can be complacency: I think we’re doing fine, I'm not getting any feedback that were losing cases or this or that, the judges throwing it out. So I think that creates a false sense of things must be fine. So creating urgency around this can be difficult. But I think that's over trying to do. There is an urgency even afford late to the game. But it's not that nothing has happened either so that's my impression of the certain culture that surrounds some of this.

Paula Wissel

And I just had a question David Angel what prompted Santa Clara to be so out ahead of everybody if they are?

David Angel

Well let me just say kind of in response to that question also maybe in front of a bullet in some ways see were talking about this earlier it's kind of important to segregate what happened before from how you want to move forward. What happened before, usually have to evaluate pursuant to meet the constitutional minimum, and if it didn't then the case gets reversed if it did, then it's okay. But you don't want to set your standard…your goal is not to achieve the constitutional minimum, right? Your goal is to achieve the best you can. For us, conviction integrity was just really central to our current DA Rosen so when he ran it was pretty central to his campaign but it had in various guises kind of like you're saying it was a function that was happening. I think this if you added value by taking that function and you can really leverage it if it's in one area. And that is something actually I have been asked to present to the California District attorneys Association in January. So I think there is advantages to doing it that way but it's not like…usually these are things that are happening anyhow. As for the eyewitness identification the real genesis for that was some ill-advised stupid things I said to my wife about a decade ago. What had happened was that she is currently representing the sheriff. She works for the city attorney. But at the time she was a public defender and she was working on something and she said, have you read this study from the DOJ about eyewitness identification? Do you do that in Santa Clara? And what with this being my wife and said yeah of course we do. And then I asked if I could borrow it and I said well we mainly do a lot of the time. And then that kind of started us going but I was really lucky because our DA, you know I then drafted up a report based on that and looked at some of the research and we have had for a while a very good relationship with law enforcement so we went to them and said hey I think this would be a good idea to do it. They were open to trying. We have always had this fairly experimental view of well let's try something and see if it runs and if it doesn't work in six months, quit doing it. So that allows us to be able to be a little more flexible because of make a mistake we can just stop doing it and try something else.

Paula Wissel

Is there another question?

Question 1:55:42

I don't know if I can state this the right way but I guess I am concerned that their cases that are valid when there is only evidence is an eyewitness identification. I am also concerned that sometimes people get it wrong. So I am just wondering if part of the equation is that jurors overvalue this evidence and that maybe the social science of the science would teach us something that we can inform the jurors of to sort of counterbalance the fact that this may be the only evidence. Hopefully jurors kind of evaluate that evidence.

David Angel

Well if I may just for a second adjustment the quick statement on that. I couldn't agree with you more because sometimes I have been at forums were like oh well you should never do a case and if it's just eyewitness identification. And I say yeah well we used to have that the law in California and it went this way: if a white man assaulted a Native American that evidence couldn't come in. Nothing that Native American could say because you always needed him unless someone else. It was the same thing for domestic violence. And you go down the ugliest parts of our history and we always said oh you certain people need corroboration. So I never want a situation where all you need to do to get away with crime is make sure there is no one else seeing you while you do it. And we have jury instructions that specifically say if you believe a witness is credible that is evidence alone. It is very rare that you are mandated to have cooperation. And that is why this stuff isn't simple. If it was simple enough well we have said this in a number of guises hey you just follow this checklist and it will all be fine you know I like to think we would've done it already. The reality is it's not that simple and you cannot just toss out eyewitness identification, but you want to train people to be mindful of his limitations as well.
1:57:37

Dr. Stephen J. Ross

Let me address one thing I'm not from the social science perspective. Do jurors overvalue witnesses? Without a doubt, hands down, they overvalue witnesses and they primarily rely on witness confidence which we have demonstrated earlier is not what you should be relying on. So what can we do about that? We can give jury instructions which some states have done. That was the argument in state versus Alan. From a scientific perspective I don't use jury instructions as an incredibly positive thing because the research that is out there on them does not demonstrate that jury instructions as they are implemented I received very well by jurors and they know how to apply them. Jury instructions in order for jurors to apply them at least what the research is suggesting is that they need to be more specific to any individual case. So blanket pattern instructions are not necessarily going to serve the purpose that they are intended to. And my concern on this is how it will be viewed as the fix when it's not fix. I view it from a scientific perspective that the best way to prevent an inaccurate identification from leading to a wrongful conviction is to prevent inaccurate identification. Focusing on these interviewing techniques these procedures that are used in the collection of eyewitness evidence may be the best way to go about it. Now that doesn't mean that there's nothing we can do post collection and at the trial level but the most effective ways that have been demonstrated so far and the research is by doing education in the trial through the course of expert testimony-- getting expert similar to testify about these factors that are what is termed estimator variables, these things that aren't within the control of law enforcement: the stress, the weapon focus that has been alluded to earlier. In educating the juror on the specifics and giving them an idea how they can apply this knowledge to that specific cases more effective than just giving a pattern jury instruction.

Paula Wissel

let's take another question.

Question 1:59:34

Detective Janes you mentioned that to persuade police officers to do this you would have to persuade them that that it's going to make a difference, that it is the right thing to do. So my first question for you is are you persuaded by the evidence that was presented here today? And if so who needs to make this happen? Does the prosecutor's office play any role in making this happen within the police department?

Detective Janes

First of all I have been meeting with Lara and Prof. Ross for years and so I was convinced long before we had this today, personally. And yes the prosecutor's office should definitely be involved with it and as we've been meeting with Mark as well. I don't see the reason not to move forward and rather quickly as long as we get the comprehensive protocols in place that we already have drafts of. So it shouldn't take too long from a practical standpoint get these protocols in place at least for SPD and then move there and then try to get them standard for the state of Washington. Does that answer your question?

Question 2:00:47

Well it does except except for hearing Mark say he's been working on this since 1997 and Travis saying these reports of been out since 1999. What's been happening for 15 years?

Detective Janes

What's been happening for 15 years?. I read the first NIJ study and have a copy of it still. And when I read it I said okay I'll start doing it that way. My Department said okay if you want to, fine. That was about it. It was a low priority. Actually the way I first met Lara was reading in the University of Washington alumni magazine because I'm an alumni, reading about Jackie McMurtrie and the innocence project and how at that time it I think it was seven cases in Washington state that had been found to be false convictions. And I thought there were none in Washington state except for the Wenatchee thing from the late 90s. So then I contacted Jackie McMurtrie to find out what happened in those cases and I wanted to know what went wrong. And that's how we started meeting and talking. I think the vast majority of law enforcement in Washington state thinks that those things that have gone wrong it's mostly in the South that's gone wrong. Because in the South it's Billy Joe Jim Bob wants to be a cop give them a gun. And there's not much training. That is the perception which is wrong but that's the perception. So they thought nothing was going wrong in Washington state so we are good so whatever. That's not the case but it is still getting that information out to all of the various agencies in Washington state is a problem. Frankly we have a problem of we are SPD-- we're the big boys on the block, right? And the little departments are like I don't want the big boys on the block telling me how to do my job. So it needs to be at the state level as well as that were working on it right now.

Lara Zarowsky

And I just want to say I think it is fair to say that there have been many conversations about how to move forward for many years and I know that Mark has been involved in a lot of those. We had a lot of preliminary conversations and for quite a while the conversations we have been having have been very behind closed doors just trying to figure out what is the most efficient and best way to move forward? And I think that honestly my opinion is that probably the most effective thing that we could possibly do right now is to try to find a way to do basic law enforcement training on these core reforms that would been talking about and then on its own could make a tremendous difference. But the goal ultimately absolutely should be some sort of standardized statewide protocol that we can all get behind and so hopefully that's where were headed with the momentum the time now.
2:03:23
Travis Stearns:There are a lot of departments and King County that do this extremely well and they have great protocols and procedures I just got it yesterday, Mercer Island's, believe it or not. It's a model policy and they have I think trained to it. So my point isn't that everybody is feeling but that it's uneven and so parochial and within department large ones that can be parochial: I got trained by this guy who showed me how to do it. So it's really bring I think some kind of clarity to that in the mandate to sort of say were going to do the same way across the state. That's important but it's also not easy to accomplish either in this environment. We are working with WASPC. I think that's our most important ally in this (inaudible).

Paula Wissel

So it's not legislative. It doesn't have to be to the legislature?

Travis Stearns

I don't think that that would be the most effective way to proceed. But they could do it in other states have legislated and other courts have certainly gotten involved as well and I think that's probably less effective.

Question 2:04:13

What has been the reception of WASPC to the protocols?

Travis Stearns

So far pretty favorably. You know this fall we have been working with them and bring in Tom McBride from WAPA has Spencer getting those folks on board. We have met with them. I had a CLE that I did with them about a month and a half ago so it's been pretty good.

Question 2:04:30

(inaudible)

Travis Stearns

Sue Rohr is now the chief of the training center she has joined a group. And actually we were talking to her about this really critical part: couldn't we get part of basic training to beat how to do interviews better? Put aside even just eyewitness cases. Other techniques that allow you to get more and richer and better data from the people that have experienced and witnessed the crimes? Yes, and right now that has not been made a priority in law enforcement curriculum. So I think that's something that I am very interested in. I want more and better data about all kinds of cases and I think that is something that can be improved. And she seemed interested in that.

Paula Wissel

We maybe have time for one more question.

Question 2:05:24

Thanks I think we all can agree how important this witness identification is. I have personally witnessed what happens when a witness identification goes wrong early in the investigation because it starts a train that is not only bad from the standpoint of integrity but it becomes extremely expensive when it results in a wrongful arrest or it gets through the process. And education is great but once his police officers are trained it doesn't seem like there's any checks and balances to detect when it's gone wrong. For example a witness might be part of this and feel that they were not or the process didn't work well but yet the witness feels they have no recourse or would they go to tell somebody? So I think that if we could get a checks and balances early in this process with the people involved it might stop this train from happening. And my question is: are there any moves to get some type of reporting system where this could be stopped earlier in the process?

Paula Wissel

Does anyone want to take that?

David Angel

Well I can just say from our experience initially when you are doing something in the criminal justice system you have very few advantages because it is this adversarial kind of creaking system that one area where we have a lot of advantages is early on somebody like a reporter says well how much compliance do you have? And I said well I think it's really high. They said what why do you think that? And I said well because a public defender would tell me in every single case if we have adhered to our own protocol. So most things when you do it are not subject to cross-examination by a paid trained attorney on the other side who is trying to find out if you didn't do something you are supposed to. And then especially if you add the component, I can't remember if this is part of the core for or not, but if you videotape it as well. But even without that if you have cross-examination and you have represented defendants you find out pretty quickly if you're not doing what you said you're going to do. I am sure that would be experienced here don’t you think?
2:07:34

Mark Larsen

You know I think that a big step of it to though is prosecutors are going to say no. I was talking to an out of County prosecutor couple of days ago I'm trying to prepare for this and he was talking about one of his most recent cases where he had had a detective who brought in his array and put his finger on top of one. And he read his list and he said I want you to know that this person may not be here but that he may be and I want you to let me know who it is and he took his hand away. Now the witness when talking to the prosecutor said I know I pitch that person I believe that it was but I also will tell you that the detective gave assigned to me. And the case went away. Now the prosecutor also said that he didn't really believe there were other factors that made him believe that that person was not the suspect in it and it was not the most serious case in the world. But knowing that and knowing that you have a prosecutor who is going to say no when they smell something like that I think makes an enormous difference in knowing that we have prosecutors like that it also makes an enormous difference. It's a combination of the defense attorney saying no and the prosecutor also saying no when it gets to the courtroom. And then I judge being willing to embrace that and not looking forward to finishing their case in two days, being willing to say yes I'm going to allow you to put the expert testimony on and I'm going to talk to you about a particularized instruction in this case. And knowing that you have that kind of integrity within the system I think is the balance that we need. And we need all parts of that to be together.

Paula Wissel

Okay I want to thank everyone. We're out of time. Thank you.

Last updated 12/30/2013