Summary
finding proffered expert testimony not sufficiently reliable to warrant its admission under Fed. R. Evid. 702, and disagreeing with the statement in McDonald,supra, 37 Cal.3d at p. 365 (relied on by the dissent, post, at pp. 1157-1158), that the results of studies in the area of eyewitness identification factors were consistent: "With due deference to the . . . Supreme Court of California, the testimony presented at the evidentiary hearing held by this court belies such consistency." The court listed areas in which the results of studies have been inconsistent, including theories as to the rate at which memory diminishes, the assimilation factor and the confidence-accuracy ratio.
Summary of this case from People v. WrightOpinion
Crim. No. 82-00223-01.
May 23, 1985.
Peter F. Schenck, Sp. Asst. U.S. Atty., Philadelphia, Pa., for the U.S.
Brian E. Concannor, Mansfield, Mass., for John W. Downing.
MEMORANDUM OPINION AND ORDER
On August 17, 1982, John W. Downing was indicted on counts of mail fraud, 18 U.S.C. § 1341 and 1342; wire fraud, 18 U.S.C. § 1343; interstate transportation of stolen property, 18 U.S.C. § 2314; and aiding and abetting, 18 U.S.C. § 2. On October 28, 1982, after a trial before this court, the jury found Downing guilty of all but the interstate transportation of stolen property counts. On January 25, 1985, the United States Court of Appeals for the Third Circuit vacated the judgment of conviction and remanded the case to this court with instructions to conduct an evidentiary hearing with respect to the admissibility of certain expert testimony which this court had declined to admit at trial. United States of America v. Downing, 753 F.2d 1224, 1244 (3d Cir. January 25, 1985). An evidentiary hearing was held on March 25, 1985. After consideration of the testimony presented at the hearing and the briefs filed by counsel, the court declines to admit defendant's expert testimony and reinstates the judgment of conviction.
Briefly summarized, defendant's indictment arose out of a scheme to defraud numerous vendors by a group of individuals calling themselves the Universal League of Clergy (U.L.C.). U.L.C. sent representatives, identifying themselves as U.L.C. clergy, to national trade shows where they expressed interest in the product lines of various manufacturers. When a vendor took an order for his product, U.L.C. would provide the vendor with a list of phoney credit references. U.L.C. established mail drops where it would collect inquiries sent by the credit departments of manufacturers. It provided positive credit reports to the manufacturers, who then shipped the goods to U.L.C. on credit. U.L.C. disposed of the goods without making payment to the manufacturers.
Defendant and co-defendants James A. Silva and Richard Piazza were indicted as representatives of U.L.C. The government sought to connect defendant with this scheme through eyewitness identifications. It produced twelve witnesses who identified defendant as a man they knew as Reverend Claymore. The witnesses personally observed Claymore for periods of 5-45 minutes during the course of business dealings that subsequently turned out to be fraudulent. At trial, defendant sought to introduce the expert testimony of a psychologist as to the unreliability of eyewitness identifications. The court declined to admit such testimony because it felt that it was the jury's function to judge the credibility of the witnesses. However, the Third Circuit vacated the judgment of conviction and directed this court to conduct an evidentiary hearing to determine whether proffered testimony could meet the "helpfulness" standard of Fed.R.Evid. 702. Furthermore, if this court determines that the proffered testimony meets the helpfulness standard, it must then consider, pursuant to Fed.R.Evid. 403, whether the danger of unfair prejudice outweighs the probative value of the evidence.
Silva and Piazza were also convicted. Piazza has served his sentence and is presently on parole. Silva is now serving his sentence, his conviction was affirmed, and he has completed almost one-third of his sentence.
In directing this court to hold an evidentiary hearing, the Third Circuit set forth various guidelines for determining the admissibility of scientific evidence that cannot be judicially noticed. The factors that the court must consider are:
1. the soundness and reliability of the process or technique used in generating the evidence;
2. the possibility that admitting the evidence would overwhelm, confuse, or mislead the jury; and
3. the proffered connection between the scientific research or test result to be presented, and particular disputed factual issues in the case.
At 1237-1238. In formulating this approach to admissibility, the Third Circuit expressly rejected the test first propounded in Frye v. United States, 293 F. 1013 (D.C. Cir. 1923). Frye held that before scientific evidence may be admitted, the methodology used to generate it must be shown to have gained general acceptance within the relevant scientific community. Id. at 1014. With the guidelines of the Third Circuit in mind, the court must now consider the evidence presented at the hearing.
I THE EVIDENTIARY HEARING
Defendant seeks to present before the jury the testimony of Dr. Robert Buckhout, a professor of psychology at Brooklyn College. Dr. Buckhout proffered testimony concerning three phenomena of memory which he claims that research in experimental psychology has identified:
1. retention interval, which concerns the rate at which memory declines over time;
2. the assimilation factor, which concerns a witness' incorporation of information gained subsequent to an event into his/her memory of that event; and
3. the confidence-accuracy relationship, which concerns the correlation between a witness' confidence in his/her memory and the accuracy of that memory.
Dr. Buckhout also testified about a checklist he prepared which purports to identify sources of bias in photographic lineups. In response to Dr. Buckhout's testimony, the government offered the testimony of Dr. Michael McCloskey, an associate professor of psychology at Johns Hopkins University. The court will now proceed to review each area of Dr. Buckhout's testimony and the response offered by Dr. McCloskey.
At the hearing, defendant objected to the testimony of Dr. McCloskey. While acknowledging that Dr. McCloskey is an expert psychologist, defendant contended that Dr. McCloskey has not done sufficient work in the area of eyewitness identification. [Transcript of Evidentiary Hearing, March 29, 1985 ("Tr.") at 63]. The court denied defendant's objection. One reason that the Third Circuit abandoned the Frye test is that it has "allowed courts to manipulate the parameters of the relevant `scientific community' and the level of agreement needed for `general acceptance.'" At 1236. Dr. McCloskey is a cognitive psychologist, a specialty within the field of psychology which concerns perception and memory. The court is satisfied that both Dr. McCloskey and Dr. Buckhout possess adequate backgrounds which render them competent to assist this court in determining whether it would be helpful for the jury to hear expert testimony.
A) Retention Interval
According to Dr. Buckhout, the research on retention interval dates back to the early 1900's when "forgetting curves" were first published. Tr. at 11. He cited one study of face identification by Professors Shepard, Ellis and Davies published in 1982 [D-1] which found a straight line decay of memory over time up to a period of eleven months. Tr. at 11-12. Beginning with 60 percent correct identifications after seven days, the group tested was only 20 percent accurate after eleven months. When asked whether he could assist the jury in evaluating the accuracy of a witness' memory after a period of over three years, Dr. Buckhout responded as follows:
I would have to project a limited body of data. The limited body of data here means by the time they reached eleven months they would not essentially have gone further with that type of testing for faces. When we look at older studies of memory without regarding verbal memory if it is well learned, such as a language structure, memory is quite good, but verbal and face memory appear to be two different processes to simplify it as much as I can.
In effect, testing beyond eleven months has not been done yet systematically because by the time you reach eleven months if you are down to close to 10 or 20 percent recognition, there isn't much further you can go. Memory loss is substantial and further extension of that would be lower than that, but I have another point of contention about this because I think a lot of it depends on the kinds of testing and what people do when they go to those tests.
Tr. at 13-14. When asked whether it was possible to extrapolate from the eleven month period of the Shepard, Ellis and Davies study to a period of three years (the approximate time period between the witnesses transactions with Reverend Claymore and the time of trial), Dr. Buckhout stated that:
Extrapolation is, of course, mathematically possible, but the trend of the curve is quite crystal clear. Unless a lot of research changes dramatically I would not expect a chance [sic] in the performance. I would expect it to get worse. They get much worse than are [sic] at the point of eleven months.
Tr. at 16.
On cross-examination, Dr. Buckhout acknowledged that in most of the studies of eyewitness identification, the individuals tested had a period of only sixty seconds or less to view the persons they were later asked to identify. Tr. at 32. Such was the case in the previously discussed Shepard, Ellis, and Davies study. Tr. at 42. Dr. Buckhout was aware of no studies in which the individuals tested had an opportunity to view the person they were asked to identify for periods of up to 45 minutes. Tr. 32. The following exchange took place between the government's attorney and Dr. Buckhout, regarding the validity of extrapolating from studies where the viewing times were a minute or less to a situation in which the viewing times ranged from five to forty-five minutes:
Q And would it be fair to say that all of those, even the shortest of those, is in excess of the time you used in the experiments?
A Absolutely. There would be no issue if it were not for the long delay and the long length in taking information. The research in this area does not approach good research, which is familiar articles [sic] where people have had years. We have one extreme or the other.
Q So really the scientific experiments, and yours don't put you in the position where you can evaluate the kind of facts that we have in this case?
A I don't agree at all. The facts [sic] pattern falls within the parameters and have become solid over the years. There was no necessity to apply research nor every single comment or element you find in the trial.
Q You will agree, the closer you get to the trial the more accurate your extrapolations would be?
A I't [sic] very close to the fact pattern with respect to long term memory. Memory has fallen off to eleven months even though the exposure was much shorter.
When people have had significantly longer times, years of association in which the memory is reported to be quite good — between those two extremes a scientist has no problem taking variations on the parameters and offering the research for what it's worth.
Research will never fit the exact combination of factors.
. . . . .
Q Do you believe you could extrapolate and give is [sic] to the jury for what it's worth?
A When a jury is facing a unique pattern every time they face a case, knows there is a combination of things happening there and some have been testified to, the person presenting the photographic evidence in court has sworn to the reliability of the tests he is giving, that all my testimony adds is the additional information about what research shows in and around that process which the jury might have to fit together which is why their job is harder than ours.
Tr. at 34-36.
The government's attorney and Dr. Buckhout also explored the problems of applying the results of laboratory experiments to real world situations:
Dr. Buckhout:
It is difficult, but in the course of my career I have moved back and forth between reasonably good live simulations of crimes. They talk to the experts to find out what you would include in the simulations. We refer to this as ecological validy [sic] which means there is some truth to the simulation in comparison to the real world. It is not an exact duplicate any more than cancer research in this area would subject a human being to a carcinogen.
In running the simulation we find sometime we have to go back to basic research.
When I studied most recently the problem of person's focus and arousal I find I had to return to the laboratory to get better control and use a less exact simulation of a crime in order to study the process. This is a two-way street and scientists are used to it.
Q In amny [sic] experiments to determine if identification can be made aren't there many variables to be made; somebody runs it from the left and somebody from the right, and isn't that accurate? A It depends on how you stage it. I have them run down center stage.
Q Have you been where a subject was sitting at a table and talked and they were to identifhim [sic] him?
A Yes, but not the time factor.
Q What rewards are there for the study or penalty if they make accurate or inaccurate identification in these tests?
A None. They are good citizens coming forward. That's not a major issue. The movitation [sic] of people tested — it is that level of it fits between our research process and the real world, that's sufficient to make the case that this is what memory is about.
Q That's generally not true of the kind of experiments you run. Those that relate to tests; robberies, burglaries and perhaps a neighbor who witnessed the incident, a bystander?
A The range of published literature include incidents which you would describe as relatively innocous at that time. They have also found this is conducive to producing very good memories, vivid enough to last a long time.
Q I am not sure that answers my question.
In each case where you are testing is your subjects — called them students — when the students observe the subject or the person to be observed they are generally in the position of a bystander on the corner seeing a robbery or someone seeing something occur in front of him; would that be accurate?
A Yes.
Q So that there would not be the added factor of financial reward if they get to know the person well and was going to make a large sale to a person in a telling situation?
A Sorry. If I could reach into the literature for a pigeon drop or a good scam, I would make everyone happy here. The particular nature of the interaction whether long or short is not probable. To the scientist, I have conceded if you are five minutes to 45-minutes at least at the start you have all the ingredients of a good memory of a person's face.
Tr. at 37-39. Dr. Buckhout went on to state that there is a disagreement in the literature as to the effect that the nature of the interaction, e.g. stressful or casual, has on memory. Tr. at 40-41.
Dr. McCloskey, the government's expert, evaluated Dr. Buckhout's testimony with respect to retention interval as follows:
I disagree with his specific statements that we can extrapolate to make a specific judgment as to how accurate a witness would be after two years or eleven months. He took the Shephard, Ellis, Davis study and said memory was poor and could only get worse. I took him saying the accuracy should be worse after three years than the eleven months. It's a generalization. It came out in the testimony that the initial period of observation was on the order of one minute. To extrapolate with that and there was a face-to-face confrontation, and the person trying to remember the person can't be properly done.
Tr. at 67.
B) The Assimilation Factor
Dr. Buckhout described research indicating that individuals tend to incorporate into their memories of an event facts learned after the event has taken place. Tr. at 16-18. This phenomenon is referred to as "the assimilation factor." When asked how he could assist the jury with respect to the assimilation factor, Dr. Buckhout responded:
Given the facts of this case, I am not so sure I can relate it specifically to the facts of this case without looking at individual witnesses and I would be frank to say that I am in a bind over that because my practice in general is not to second-guess the evidence as it evolves.
There is indication here, for example, that there is a bit of protocology. The witnesses' descriptions from day one vary all over the place, yet all were willing to sit in front of a line up with bearded men. We have at least one instance in the case where at least the beard that is evident in the line up was not added by the witness until the time of the trial. That could have resulted from assimilating the pictures prior to trial. It is something that can be avoided by well controlled testing but in a hearing I can say this, but at trial I cannot be in a position to blame or characterize the witnesses.
To point out the superiority of taking the original descriptions, which to summarize our research, we trust more, however complete or incomplete it may be. The original description did not include a beard. These were the bases for setting up a test of memory. Failure to do that may show it a typical assimilation process, meaning the witnesses change to what they pick up later.
Tr. at 19-20.
Dr. McCloskey testified that his own research regarding the assimilation factor contradicts that of others and suggests that no such phenomenon exists. Tr. at 68-69. He also stated that from his reading of the testimony, he could find no facts to indicate that any of the witnesses received additional information after their contact with Reverend Claymore. Tr. at 68, 78. [Dr. McCloskey testified that he reviewed the transcripts of the suppression hearing and the trial. Tr. at 67. Dr. Buckhout stated that he read summaries of the testimony prepared by counsel. Tr. at 34.]
C) The Confidence-Accuracy Relationship
Dr. Buckhout stated that studies he and other researchers have conducted have consistently shown no correlation between the confidence that witnesses express in their identifications and the accuracy of those identifications. Tr. at 20-21. He opined that confidence is a function of social pressures. He suggested, for example, that a witness might express 100% certainty in his/her identification due to the desires of authorities. Tr. at 21. Stating that Dr. Buckhout had "ignored a large part of the literature," Dr. McCloskey testified that studies of the confidence-accuracy were inconclusive. Tr. at 69-70. He further testified that the circumstances under which there is either a strong or a weak correlation between confidence and accuracy are unknown.
D) The Checklist
Dr. Buckhout proffered a checklist which he has designed to identify potential sources of bias in photographic lineups. [D-2]. The checklist contains thirty questions, including the following examples:
The Third Circuit rejected defendant's argument that this court erred in refusing to suppress the photographic identifications and in-court corporeal identifications of defendant. At 1244, n. 28. Of course, as part of his defense at trial, defendant could challenge the fairness of the photographic lineup in an attempt to impair the credibility of the witnesses' identifications.
Do the participants have different skin tones?
Are the participants different in modes of dress?
Did the officer in charge make the witness aware that a suspect is present in the photospread?
Are any of the photographs distinctive in regard to their color or texture?
A "yes" answer is supposed to indicate a potential source of bias. Dr. Buckhout testified that he compiled the checklist as a result of his career-long study of testing methods and his experience in working with police departments on lineups and photospreads. Tr. at 22-23.
Dr. McCloskey criticized the checklist as being subjective, i.e. different people could answer the same question in different ways. Tr. at 74-75. He stated that another problem with the checklist is that it simply makes note of variations among the photographs, but cannot indicate whether such variations result in bias to the defendant. Tr. at 73. For example, Dr. McCloskey noted that while the photograph of defendant presented the darkest background, photograph # 6 contained the lightest background. Thus, it was not possible, in Dr. McCloskey's opinion, to conclude that the dark background of defendant's photograph necessarily biased the photospread against him. Tr. at 74.
II ADMISSIBILITY UNDER FED.R.EVID. 702
Having reviewed the testimony of the evidentiary hearing, the court must assess it against the guidelines set forth by the Third Circuit.
A) The Soundness and Reliability of the Process or Technique Used in Generating the Evidence.
While the Third Circuit rejected the Frye test, the degree to which a scientific technique has gained acceptance may still be decisive. At 1238. Other factors which the court identified as bearing on the question of reliability include: (1) the relationship of a new technique to established modes of scientific analysis; (2) the existence of a specialized literature dealing with the new technique; (3) the qualifications and professional stature of expert witnesses; (4) the non-judicial uses to which the scientific technique are put; and (5) the frequency with which a technique leads to erroneous results. At 1238-1239.
In the case sub judice, the testimony of the two experts focused on conclusions which have been drawn from research on perception and memory. Neither expert presented, in any detail, the raw data upon which such conclusions have been based. Nor did they discuss the methodology used to generate the data. These omissions, standing alone, make the court hesitant to admit expert testimony. Absent a presentation of either the methodology underlying the perception and memory studies or the data such studies have generated, an important basis for assessing the reliability of the experts' testimony is lacking.
This court is mindful of the Third Circuit's statement that, "From the facts available on the record and otherwise, it would appear that the scientific basis for the expert evidence in question is sufficiently reliable to satisfy Rule 702." At 1241. The court went on to quote People v. McDonald, 37 Cal.3d 351, 208 Cal.Rptr. 236, 690 P.2d 709 (1984), a case approving of the use of expert testimony regarding eyewitness identifications, where it was stated that "the consistency of the results of these studies [i.e. research into the psychological factors affecting eyewitness identification] is impressive." 208 Cal.Rptr. at 245, 690 P.2d at 718. With due deference to the Third Circuit and the Supreme Court of California, the testimony presented at the evidentiary hearing held by this court belies such consistency.
For example, in State v. Chapple, 135 Ariz. 281, 660 P.2d 1208 (1983), a case which held that it was reversible error for the trial court to exclude psychological testimony on eyewitness identification, the proffer included testimony by Dr. Elizabeth Loftus to the effect that memory does not diminish at a uniform rate. 660 P.2d at 1220. However, the Shepard, Ellis, and Davies study submitted to the court found a uniform rate of forgetting. The authors of the study stated that, "This is an unusual effect in memory research, and may be peculiar to facial or person identification." [D-1 at 86]. Without adequate information as to the shape of the forgetting curve, the jury is merely left with the common sense notion that memory declines over time.
Other areas of inconsistency concern the assimilation factor and the confidence-accuracy relationship. Dr. McCloskey denies the existence of the assimilation factor and asserts, in the alternative, that if it does exist, the circumstances under which it will arise are unknown. Tr. at 69. Dr. McCloskey also denied that it was possible to draw any conclusions with respect to the confidence-accuracy relationship. Both Dr. Buckhout and Dr. McCloskey were able to cite their own studies and the studies of others to support their respective opinions on the confidence-accuracy relationship. Tr. at 20-22; 69-71. Significantly, the Shepard, Ellis, and Davies study relied on by Dr. Buckhout for its conclusions with respect to retention interval was far from conclusive in its findings on the confidence-accuracy relationship. Their report states that:
From a practical standpoint, the results have an intriguing ambiguity. It would appear on the one hand that the chances of a person identifying a suspect after almost a year are much less than his chances after a week or even three months; but on the other hand, anyone who makes an identification with a high level of certainty after such a period is likely to be correct. However, it must be stressed that the instructions given were designed to discourage guessing by those who were uncertain, there were no consequences for the person selected, and only two people made a correct identification after the longest delay, far too few to enable any firm conclusions to be drawn.
[D-1 at 86].
In view of the inconsistent results produced by the studies and the lack of testimony regarding either the methodology of those studies or the underlying data on which the test results are based, the court finds that the proffered testimony of Dr. Buckhout does not carry with it a sufficient degree of reliability to warrant its admission. Although the court finds this evidence unreliable, the court will proceed to a consideration of the other two guidelines set forth by the Third Circuit in the interest of developing a complete record.
B) The Possibility that Admitting the Evidence Would Overwhelm, Confuse, or Mislead the Jury
Assuming that this court found defendant's evidence reliable, it would then have to balance its reliability against the possibility that the evidence might confuse or mislead the jury. Of course, expert testimony is generally accorded a "`presumption of helpfulness.'" At 1241 (quoting In Re Japanese Electronic Products Antitrust Litigation, 723 F.2d 238, 279 (3d Cir. 1983)). The Third Circuit noted, however, that caution is appropriate, particularly in the criminal context, "whenever proffered novel scientific evidence, if unreliable or likely to mislead, will increase the likelihood of an erroneous verdict." At 1241. The court pointed out that one way in which expert testimony could mislead a jury is "where the jury is not presented with the data on which the expert relies, but must instead accept the expert's assertion as to the accuracy of his conclusions." Id. at 1239.
The court reiterates that the two experts in the case sub judice neither presented the data on which they relied nor the methodology used to generate that data in any great detail. This corresponds precisely to the situation in which the Third Circuit warns of a danger of misleading a jury. Absent such information, a jury has little basis for evaluating the testimony they hear. Accordingly, the court finds that even if the evidence offered by defendant was reliable, it could not be admitted due to its risk of misleading the jury.
C) The Proffered Connection Between the Scientific Research or Test Result to be Presented, and the Particular Disputed Factual Issues in the Case.
[A] defendant who seeks the admission of expert testimony must make an on-the-record detailed proffer to the court, including an explanation of precisely how the expert's testimony is relevant to the eyewitness identifications under consideration. The offer of proof should establish the presence of factors (e.g. stress, or differences in race or age or between the eyewitness and the defendant) which have been found by researchers to impair the accuracy of eyewitness identifications.
At 1242.
It is perhaps in establishing the "fit" between the scientific research presented by Dr. Buckhout and the disputed factual issues of this case that defendant's argument for the admission of expert testimony is weakest. The following observations identify some of the more important differences between the research on retention interval and the facts of this case.
1. The research has not generally tested the accuracy of a person's memory for periods of up to three years.
2. In most studies, the exposure of a perpetrator" to a "witness" occurred for periods of less than one minute. In contrast, the twelve witnesses who identified the defendant as the man they knew as Reverend Claymore witnessed defendant for periods ranging from five to forty-five minutes. Even Dr. Buckhout conceded that at exposures of this length "at least at the start you have all the ingredients of a good memory of a person's face." Tr. at 39.
3. Most studies which have attempted to simulate crimes have simulated crimes such as rape, robbery, and mugging. In these studies, the motivation of the witnesses was not an important factor. Rather, the witnesses are merely supposed to be good citizens coming forward. Tr. at 38. The twelve witnesses who identified the defendant, however, not only were not confronted by the fear and stress presented by a violent crime situation, but stood to realize a financial gain as a result of their transactions with Reverend Claymore.
These differences, taken together, reveal that there is, at best, only a weak connection between the research on retention interval and the disputed facts of this case. Accordingly, even if defendant's proffer on retention interval survived the first two prongs of the analysis for admissibility under Rule 702, such testimony would still have to be excluded. With respect to the assimilation factor, it does not appear from the testimony of any of the witnesses that they assimilated any facts learned subsequent to their initial contact with Reverend Claymore. Therefore, the expert testimony on this issue cannot be connected with any facts of this case. Finally, as to the confidence-accuracy relationship, the testimony of Dr. Buckhout does not reveal the situations that the studies have tried to simulate. Accordingly, the court cannot determine that there is any "fit" between the confidence-accuracy studies and the disputed facts of this case.
In view of the foregoing, the court declines to admit the expert psychological testimony proffered by defendant under Rule 702.
The court takes this opportunity to address the checklist proffered by Dr. Buckhout. As the court understands the proffer, the checklist was offered as a mere guide to identifying potential sources of bias in photospreads. It was not claimed that the checklist represented a scientifically reliable method for predicting when a photospread will, in fact, be biased. For example, Dr. Buckhout did not claim that any studies of his checklist had been conducted to determine its reliability for predicting bias. Dr. Buckhout did not indicate how many "yes" answers on the checklist were required for a showing of bias. Moreover, as Dr. McCloskey correctly pointed out, the checklist merely indicates variations among the photographs, but does not indicate whether these variations will necessarily result in bias to the defendant. Since testimony about the checklist by Dr. Buckhout could give it an improper aura of scientific reliability, the court declines to admit this testimony.
ORDER
The court determines that the proffered testimony by the defendant, John W. Downing, of his psychological expert should not be admitted.The judgment of conviction is REINSTATED.
The court's prior sentence of defendant is AFFIRMED.
Defendant shall report to the institution to be designated by the Bureau of Prisons to serve his sentence fifteen (15) days from the filing of this order.
IT IS SO ORDERED.