Tuesday, March 29, 2011

Texas Court Puts Another Nail in the Scent Lineup Coffin, but Should the Concept Be Buried Entirely?

A recent appellate decision continues the trend of Texas courts to reject scent lineup evidence, as to which those courts were once the most welcoming in the country. The case, Texas v. Smith, 2011 WL 480600 (Ct. App. 2011), adds some nuance to prior decisions regarding scent lineups, justifying an analysis.

Officers collected scent samples from inside the car in which a murder victim was found in 2003. Two years later, a dog handler, Deputy Keith Pikett, conducted a scent lineup with six individuals, all black males, standing 25 to 30 feet apart in a wagon wheel formation. Three bloodhounds were scented to samples that had been recovered from the vehicle. The dogs alerted to two individuals in the lineup depending on what they were scented to, but all alerted to Jason Smith, the defendant, after being scented on a Wessen .357 Magnum recovered during the investigation. Another scent lineup (or lineups) were performed regarding another defendant besides Smith, but produced no identifications.

Smith moved for a hearing on the admissibility of the evidence under the state’s case law on scientific evidence. The trial court initially approved admission of the scent lineup, but changed its mind on a subsequent motion. The court made 16 findings of fact:

1. Possible cross-contamination of the scents in the lineup in question, as admitted by the handler himself to explain one misidentification in one of the lineups.
2. Lack of any “blind” scent lineup without the defendant (though Pikett did conduct a scent lineup without the defendant but with a co-defendant).
3. Failure of the handler to keep complete records of scent lineups his dogs have performed.
4. Incomplete training records of the dogs.
5. Lack of accuracy or error rates because of the failure to keep records.
6. Handler’s records, such as they were, were not subject to peer review.
7. Failure of the handler to record the dispositions of cases in which his dogs participated.
8. Failure of the handler to perform validation testing on his dogs during scent lineups.
9. Handler testified that no one reviewed his work.
10. The bloodhounds were not certified and, in any case, the court noted the lack of any recognized industry standard on bloodhounds and no certification program for bloodhounds.
11. While the idea that bloodhounds can track and identify scents was accepted as valid, there is no clearly accepted method for conducting scent lineups.
12. No literature offered by the prosecution in support of the manner in which the scent lineups were conducted.
13. No independent evidence presented by the prosecution regarding the potential rate of error in the lineups.
14. No evidence presented by the prosecution regarding the availability of other experts to test and evaluate the manner in which the scent lineup in question was conducted.
15. The defense presented evidence that the dogs in question could have been intentionally or unintentionally influenced by the dog handler because the manner in which the scent lineup in question was conducted (i.e., potential for cueing).
16. No showing that the scent lineup results could be duplicated by others following the same methods.

The use of the term “blind” in finding 2 is confusing. Blindness in a scent lineup usually refers to procedures in which the handler and dog are ignorant of which person in the lineup is the suspect, or which station contains the suspect’s scent. Here, blindness appears not to refer to the knowledge of the handler and others present during a trial about the position of the suspect, but rather to a trial in which no suspect would be present. A trial in which no alert will be correct is usually called a blank, a zero trial, a negative trial, or a non-match run (see SWGDOG SC 9: Human Scent Dogs: Scent Identification Lineups §2.3.2.5.3).

The trial judge concluded that Pikett’s methods in the case had not been shown to be sufficiently reliable to be admitted and that he could not testify as an expert. The prosecution appealed.

The Texas Court of Criminal Appeals reviewed the test enunciated in Nenno v. Texas, 970 S.W.2d 549 (Tex.Crim.App. 1998), which applies where training and experience, rather than scientific principles, are determinative of the reliability of the evidence. This test requires a showing that (1) the field of expertise is a legitimate one, (2) the subject matter of the expert, in his testimony, is within the scope of the field, and (3) the expert's testimony properly relies upon or utilizes the principles involved in the field. The test had been applied to admit scent lineup evidence in Winston v. Texas, 78 S.W.3d 522 (Ct. App. 2002). More recently. the Texas Court of Criminal Appeals in Winfrey v. Texas, 323 S.W.3d 875 (Tex.Crim.App. 2010) held that a scent lineup could not support a conviction, but Winfrey did not address admissibility under Kelly and Nenno. Kelly v. Texas, 824 S.W.2d 568 (Tex.Crim.App. 1992) is the Texas case describing the admission of evidence from the hard sciences.

In reviewing the case against Smith, the Court of Appeals agreed with the trial court’s determination that Deputy Pikett’s records were incomplete:

"Deputy Pikett's testimony reveals that though he maintains some records for training and practice lineups, these records are incomplete as are his records verifying the outcome of cases in which his dogs have identified suspects and the records of the dogs' success rates. Given the evidence before the trial court at the September 12, 2007 hearing, the trial judge reasonably could have concluded that the State had failed to carry its burden of showing that the proffered expert testimony was reliable."

The appellate court then distinguished Winston, which also involved Pikett, but where Pikett’s testimony was admitted, by noting that in that case, unlike Smith, Pikett had referred to his procedures being consistent with the manual of the National Police Bloodhound Association. In Winston, Pikett had produced success rates, making the case “factually distinguishable” from Smith. Perhaps this is a way of saying that the prosecution was better prepared in Winston than it was here, or perhaps the Smith court was saying that the Winston court had not looked at the evidence as carefully. Alternatively, Pikett was testifying in a more friendly environment in the earlier case and did a better job of covering over the deficiencies in his recordkeeping.

The appellate court also noted that in Smith Pikett used multiple dogs and ran multiple tests, “contrary to expert recommendations and accepted methods.” The court notes that “multiple dogs can create an unwanted pattern for the dogs or trainer.” This is correct insofar as it goes. By leaving a trail for other dogs to follow, or saliva on an item, dogs may be following each other rather than making independent findings. (See Jezierski T, Walczak M, Gorecka A (2008). Information-seeking behaviour of sniffer dogs during match-to-sample training in the scent lineup, Polish Psych. Bull. 39(2), 71-80.) Nevertheless, as I have summarized elsewhere, use of multiple dogs can increase accuracy rates for positive identifications (though reducing the number of positive identifications overall since the more dogs used, the less often they will all make the same match). However, procedures involving multiple dogs assume that the testing environment of a scent match lineup is rigorously cleansed between trials. Also, the use of zero trials without the defendant (finding 2) would presume additional trials should have been conducted, so some of the court’s findings may be inconsistent.

The appellate court, like the trial court, referred to blindness, using the term “double blind” in stating that “Deputy Pikett indicated that he keeps the process simple and does not conduct a ‘double blind’ lineup.”

"According to the record, Deputy Pikett testified that he did not run a blind test to validate the accuracy of dogs' original identification of Smith. Deputy Pikett cited his science background to 'keep it simple, stupid, and get out of all that crap,' when others tried to complicate the process."

Given the lack of discussion of scent lineup research, it is unclear what others Pikett thought were complicating the process, or what research or protocols he thought were crap, but his criticism of double-blinded testing certainly conflicts with SWGDOG and other protocols. See SWGDOG SC 9 Human Scent Dogs: Scent Identification Lineups §2.3.3.6 (“No one present in the room, including the dog and handler shall know the correct outcome of the lineup.”). Again, it is uncertain exactly how either Pikett or the court were defining “blindness.”

The defense presented evidence that Pikett may have “intentionally or unintentially influenced” his dogs, i.e., cued them. Cueing has recently been the subject of a study by three scientists at the University of California at Davis, research which is likely to lead to this issue being referenced more often as a flaw in the use of dogs in a wide range of detection and identification work by police. On the appeal in Smith, the prosecution argued that Pikett “could not have influenced his dogs in their identification of Smith because he did not know in which position Smith stood….” The appellate court rejected this assertion because Pikett had access to Smith’s case file and set up the lineup.

The prosecution made another argument on appeal, saying that the trial court had effectively applied Kelly, not Nenno—i.e., that the trial court had imposed the standards for the admission of evidence derived from the hard sciences. The prosecution said that findings 6 (peer review of records not performed), 9 (lack of review of handler’s work), 12 (no scent lineup literature offered), and 14 (no supportive experts), should not have gone to the admissibility of the evidence, only its weight. It is not clear to me where peer review was expected to come from, or if items 6 and 9 are distinguishable. Perhaps item 6 refers to review of Pikett’s findings in the particular case, while item 9 might involve periodic testing of Pikett to validate his procedures and their reliability. The latter might have been provided by an organization like that National Police Bloodhound Association, but perhaps Pikett had not been attending such gatherings despite his reference to the association in prior cases. The Smith court noted that “hard science methods of validation, such as assessing the potential rate of error or subjecting a theory to peer review” might be appropriate “for testing the reliability of fields of expertise outside of hard science in appropriate cases.” This is an important observation and could have significance for other canine evidence cases.

After going through these and other arguments, the Houston branch of the Texas Court of Appeals affirmed the trial court, stating that “the trial court could have concluded that there was too large of an analytical gap between the data offered and Deputy Pikett’s proffered opinion,” and that a “trial court should not admit expert testimony that is connected to existing data only by the expert’s own assertions.” Therefore, the trial court did not abuse its discretion in excluding the scent lineup evidence.

Regardless of which evidentiary standard was applied by the trial court—the standard applied to hard scientific evidence or a less rigorous standard of the sort often applied to scent lineups, or something in between—the lineups in this case were highly deficient. The possible cross-contamination may not have been the handler’s fault, but no contact should have been permitted. Training and performance records should be complete, and should include, or easily produce, accuracy rates. Dogs and handlers should be tested periodically. The failure of the prosecution to introduce scientific literature may have been due to the fact that the scientific literature by Schoon and more recently by Jezierski and others would not support use of the human scent lineup format apparently preferred by Pikett, and would certainly have indicated that many of his procedures were unsatisfactory from a scientific perspective. Alternatively, the prosecution’s failure may be explained by the frequent tendency of trial lawyers to overlook the significance of and the need to prepare for the introduction of canine evidence.

The pendulum is swinging strongly against the admission of scent lineup evidence. I do not quarrel with the decisions so far since the U.S. scent lineups I know anything about should not have been admitted in the first place. On that, Professor Taslitz is still correct. Nevertheless, I am bothered by the blanket junk science label since this implies that scent lineup procedures can never be refined and controlled enough to provide reliable evidence. As I have argued in the draft paper mentioned above, research on scent identification has begun to produce results comparable to other forensics procedures that are widely admitted in criminal trials.

Thanks to Gene Papet for helpful comments on this note.

No comments:

Post a Comment