• Selected
• Entire List

Abstract:
The CSI Effect, if there is one, should be a matter of utmost concern within the legal community. Is there a CSI Effect? How would we know whether there is such an effect? And what precisely is the CSI Effect anyway? This Article describes the popular television program CSI: Crime Scene Investigation and analyzes the distinct features of the show that have led some to claim it could impact legal behavior. It then describes the media coverage which has largely been responsible for claims that there is a CSI Effect. We note that the media has applied the term CSI Effect loosely and that the term has been used to mean rather different things. We propose a typology of six different effects, all of which have been characterized by the shorthand term CSI Effect. We argue that clarity over which effect is being described is crucial in any discussion of whether the CSI Effect is real. In particular, we caution against arguments which use evidence of one effect to support claims about another. We then discuss four types of evidence that have been marshaled in support of the claim that there is a CSI Effect - anecdotes, surveys of legal actors, simulated jury decision-making studies, and acquittal rates in criminal cases - and analyze the strengths, weaknesses, and current findings of each. We conclude that there is little support for the gravest of the CSI Effects, which is that jurors who watch CSI are wrongfully acquitting in cases lacking forensic evidence or that they are wrongfully convicting based on an unrealistic belief in the infallibility of forensic science. In light of the weak empirical support for the CSI Effect, we discuss some possible alternative explanations for the media attention to the CSI Effect. Finally, we discuss one of the most interesting, from a legal point of view, aspects of the CSI Effect, the claim that it has altered the burden of proof. We will argue that courts have correctly concluded that there is no legal merit to this argument, but we will use the argument as a clue that helps us posit another possible explanation for the CSI Effect phenomenon.

CSI effect, forensic science, burden of proof, reasonable doubt, media, juries

Abstract:
While a fair amount of judicial and scholarly attention has been devoted the admissibility of latent print evidence under the Daubert standard for expert evidence, there has been no evaluation of its admissibility under the Frye standard. This is due the widespread assumption that latent print evidence is obviously admissible under Frye. This, in turn, is based on two assumptions: that latent print individualization is generally accepted in the relevant scientific community and that non-novel evidence is immune to the Frye test. Both assumptions are shown to be false. The article introduces the concept of meta-expertise to denote scientists and scholars who evaluate the knowledge claims of other experts. An analysis of the state of opinion in the scientific community shows that latent print individualization is not generally accepted. This lack of general acceptance points to a broader issue beyond the admissibility of the evidence: the continuing dramatic disconnect between the legal and scientific communities regarding the validity of latent print individualization.

Frye, fingerprint, expert evidence, general acceptance, relevant scientific community, meta-expert

Abstract:
The recent exposure of an erroneous latent print identification by the FBI that led to the false arrest of Oregon attorney Brandon Mayfield has punctured the myth of the "infallibility" of fingerprint identification and generated renewed interest in the "error rate" of fingerprint identification. This article undertakes a comprehensive review of what is known about the potential error rate of latent print identification. The article first presents a compilation of all known exposed cases of fingerprint misattributions. Although only twenty such cases have been documented, an analysis of these cases suggests that these cases likely represent only a small portion of the true set of latent print misattributions. Then, the article compiles and analyzes proficiency test data that sheds some light on the potential error rate of fingerprint identification. The second half of the article is devoted to the fingerprint profession's and courts' rhetorical accounts of the potential error rate of latent print identification. This section analyzes efforts to minimize, dismiss, or otherwise account for fingerprint error. Fingerprint examiners make claims of error-free practice that belie the reality of error. The article concludes that we must confront, analyze, and seek to understand error if we want to reduce it.

fingerprint, error

Abstract:
The death penalty debate in the United States has recently undergone a fundamental shift. The possibility of executing the innocent has emerged as some abolitionists’ most salient argument, displacing debates over such issues as fairness, deterrence, and cost. Innocence has managed to move to the fore of the debate in part because of the epistemological certainty attached to one particular kind of postconviction exoneration, one vouched for by the authority of DNA evidence. We suggest that such rhetorical moves draw upon the epistemic authority of science as lever with which to challenge law’s claims to truth-making authority. A few abolitionists and other scholars have expressed misgivings about the abolitionist embrace of the innocence argument. We push this concern further, suggesting that both abolitionists and death penalty reformers, who seek to promote a “scientific” death penalty centered on DNA evidence, draw upon a mythologized notion of “science” as a producer of epistemic certainty. Paradoxically, this association of science with certainty is inconsistent with contemporary notions of science as characterized by efforts to measure, manage, but always acknowledge, uncertainty.

death penalty, capital punishment, science, DNA, innocence, abolition

Abstract:
This paper reconceptualizes the Daubert admissibility regime using the "evidence-based" metaphor. Although contemporary society is pervaded by calls for such things as medicine, policy, corrections, and crime prevention to be "evidence-based" and evidence is firmly associated with law, there has been little application of this notion in law and little recognition of the homology between evidence-based medicine and the Daubert inquiry. The paper argues that the Daubert inquiry may be conceived as a demand for "evidence about evidence," or "evidence-based evidence." It then uses the recent controversy over the admissibility of latent print (fingerprint) evidence to illustrate this notion. It shows that proponents of latent print evidence have had difficulty producing evidence about the reliability (or accuracy) of latent print evidence. Instead, trial courts have tended to find latent print admissible based on evidence that does not pertain directly to the accuracy of latent print evidence. Therefore, latent print evidence, as yet, is not "evidence-based evidence." Finally, the paper suggests that this state of affairs may explain the exclusion of latent print evidence in one recent case.

Daubert, fingerprints, evidence, evidence-based medicine, forensic science

Abstract:
Beginning in 2002, popular media disseminated serious concerns that the integrity of the criminal trial was being compromised by the effects of television drama. Specifically, it was widely alleged that the popular CSI franchise, one of the most watched programs on television, was affecting jury deliberations and outcomes. It was claimed that jurors confused the idealized portrayal of the capabilities of forensic science on television with the actual capabilities of forensic science in the contemporary criminal justice system. Accordingly, jurors suffered from inflated expectations concerning the occurrence and probative value of forensic evidence. When forensic evidence failed to reach these expectations, it was suggested, juries acquitted. In short, it was argued that, in circumstantial evidence cases in which juries would have convicted before the advent of the CSI franchise, juries were now acquitting. As we have argued elsewhere, such charges, if true, would constitute a serious challenge to law's fundamental faith in the jury and thus raise serious questions about the integrity of the criminal justice system itself. However, the media perpetuated these claims in the absence of any convincing evidence that there was any such effect. Thus far, social science studies intended to detect the CSI effect have found little evidence of it. Unlike other studies that seek to test the CSI effect by thorough juror surveys and simulations, our approach has been to analyze actual acquittal rates in criminal trials. In an earlier analysis of federal trial data, we found no change in acquittal rates correlated with the advent of CSI. Although this finding does not itself disprove the CSI effect, it does suggest that the media claims that there is such an effect were premature. In the first part of this paper, we present additional data derived from state criminal trials that finds only equivocal evidence of a significant change in acquittal rates in response to CSI. We then present a content analysis of media claims about the CSI effect. This analysis shows that, contrary to the social science evidence, a consumer of popular media would have the impression that CSI is causing wrongful acquittals in our criminal justice system. In the final part of this paper, we draw parallels between the media storm concerning the CSI effect and another episode in which media claims about a severe social problem in the legal system were widely disseminated despite the absence of any convincing evidence that the problem was in fact occurring. The episode was the widespread concern about a supposed litigation explosion or hyperlexis during the 1970s and 80s. Echoing the legal scholarship concerning the litigation explosion, we explore the underlying anxieties that may generate media claims about social problems in the legal system.

CSI, juries, media, litigation explosion, forensic science, television, acquittal

Abstract:
Taking stock more than a decade after Daubert v. Merrell Dow Pharmaceuticals, this paper argues that legal scholars, including the author, have perhaps focused too much on the validity of expert evidence and not enough on the propriety of expert testimony. It argues that legal scholars who disagree about many aspects of expert evidence nonetheless agree that "fit" between the testimonial claim and the empirical support for that claim is a crucial issue. The paper then suggests that much of the recent controversies over forensic evidence may be attributed to what Professor Friedman has called "over-claiming," the exaggeration of the probative value of the evidence during expert testimony. The paper concludes by presenting some empirical data from a study of actual expert testimony about latent print analysis. The testimonial claims vary significantly from case to case, but fall into several broad categories. Although the testimonial claims are not always logical on their face, the paper argues that these testimonial claims make juries comfortable with conviction. The paper concludes by calling for more scholarly attention to expert evidence as expert testimony.

Daubert, expert evidence, fingerprints, forensic science, expert testimony, science and technology studies

Abstract:
Beginning around 1999, a growing number of scholars have claimed that validation studies for forensic fingerprint identification do not exist. This article revisits that claim by reviewing literature produced by proponents of fingerprint identification in response to that charge. It shows that fingerprint proponents employ rhetorical tricks in which they claim to address the validity question, but then subtly shift the question to ones that are easier to address. The article explores several different rhetorical strategies fingerprint proponents use to appear to be demonstrating validity, while in fact demonstrating other things. These include the fingerprint examiner's fallacy and the casework fallacy. The inability of fingerprint proponents to refute the charge that validity studies are lacking is further evidence that the charge is, in fact, correct.

Abstract:
Efforts to harness computer fingerprint databases to perform studies relevant to fingerprint identification have tended to focus on 10-print, rather than latent print, identification or on the inherent individuality of fingerprint images. This paper reports on three experiments that measure the accuracy of a computer fingerprint matcher at identifying the source of simulated latent prints. The first experiment used rolled prints supplied by the National Institute of Standards and Technology (NIST) to simulate latent prints. The second experiment used our own manufactured latent prints. The third experiment used latent prints supplied by NIST. An Automated Fingerprint Identification System (AFIS) was used to simulate the task that a human latent print examiner is typically asked to perform as part of ordinary casework. The AFIS performed this task, for which it was not designed, fairly well. However, there are non-mate images that scored very highly on the AFIS's similarity measure. These images would be susceptible to erroneous conclusions that would be given with a very high degree of confidence. Not surprisingly, the same was also true of the simulated latents which contained less information. We suggest that measuring the accuracy and potential for erroneous conclusions for AFISs might provide a basis for comparison between human examiners and automated systems at performing various identification tasks. Such comparisons might stimulate competition, innovation and improvement in the performance of these tasks.

biometrics, error analysis, fingerprint identification, latent prints

Abstract:
The probative value of comparative bullet lead analysis (CBLA), a now discontinued technique that was used by the Federal Bureau of Investigation for more than 30 years, has been hotly debated over the last several years. One issue that has received relatively little attention concerns the degree of geographic dispersion of bullets as they pass from manufacturers to retailers. Proponents and critics of CBLA alike agree that geographic distribution is such a major consideration, if not a predominant one, that it could significantly diminish, or completely erode, the probative value of a CBLA 'match' or, in some cases, even make a match counter-probative. The inattention to this issue to date appears to be a consequence of lack of data, rather than lack of importance. Until now, no datum concerning bullet distribution has been presented in the public domain, critically hampering the proper estimation of the probative value of a CBLA match. In this paper, we use manufacturer packing codes on boxes of bullets in retail outlets at four sites in the United States as a surrogate measure of bullet lead compositions to gauge local retail bullet distribution. Using a weighted average packing code match probability, we found very high degrees of geographic concentration of bullet packing codes. Although these findings can only offer a rough estimate of the degree of geographic concentration of actual chemical compositions of bullets, they are sufficient to establish that geographic concentration does, in fact, exist. Such a concentration would have a significant impact on the probative value of any claimed CBLA match.

comparative bullet lead analysis, compositional analysis of bullet lead, geographic concentration, geographic distribution, probative value, National Research Council, forensic bullet composition comparisons, forensic bullet evidence

Abstract:
DNA profiling and searchable databases enhance the ability of policing organizations to search for criminal suspects. In many respects, these technologies are incorporated within traditions of police work, supplementing familiar "subjective" methods of constructing suspects. In other ways, however, the construction of DNA databases in Britain, the United States, and elsewhere shifts criminal investigation toward suspect populations and statistical suspects. Not only is DNA evidence used to confirm that a criminal suspect is the source of crime scene evidence, it can be used to search freely through a suspect population for a possible source of such evidence. This method, commonly known as database trawling, comprises a new way of constructing suspects, one that bears close connections with new data mining technologies for prospectively identifying terrorist suspects.

DNA evidence, criminal databases, suspect populations, forensic science