DNA as Evidence: Skirting the Fallacies - Essay Example

?DNA as Evidence: Skirting the Fallacies 0 Introduction The use of DNA, which is short for deoxyribonucleic acid, in identifying a person’s geneticmake-up and distinguishing it from others is a major breakthrough in criminal investigations. This began when Dr Alec Jeffreys, an English geneticist, isolated a region in the DNA strand that contains a consistently replicated DNA sequence, which is unique to an individual. This further led him to formulate DNA typing or the process of isolating that region of the DNA strand and identifying it. [1] Since then, the use of DNA as evidence has become widespread. News about convicts who had spent years in prison, but were suddenly released because they were proven innocent of the crimes they were convicted of on the strength of new DNA evidence has made this new technology a source of wonder and hope for many who may have languished in jails for crimes they did not commit or those who were victim of unsolved crimes. Nevertheless, the use of DNA as evidence in court is actually a complicated issue and may become a source of technicality that can change the course of cases. In the UK, some cases had been overturned and decisions of convictions by trial courts quashed because of technicalities found in the presentation of DNA evidence during trials. These technicalities, often called fallacies, are the topic at issue in this paper. 2.0 The Emergence of DNA as a Vital Tool of Evidence: Background Immediately after the English geneticist Dr Jeffreys invented DNA fingerprinting, its use immediately debuted in the court in an English paternity case. [2] It was a year later however, that DNA typing proved its usefulness in a criminal case investigation in Enderby where two girls were raped and then slain in a span of three years, but had startling similarities in the execution of the crime. The same semen sample was also shared in the two cases. The first arrested suspect who had confessed to the crime was eventually released when the semen sample and the fingerprints found in both crimes scenes were not matched to him. With the help of Dr Jeffreys, the police eventually arrested Colin Pitchfork whose DNA matched those of samples found in both crime scenes. Dr Jeffrey used the VNTR (or variable of number repeats) genotyping through the SLP (single locus probe) system [3]. In 1987, both the UK and the United States have began allowing the use of DNA as evidence in criminal courts and in 1988 the Home Office and Foreign and Commonwealth Office allowed its use in immigration disputes. In 1989, the US Congress was forced to instigate a review of the matter when its use became the subject of attack in that country. Subsequently, relevant US government agencies issued results validating the scientific basis DNA fingerprinting and providing guidelines therefore [4]. Jeffrey’s VNTR genotyping was subsequently adopted by the Forensic Science Services, but the PCR-based (or polymerase chain reaction) system was soon developed, which targeted the SNPs (or single-nucleotide polymorphism) as opposed to the SLP in Jeffrey’s 1988 VNTR genotyping. Subsequently, the PCR-based system began to target the six-locus STR (or short tandem repeats), which eventually replaced the SLP system. In 1999, an expansion of the system was made with the introduction of the SGM Plus, an 11 loci system, and is presently the standard method being used in the UK. [5] Unlike the RNLF method of isolating DNA, which a required microgram amounts of intact DNA, PCR can make do with subnanogram amounts by amplifying them and detecting them in mere hours and not days. [6] Presently, 25 countries, more or less, are employing DNA fingerprinting in their respective forensic, both public and private [7]. 3.0 Presenting DNA as Evidence in Court: Fatal Errors and Fallacies 3.1 R v Doheny and Adams [1996] EWCA Crim 728 The R v Doheny and Adams case is a consolidation of two cases that were appealed to the Criminal Division of the Court of Appeal after the appellants were convicted of rape and buggery, and buggery, respectively. Both appeals were grounded, among others, on the unsafe verdict of the Jury due to the erroneous presentation of the DNA evidence during the trial. In the Doheny case, the victim did not have a clear look at the suspect’s case and described him only as having a local accent and by his general appearance and age. There was no other witness except that traces of semen were found on the victim’s clothing, a pubic hair on the mattress and another on the victim’s anus. This case, therefore, largely relied on forensic evidence. During the trial, two expert witnesses for the Crown were called to the stand: the first testified that the pubic hairs found on the mattress and the anus of the victim were not of the suspect, and; the second witness gave testimony as to the DNA tests conducted on the semen found in the scene of the crime and the semen sample provided by the suspect [8]. The bone of contention in this case, and for which an appeal was brought to the CA, was the manner in which the second expert witness was allowed to adduce evidence on behalf of the Crown. In addition, the manner with which he calculated the random ratio of occurrence was also called into question. In conducting the DNA examination, the expert witness made both multi-locus and single-locus probes. The first entails a selection of family of stutters, or distinctive portions of DNA, whilst the second necessitates a selection of only a pair of stutters from a known region of the DNA for laboratory examination. In the multi locus probe technique, the bands that form patterns are given a value of .26 each and are multiplied together to get the random occurring ratio of the matching pattern. In the single locus probe technique, on the other hand, matching bands are given specific ratio in accordance to their size and are all multiplied together to come up with the random ratio of occurrence. The expert witness in this case, multiplied the product in the multi-locus probe and the single-locus probe [9]. During his testimony, the expert witness testified that the likelihood that the owner of the semen stains on the victim’s mattress is 1 in 40 million, which is the product of the multi-locus probe multiplied with the product of the single locus probe result. Moreover, when the prosecutor asked him whether he was certain that it was the suspect, he answered in the affirmative, both question and answer unclear as to in what context the certainty was to applied to [10]. In quashing the conviction, the CA grounded its decision on the unsafeness of the conviction due to the manner in which the DNA evidence was presented by the expert witness not only because of the kind of question posed to him by the Crown prosecutor, but also by the lack of legitimacy in the procedure used by the expert witness in combining the results of the multi-locus probe and single-locus probe results. The Court stated that the question on the likelihood that the semen stains were that of anybody other than the suspect is a prosecutor’s fallacy and therefore, inappropriate whilst the second question should have not been propounded at all because it was for the jury to determine [11]. The case of Adams involves the crime of buggery committed by a member of the Samaritans against a psychologically disturbed woman who often called the organisation to seek help and comfort. In this case, the victim had given a complete description of the assailant and semen was recovered from a cushion in the victim’s house [12]. The forensic scientist who conducted the DNA test on the semen recovered from the cushion and compared it with the sample taken from the suspect did four single locus probe tests. The tests resulted into eight matching bands and on that basis calculated the random occurrence ratio at 1 in 27 million. During his testimony, however, the prosecutor for the Crown asked him to compare the semen recovered and the semen sample from the suspect and he answered that they were the same. Furthermore, he was asked the likelihood of the recovered semen from the scene of the crime belonging to another other than the suspect and his answer was 1 in 27 million [13]. On appeal, the forensic expert’s testimony was contested on the ground that it constituted prosecutor’s fallacy. Although the CA agreed that it did constitute prosecutor’s fallacy, it nevertheless, did not find this objection enough to warrant the allowance of the appeal considering that it cannot possibly overcome the existence of other pieces of evidence, such as the positive identification of the suspect by the victim that pointed to the guilt of the appellant [14]. 3.2 R v Deen [1994] Times LR 11 The Deen case is the first successful appeal of a convicted appellant involving an issue in the presentation of DNA. Here, the appellant was convicted of three counts of rapes in a series of rapes of young women in ground-floor apartments in the South Manchester area between 1987 and 1988. The DNA evidence against the suspect was one of the most important evidence presented by the Crown against him. The forensic expert had determined after conducting an examination of the DNA samples a match of 1 in three million. This figure represented the probability of the odds that that the specific combination of DNA and blood samples, also taken in the crime scene, would match in analysis of a blood sample from a randomly chosen person in a specific population. During the trial, the forensic expert who examined the DNA gave his testimony before the court by answering the following question propounded to him by the prosecution: “So the likelihood of this being any other man but Andrew Deen is one in three million?” The forensic expert answered in the affirmative and the jury subsequently convicted Deen to three counts of rape and sixteen years of prison life [15]. The Court of Appeal quashed the conviction when the case was elevated for appeal by the defence. The defence had appealed on three grounds, one of which is that the prosecutor’s fallacy was committed during the presentation of the forensic expert of his testimony. In essence, the prosecutor’s fallacy was committed when the forensic expert gave the same answer to two questions: the probability that an innocent person would match the DNA profile in the crime scene, and; the probability that a person is innocent given that he matched the DNA profile from the blood sample recovered in the crime scene [16]. 3.3 R v Adams [1996] 2 Cr. App.R. 467 This case also involves a rape victim who was attacked while walking home one night. The assailant initially approached the victim feigning to ask her for the time and suddenly raped her from behind. Semen from a vaginal swab was recovered and subjected to DNA testing and Adams was picked up for questioning and paraded for identification. The victim, who had earlier told the police that her assailant was white, clean shaven and between 20 and 25 years old, was not able to identify him from the line-up. The appellant was in his late 30s and had an alibi that was corroborated by his girlfriend. Nonetheless, on the basis alone of the DNA evidence, he was found guilty and was sentenced to 7 years imprisonment [17]. During the trial, the prosecution adduced evidence showing that the DNA match occurred in 1 to 200 million in the population. This was objected to by the defence as incorrect because it did not allow the possibility that the assailant might be a close relative of the suspect, who happened to have half-brother in his 20s. The defence presented an expert who suggested the use of a statistical formula that the jury can use to resolve the conflicting evidence in the case, viz. the evidence pointing to the innocence of the accuse and the DNA match. This statistical formula made us of the Bayesian theorem developed by Professor Donnelly of Oxford University. This formula entails the multiplication of values assigned to various probabilities such as the probability that the crime was committed by another man, the possibility of the suspect not being identified in a line-up by the victim and the probability of the alibi. The judge agreed and directed the jury to use the formula, but Adams was nevertheless, found guilty [18]. At the appeal court, the Bayesian approach was assailed by the Court as having trespassed into an area specifically reserved for the jury to determine. The Court held that it was the jury’s duty and function to consider all evidence gathered during the trial together and determine their comparative weight in handing out their decision as to whether the accused was guilty or not. To do this, the jury must make use only of their common sense and universal knowledge of the things around them [19]. 5.0 Discussion From the three cases discussed in the preceding paragraphs, it is evident that the manner of presentation of DNA evidence is crucial to a case. Even if lapses in the presentation of the DNA evidence may not be noted in the trial court and results in the conviction of the suspect, that conviction may be quashed on technicality by an appeal court and a retrial ordered or some other decision not desirable to the prosecution may ensue. As also illustrated in the aforementioned cases, theses lapses and technicalities in the presentation of DNA as evidence in court may come in the form of prosecutor’s fallacy and the application of the Bayesian theorem. Also belonging to this class of erroneous presentation of DNA evidence in court are the interrogator’s fallacy, the uniqueness fallacy and the defense fallacy. 5.1 The Prosecutor’s Fallacy The prosecutor’s fallacy, according to author de Macedo is a “conflation of two conditional probabilities” resulting in an erroneous conclusion because of the two distinct natures of such probabilities. For example, the formula p(M/I), where M stands for the match of the suspect with the DNA sample recovered in the crime scene and I stands for the innocence of the suspect, implies that the chances of the suspect matching the DNA sample is very low if he is innocent. This same formula cannot be equated with the formula p(I/M), which means an improbability of the suspect being innocent if a DNA match is found. The prosecutor’s fallacy stems from the desire of prosecutor’s to link the statements “it is improbable that there would be a DNA match if the defendant is innocent” and “it is improbable that the defendant is innocent if there is a DNA match” together [20]. In the preceding paragraph’s example, the two statements are not equivalent to each other, but are two distinct probabilities. The prosecutor’s fallacy, also called transposed conditional, is therefore, a misinterpretation of misapplication of statistics, which is evident in the Deen case [21]. The Deen case was the first case whose conviction was hinged entirely on the DNA evidence, which was presented as evidence to court after forensic experts were able to match 10 bands using the multi-locus probe technique. The occurrence rate of each band in the population was pegged at 0.26, whilst the blood group was said to be common in 26% of the population. The formula used to obtain the random frequency ratio was: P(E/H2 ) = (0.26)10 x 0.25 = 3.5 x 10-7, where the occurrence rates of all ten bands were multiplied altogether and their product further multiplied to the blood group occurrence. The product, which is 1 in 2.8 million, represents the random frequency ratio [21]. On the witness stand, however, the prosecutor made the mistake of asking the forensic expert the probability of the semen recovered from the crime scenes belonging to someone other than the defendant as well as asking the probability of the same semen belonging to any person other than the defendant, both of which were answered by the forensic expert as 1 in 28 million. This was compounded when the judge himself stated subsequently that the figure given by the forensic expert “approximates pretty well to certainty” [22]. The case became illustrative of the prosecutor’s fallacy considering that there was a confusion of the probabilities involving chance match and innocence, which are two different things. Author Adams calculates probabilities of both guilt and innocence of the defendant and finds two distinct almost opposing figures. Using 19 million as the estimated number of sexually active males in the UK, Adams developed the following equations: PO ~ ________1_______ = 5.3 x 10-8 19000000 (this represents the defendant’s probable guilt in a population of 19 million sexually active male) LR = ______1________ = 3 x 10-6 1/3000000 (this represents the defendant’s probable innocence expressed by the forensic expert’s figure of 1 in 3 million) The above formulas led Adams to the following conclsuion: P(H/E) = _____________1_____________ = 0.14 1 + (3 + 106 x 5.3 x 10-8) -1 P(H2/E) = 1- 0.14 = 0.86 Adam’s conclusion is that the guilt of the defendant in the Deen case can be estimated at the probable 14% and his innocence as 86%, which according to him, does not justify the forensic expert’s statements at the trial and the follow-up remark of the judge [23]. The prosecutor’s fallacy is also evident in the R v Donehy and Adams case when the forensic experts testified that the random occurrence ratio resulting from the examination of the DNA taken from the forensic traces in the crime scenes were 1 in 40 million and 1 in 27 million, respectively. In both cases, the same forensic experts also testified that the possibility that the same DNA matching would be less than the stated random occurrence ratio. The Court held that the forensic experts went beyond what they were supposed to testify, which was just to give the random occurrence ratio and the number of persons in the locality with the same DNA profile [24]. Nonetheless, the prosecutor’s fallacy was not the primary reason why the Doheny case was quashed, but the use of the Bayesian approach in calculating the guilt of the accused by the jury. On the other hand, the Court in the Adams case believed that the commission of the prosecutor’s fallacy cannot overcome other pieces of evidence such as the positive identification of the victim of the defendant [25]. In line with this, the Court has, in R v Doheny and Adams, given an outline of what a forensic expert must testify to during the presentation of DNA as evidence in court: first, he or she must state the comparative DNA findings together with the random occurrence ratio; second, the Crown must inform the defence in detail how the calculations as to the ratio and other statistics have been done, and; third, the Forensic Science Service should provide, if requested, relevant databases to the expert [26]. 5.2 The Use of Bayesian Theorem in Presenting DNA Evidence The Bayesian theorem was developed by Thomas Bayes in the mid-18th century and it is a method of assigning probabilities in accordance to accumulating information. It entails a continuous revision of probability assignment to accommodate and to correspond to additional data. An example would be if X is informed that his classmate A was shot and he is asked to give his opinion as to the probability that B committed it. If X gives the probability as p = 10 for the reason that he thinks that B is unfriendly and seems capable of doing, but ups this to 20 after being told that B owns a gun that looks like the kind that was used to shoot A. However, after being told that B was seen in movie house with a friend at the time that A was shot X lowered his probability estimate to 5. In this example, it is evident that the Bayesian theorem is largely subjective in nature [27]. The theorem can be expressed in the following formula: P (h/e & k) = P(h/e & k) P(h/k) P(e/k) UK courts, however, are not too appreciative of the Bayesian approach in the DNA presentation of evidence as illustrated in the cases of R v Doheny and Adams and R v Adams [28]. In R v Adams, for example, the CA stated that the Bayes Theorem is not ideal for use during evidence presentation in jury trials or even as a means of assisting juries in deciding cases. This is primarily because, according to the CA, the use of a mathematical formula is inconsistent with the task of a jury. The jury is supposed to decide cases by evaluating evidence using their common sense and universal logic and not through the application of any mathematical formula. Moreover, the CA stated “to introduce Bayes theorem, or any other similar method, into a criminal trial plunges the jury into inappropriate and unnecessary realms of theory and complexity deflecting them from their proper task” [29]. The implication in the Adams case is that the Court objects to the use of the Bayesian approach not because it is deemed flawed, but because it is perceived as “inappropriate” for its mathematical nature. The soundness of the theorem is not what is being questioned by the Court, but the fact that it might not be easy for the jury to use and divert it from its primary function. This is especially true in the Adams case where pieces of evidence are contradicting each other: the defendant was not identified in a line-up; the description of the defendant did not match the description given by the victim; the defendant’s alibi was corroborated and was not controverted, and; his DNA matched the DNA in the semen from the vaginal swab of the victim [30]. 5.3 The Defendant’s Fallacy Closely allied with the defendant’s fallacy is another fallacy called defendant’s fallacy. This fallacy is committed when the defence attempts to use DNA statistics to its advantage. For example, the forensic expert gives the random occurrence ratio of 1 in 1 million and the population in the country stands at 100 million. The defence can use this figure to its advantage by harping on the fact that since there are 100 million people, there may 100 people who may share the same DNA profile and therefore, the probability that the defendant is the perpetrator is only 1 to 100 or 1 % [31]. Such an argument, however, is only true if all the other 99 people are likely to be as equally guilty as the defendant without taking into account the DNA evidence. For example, in the R v Doheny & Adams case where the random occurrence ratio was calculated as 1 in 27 million, the defence may offer the argument that there may be three persons within the UK who share the same DNA profile as the defendant if the population of the male is about a hundred million. It can also conclude that the probability of the defendant being the actual culprit is therefore, only 1 in three. This is a fallacy, however, because there is no information whatsoever about the two other persons or that the likelihood that they were near the scene of the crime when it happened [32]. 6.0 Conclusion The discovery of DNA and its subsequent use in courts of law as evidence is a major breakthrough not only in criminal investigations, but also in criminal, and civil, trials. Many lives, already rotting in prison houses for crimes they did not commit, were saved because of it and many victims of unsolved crimes are hoping that justice will still be theirs in the long run. Prosecutors and defence lawyers, however, need to be reminded that the presentation of DNA evidence in court needs to be done correctly. Particularly, the so-called fallacies must be avoided and the use of the Bayesian theorem should only be carefully utilised. The cases mentioned in this paper attest to the fact that fallacies and errors in the presentation of DNA evidence and the manner by which a forensic expert is questioned in court in the presence of the judge and the jury is a delicate matter and may lead to a undesirable outcome. End Notes [1] T. Yeatts, Forensics: solving the crime, USA: The Oliver Press, Inc., 2001, p. 113. [2] S. Pena, DNA fingerprinting: state of the science, Birkhauser, 1993, p. 1. [3] A. Semikhodskii, Dealing with DNA evidence: a legal guide, Routledge, 2007, pp. 21-22. [4] Pena, pp. 1-2. [5] Semikhodskii, p. 22. [6] A. A. Kloosterman, Current and Future Developments in Forensic DNA Typing, Nederlands Forensisch Instituut 2003, p 5. [7] Pena, p. 2. [8] R v Doheny and Adams [1996] EWCA Crim 728 [9] ibid. [10] ibid. [11] ibid [12] Ibid [13] ibid [14] ibid [15] R v Deen [1994] Times LR 11. [16] ibid [17] R v Adams [1996] 2 Cr. App.R. 467. [18] ibid [19] ibid [20] C. de Macedo, Guilt by statistical association: revisiting the prosecution’s fallacy and the interrogator’s fallacy, The Journal of Philosophy (2008): 320-32. [21] C. Adam, Essential mathematics and statistics for forensic science, John Wiley and Sons, 2010, pp. 295. [22] P. Hughes, DNA fingerprinting, House of Commons Library, Research Paper 96/44 (1996), p.16. [23] Adam, p. 296. [24] M. Finkelstein, Basic concepts of probability and statistics in the law, Springer, 2009, p. 13. [25] R v Doheny and Adams [1996] EWCA Crim 728. [26] Ibid. [27] D. Howell, Statistical methods for psychology, seventh ed., USA: Cengage Learning, 2009, p. 123. [28] J. Ledward, Understanding forensic evidence: Do lawyers and the judiciary understand forensic evidence and the Bayesian approach? QEB Hollis Whiteman Barristers (2004), pp. 1-2. [29] K. Ritz, L. Dawson & D. Miller, Criminal and environmental soil forensics, Springer, 2008, pp. 25-26. [30] R v Adams [1996] 2 Cr. App.R. 467. [31] D. Balding, Weight-of-evidence for forensic DNA profiles, John Wiley and Sons, 2005, p. 147. [32] A. Semikhodskii, p. 116. Read More