In recent years, people who had been convicted due to errors in the process of investigation and forensic sci- ence are acquitted in the retrial. In these cases, it appears that forensic science, which should be rational and neutral in nature, was distorted in various ways by psychological and/or cognitive biases and consequently be- came one of the factors that led to false charges. In April 2016, “Innocence Project Japan,” was launched, which is based on “Innocence Project,” an American organization committed to providing relief to people who are wrongly convicted. In this symposium, along with researchers and practitioners involved in Inno- cence Project Japan, we discussed the structure in which the process of investigation and forensic science are distorted.

View full abstract

In this paper, discussions on “tunnel vision” are introduced. Tunnel Vision is a phenomenon in which one’s range of cognition narrows. For example, an investigator might focus on a particular person in a criminal case and choose evidence to convict the person. On the other hand, exculpatory evidence might be disregarded or excluded. The bias of the investigator might influence interviews with witnesses, eyewitness identification proceedings, interrogation of suspects, or interactions with informants. Tunnel vision is reinforced by external factors (various institutional pressures) and internal factors (cognitive bias). Not only the police but also prosecutors, defense attorneys, and judges are affected by tunnel vision. In the United States and other coun- tries, there have been studies since the 2000’s on tunnel vision and cognitive bias in the criminal justice sys- tem. This paper will also introduce reform proposals to reduce the effects of tunnel vision.

View full abstract

In this paper, the problem of the forensics of an image from a security camera in a murder case in Kyoto is re- ported. Mr. N was accused of beating and killing the victim for reasons of obscenity (Mr. N was later acquit- ted). A security camera near the site where the dead body was found recorded a person considered to be the victim and a man pushing a bicycle. An expert who gave an opinion of the recorded image was considered as evidence. In his opinion, the image showing the person’s clothes, bicycle, and ear lobes closely resembled Mr. N. Thus, he concluded that the person in the image was Mr. N. When another expert’s opinion was sought, the quality of the image was considered: the number of pixels and the visibility thereof. We concluded that the identity of the person could not be determined from the security camera images; for instance, the image of the earlobe was approximately 2.5 × 2.5 cm per pixel only. In my opinion, it is necessary to verify criteria and expressions of identity judgment in forensic sciences.

View full abstract

In this paper, I will report on the problem of DNA testing and the prejudice involved in the “Ten-mon-kan rape case” in Kagoshima prefecture. The accused was convicted at the first trial because the judges trusted the plaintiff’s statements, semen was found on the plaintiff’s body (although its DNA type could not be identi- fied), and the accused’s DNA type was detected on the plaintiff’s chest. In the appeal, the judges handed down a not guilty verdict and raised various issues with the first conviction. First, the judges in the first trial blindly believed the testimony of a scientist from the forensic science laboratory of the prefectural police de- partment that “sperm existed in the detected semen.” Second, the judges believed that the detected sperm be- longed to the accused because they did not think that the victim would lie in a rape case. Third, the judges lacked knowledge of deposition psychology and fully trusted the plaintiff’s statements. These findings show that there was a problem with trusting the forensic science lab’s appraisal of the case. Drastic improvements are required in DNA-type appraisals at the laboratory, such as verifying DNA-type samples with an indepen- dent medical doctor, preserving any evidence for possible re-appraisal, and reviewing all procedures in the lab, from preservation to appraisal.

View full abstract

Based on my work experience as a judge, I will examine the causes of faulty judgments due to various biases in criminal trials. Criminal judges tend to blindly believe the evidence submitted by public prosecutors. These judges also follow “the principle of discretionary prosecution,” which gives them an impression of guilt before a trial because they believe that cases with factual disputes will not be prosecuted. I argue that there are three types of judges: the bigoted type (type A), the undecided type (type B), and the thoughtful-decisive type (type C). Type A has a tendency not to believe the defendants in a trial, but they trust the investigators. Type B is easily affected by others’ opinions, even though the judge may have a feeling that the accused could be innocent. Type C has a tendency not to be biased, although these judges could still misunderstand the facts because they are affected by certain biases depending on the nature of the case. In a case where a confes- sion is the main evidence, for example, the belief is created that a detailed confession by the accused is evi- dence of the real culprit. In a case where indirect facts are the main evidence, judges think that the accused is the real culprit based only on a feeling that the accused “seems like a criminal.” To prevent such biases, it is necessary to (1) deepen the understanding of the process of generating false confessions, (2) appropriately evaluate any weak evidence in cases where indirect facts are the main evidence, and (3) understand how the influence of interview recordings affects one’s judgment.

View full abstract

From the perspective of the psychology of statements, I examined bias issues such as tunnel vision in relation to the evidence of a criminal trial. Along with the tunnel vision resulting from the narrowing of one’s field of view at the stage of deciding on the evidence, the tunnel vision of the investigator’s predictions that “the sus- pect must be the culprit” at the stage of collecting statements also may distort the statements or result in the creation of false statements. For over 40 years, I have been involved in statement analysis of false accusations. Working on the possibility of both guilt and innocence, I have attempted to eliminate bias as much as possi- ble. The Criminal Procedure Act stipulates that the “accreditation of facts is due to evidence”. Apart from the court’s decision, it is necessary to gather various evidence initially with an open mind. However, investigators are influenced by assumptions, including the assumption that initial predictions work. In addition, when an investigator questions suspects and relevant people, there are cases where the questions are chosen based on those assumptions. Investigators must keep in mind that an investigation is a verification of a hypothesis. In- deed, interrogations without considering the possibility of innocence are likely to produce false confessions. From that viewpoint, the visualization of an interrogation was introduced in order to check and prevent bias. However, if interrogations are recorded so that false confessions occur spontaneously, it is difficult to check their arbitrariness. Rather, visualized interrogations can provide useful data for checking the origin of state- ments.

View full abstract

Biases in forensic science and in investigation systems are considered from the forensic scientist’s perspec- tive. There are risks involved in making an invalid decision based on the threshold level of DNA electro- pherograms, where incorrect decisions might be made as to whether peaks belong to a signal or to noise. Sim- ilarly, there is a risk of making incorrect decisions regarding latent fingerprint analysis of results in the “gray zone” between “identified” and “not identified”; the possibility of a “double standard” in the decision to ex- amine or not also exists. Other biases are found in the investigation headquarters (HQ), which play a landmark role in investigating heinous crimes. The investigation HQ tend to adopt evidence to support criminality when proofs of contra- diction are not sufficiently verified and when effectual evidence is very low. Since biases in the investigation, where detectives only have to arrest the criminal, are considered one of the causes of false charges for inno- cent people, measures to minimize these biases are necessary.

View full abstract

Children sometimes have the opportunity to give testimonies or statements of their own actions. However, research on testimonies by children with disabilities has not been comprehensive. This study aimed to investi- gate the effects of a question-answering session to promote accurate reporting of one’s own actions among children with autism spectrum disorder and an intellectual disability, using a single case design. In the ses- sion, participants were asked to report what they studied or played in one session. The dependent variables were accuracy of reporting, number of topics, and number of words; the independent variable was the ques- tion, “What did you study (or play)?” asked by the listener. The listener repeatedly asked the participants for the correct name of the action just after they had studied or played. After this intervention, they could report the name correctly and two of them were consistent in their reporting. Furthermore, this study suggested the possibility of establishing accurate reporting and stated the limitations in obtaining statements from children with autism spectrum disorder and intellectual disability.

View full abstract

The Self-Administered Interview© (SAI©) is a questionnaire employed to obtain eyewitness evidence. Given that information from eyewitnesses is critical for police investigations and that memories are forgotten and contaminated by post-event information as time passes, interviews with eyewitnesses should be conducted as soon as possible. However, this requires time and human resources; therefore, conducting interviews may of- ten be delayed. The SAI includes instructions to enhance memory recall and thus, enables police officers to obtain eyewitness evidence from multiple eyewitnesses at once. The SAI was originally developed in English and has been translated into several languages so as to study its effect. In the present paper, the SAI is intro- duced and studies that investigate the effects of the SAI are reviewed. The applied use of the SAI in situations other than police investigations is also addressed.

View full abstract

In theoretical criminology, it is argued that the recent heightened punitiveness toward criminals has its root in the “emotionalization of law” and “irrational” thinking. To test this theoretical suggestion, this question- naire-based research attempted to show how variables such as dichotomous thinking, social dominance orien- tation, assumed self-competence, and information processing style affect punitiveness. Our results show that although all these variables are related to punitiveness, rational processing style and dichotomous thinking are stronger predictors of punitiveness than the other variables. This implies that some “irrational” elements are embedded in punitive attitudes, which must be taken into consideration if we wish to understand public opinion on punishment. In addition, social and theoretical implications are discussed.

View full abstract

Information associated with suicide can cause confirmation bias among police officers, misleading them to make wrong determinations. In general, police officers are more likely to have a confirmation bias compared with coroners. This case study aimed to gain insight into the psychological phenomenon of confirmation bias in a death case by false-hanging. Employing the text mining approach, we analyzed the words describing the corpse from information in the published inquest records of the police officer, coroner, and forensic doctor. The result of a chi-square test and residual analysis showed that there was less corpse information in the re- cords of the police officer, who made a wrong determination, than the coroner and the forensic doctor. Fur- thermore, the police officer focused on information related to the instrument used for hanging (cord), neck pressure, and body weight more than the coroner and the forensic doctor.

View full abstract