| Summary: | 碩士 === 國立交通大學 === 科技法律研究所 === 107 === Governmental misconduct, including coercive interrogation, suggestive eyewitness identification, and biased forensic science evidence, is regarded as one of the leading causes of wrongful convictions. Oftentimes these errors are honest mistakes, however, from investigators affected by “tunnel vision” as all human beings. Under tunnel vision, investigators would search for information consistent with their own hypotheses, overlook facts that point to alternative suspects, and focus on the suspect even if contradictory evidence exists.
While tunnel vision is a natural phenomenon of human beings, it can be reinforced by rules and institutional pressures. First, in terms of questioning of suspects, the accusatorial model of interrogation, which encourages interrogators to use psychological tools to obtain a confession once they spot signals of deception from a
suspect, can reinforce tunnel vision and lead to false confessions. Likewise, suggestive eyewitness identification procedures can lead witnesses to misidentify innocent people. In such situation, for instance, if a lineup is biased against a suspect, a witness may select whoever stands out in the lineup. Last but not least, forensic scientists may identify themselves as scientific crime fighters, be biased by contextual information, and thus interpret results of experiment favoring the prosecution.
To reduce negative consequences that occur during tunnel vision, legislative reforms are required. The thesis proposes “investigative interview” as the best practice
in questing of suspects; investigators should aim at “searching of information” rather than “obtaining a confession”, and treat suspects with trust and respect. For eyewitness identification, the best practice includes neutral eyewitness identification instructions,
implementation of double-blind procedure, and prohibition of repeating identification procedures. To ensure the quality of forensic evidence, Standard Operating Procedures in laboratories, for instance blind testing and Linear Sequential Unmasking, are recommended.
Chapter 1 outlines objectives and research methods of the thesis. Chapter 2 discusses the causes of tunnel vision in three parts. Chapter 3 then compares between accusatorial model of interrogation and investigative interview. In chapter 4, how suggestive eyewitness identification procedures may influence eyewitness
identification decisions are discussed. Chapter 5 analyzes two possible factors that affect reliability and validity of forensic science evidence: role effect and contextual information. Chapter 6 is the conclusion of the thesis.
|
|---|
