The effects of using a scripted or unscripted interview in forensic interviews with interpreters
The current study set out to replicate and expand the results of a study by Pruss (2007) which found that information is lost when interviews are conducted through interpreters. In the present study, Viewers (i.e., mock eyewitnesses) fluent in Spanish watched a video of a burglary and then were interviewed in Spanish about what they had seen. Half of Viewers were randomly assigned to be interviewed by an English-speaking Interviewer through a bilingual Interpreter (Interpreter condition), and the other half were randomly assigned to be interviewed directly by a Spanish speaking Interviewer with no Interpreter (No Interpreter condition). Within each of these two conditions, half of interviews were conducted following a script of open-ended questions (Scripted condition), and the other half were conducted without a script (Unscripted Condition). The total number of groups was 120 and the total number of participants was 300. Interviews were coded for the amount of information transmitted during the interview. Consistent with the findings of Pruss (2007), the present study found that Interviewers extracted more information when they questioned Viewers directly than when they questioned Viewers through an Interpreter. The open-ended script significantly increased the yield of information when Interviewers questioned Viewers directly, but not when Interpreters were used. More generally, the results of this study support the use of open-ended questions in investigative interviews and the practice of tape recording such interviews.
Pruss, Nicole, "The effects of using a scripted or unscripted interview in forensic interviews with interpreters" (2008). ETD Collection for University of Texas, El Paso. AAI3341651.