Leakage of relevant information to innocent examinees in the GKT: An attempt to reduce false-positive outcomes by introducing target stimuli.

Abstract

This study focused on the Guilty Knowledge Test (GKT) -a psychophysiological detection method based on a series of multiple-choice questions, each having one relevant and several neutral (control) alternatives. The study examined a new method, designed to reduce false-positive outcomes, due to leakage of relevant items to innocent suspects, by introducing target items (i.e. items known to all examinees, but unrelated to the crime), to which subjects have to respond (e.g., by pressing a key), while answering the GKT questions. Informed innocents showed relatively larger electrodermal responses to the critical items than uninformed participants, but not as large as the responses made by guilty participants. No differences between informed and uninformed innocents were obtained with a respiration measure. The use of the target items tended to reduce the differences between informed and uninformed innocents. The results further demonstrated that electrodermal responding to the relevant items was correlated with memory of these items.The Guilty Knowledge Test (GKT) is a method of psychophysiological detection that has been extensively researched and discussed in the literature (e.g., Ben-Shakhar, 1991;Furedy & Heslegrave, 1991;Lykken, 1974Lykken, , 1981. It utilizes a series of multiple-choice questions, each having one relevant alternative (e.g., a feature of the crime under investigation) and several neutral (control) alternatives, chosen so that an innocent suspect would not be able to discriminate them from the relevant alternative (Lykken, 1981). Typically, if the subject's physiological responses to the relevant alternative are consistently larger than the responses to all other alternatives, knowledge about the event (e.g., crime) is inferred.This method protects innocent suspects because, as long as information about the event has not been leaked out, the relevant and control questions should be equivalent from their perspective. Under these conditions, the probability that an innocent suspect would show consistently larger responsivity to the relevant alternative than to the neutral ones depends only on the number of questions and on the number of alternative answers per question, and this probability can be controlled such that maximal protection for the innocent is provided. Indeed, a relatively large number of studies, which were conducted to assess the accuracy of the GKT, reveal that the rate of false-positive errors (innocent suspects classified as guilty) is around 5% (for a review of these studies, see, Ben-Shakhar & Furedy, 1990;Elaad, 1998).However, a successful implementation of the GKT depends on the validity of the assumptions underlying it, and in particular on the assumption that the relevant information is not known to innocent suspects. This seems crucial because it has been

Successful detection of verbal and visual concealed knowledge using an RT‐based paradigm

“…While a CKT constructed from these materials can successfully indicate whether the unauthorised examinee has viewed privileged information, it may not reveal whether malicious intent was involved. While no existing test can accomplish this, previous research using polygraph measures have shown some success in distinguishing those who are 'guilty' from those who are 'innocent but aware' of crime details when familiarity with actions, rather than knowledge is stressed (BenShakhar, Gronau, & Elaad, 1999;Bradley, MacLaren, & Carle, 1996;Bradley & Warfield, 1984). Although a promising approach, in some cases researchers report a significantly high false-alarm rate for innocent participants who are aware of crime information (e.g.…”

Estimating the Validity of the Guilty Knowledge Test From Simulated Experiments: The External Validity of Mock Crime Studies.

“…This study focused on the external validity of the mock-crime paradigm, which has been used extensively in the past 3 decades to evaluate the validity of the GKT (e.g., Ben-Shakhar & Dolev, 1996;Ben-Shakhar et al, 1999;Bradley, MacLaren, & Carle, 1996;Bradley & Rettinger, 1992;Davidson, 1968;Lykken, 1959). We showed that the standard mock-crime procedure, applied in most of these studies, may have weak external validity because it does not tap several factors, which operate in the realistic situation and may reduce memory of some relevant items.…”

“…A person (the mannequin), whom you know as Frank, is seated in this room", see Bradley & Warfield, 1984, p. 684). In addition, in many of the mock-crime studies the guilty participants are presented with these details just before the GKT is administered (e.g., Ben-Shakhar, Gronau, & Elaad, 1999), and in some studies, data from participants who could not recall the relevant items in a postexperiment recall test were discarded (e.g., Ben-Shakhar & Gati, 1987). In this study, we manipulated this factor by using the typical mock-crime procedure for half our participants (the standard condition), while using a more realistic procedure for the other participants (the realistic condition).…”

Detecting concealed information with reaction times: Validity and comparison with the polygraph

“…Participants were requested to name the catch number instead of answering 'no'. Inclusion of catch items has been shown to leave detection efficiency in 'guilty' individuals unaffected (Ben-Shakhar, Gronau, & Elaad, 1999).…”