Validity of the concealed information test in realistic contexts

Overview: Earlier field studies on the Concealed Information Test (CIT) are reviewed from a new perspective. Limitations of the studies as well as factors that account for the relatively large false negative rates are discussed. Two types of CIT practice, the pre-interrogation practice, which is mainly used in Japan, and the interrogation practice are identified and discussed. The advantages of each type of practice are described and suggestions for improvement are made.

Critical review and reanalysis of two foundational Israeli field validity studies using confession as ground truth criterion: 1990 study (n=98 suspects, skin resistance responses) and 1992 study (n=80 suspects, respiration and skin resistance measures). Both employed suboptimal CIT implementation with minimal questions.

Elaad's 2011 chapter addresses the critical gap between exceptional laboratory validity of the CIT (effect size 3.12, ROC 0.95) and unknown field performance by reanalyzing Israeli Police field data. Two rigorous field studies revealed asymmetrical error patterns: false-positive rates remained laboratory-like (2-5%) but false-negatives substantially increased (20-42%), attributed to suboptimal implementation rather than fundamental technique weaknesses. Using optimal decision rules, guilty detection improved to 65-75.8% while maintaining 94% innocent classification accuracy. The chapter also examines international variations in CIT practice, comparing Japanese pre-interrogation protocols with Western interrogation-based approaches.

Field validity concerns for the CIT appear addressable through proper implementation protocols, suggesting the technique can serve as a protective tool for innocent suspects in criminal investigations. Results highlight the critical importance of field research and optimal procedural conditions for forensic application of CIT technology.

Background & Context

The Concealed Information Test has demonstrated exceptional validity in controlled laboratory settings, with meta-analyses showing effect sizes of 3.12 and ROC areas of 0.95 under optimal conditions. However, a critical gap existed in understanding how the CIT performs in the messy reality of criminal investigations. Field conditions typically differ drastically from experimental settings, making field studies absolutely essential for a comprehensive picture of forensic application validity.

Elaad's 2011 chapter addresses this crucial knowledge gap by reexamining earlier field research from a fresh perspective. The work is particularly significant because only two field-validity studies had been published at that time (Elaad, 1990; Elaad et al., 1992), a situation attributed to inherent difficulties in conducting proper field research. The chapter also explores international variations in CIT practice, contrasting approaches used in different jurisdictions.

Research Design & Methodology

This chapter is a critical review and synthesis rather than an empirical study. Elaad systematically reanalyzes data from the only available field validity studies—his own Israeli Police research from 1990 and 1992. The methodology of these foundational field studies was distinctive and rigorous:

• The 1990 study analyzed skin resistance responses from 50 innocent and 48 guilty suspects, with ground truth established by confession
• The 1992 study compared respiration line length and skin resistance response in 40 innocent and 40 guilty suspects, also using confession as the criterion
• Both studies used very few CIT questions (mean of 2 and 1.8 questions respectively)—far below recommended levels
• Tests were administered immediately after a Comparison Question Test, which may have attenuated sensitivity

The chapter also examines two distinct types of CIT practice: pre-interrogation practice (mainly used in Japan) and interrogation practice, analyzing the advantages of each approach.

Results & Key Findings

The field validity data revealed a striking asymmetry between error types. False-positive error rates remained as low as laboratory studies (2% in the 1990 study and 5% in the 1992 study), but false-negative rates were much larger (42% in 1990 and 20% in 1992).

Specific accuracy findings include:

• Using a priori decision rules, 98% of innocent suspects and 42% of guilty suspects were correctly classified in the 1990 study
• Optimal decision rules improved guilty detection to 65% while maintaining 94% innocent classification accuracy
• In the 1992 study, combining skin resistance and respiration measures improved guilty detection to 75.8% while maintaining 94.1% innocent detection

Despite the elevated false-negative rates, results suggested the CIT could be a useful investigative tool that protects innocent suspects, with false negatives attributed to realistic field conditions. The chapter identifies factors accounting for these relatively large false-negative rates and discusses limitations of previous field research.

Discussion & Significance

The critical insight from this review is that external validity concerns are real but not insurmountable. The CIT's use in the Israeli criminal cases studied was not optimal, particularly regarding the low number of questions—much smaller than recommended. This suggests that apparent limitations in field validity may reflect suboptimal implementation rather than fundamental weaknesses in the technique.

The chapter's comparative analysis of Japanese and Western CIT practices reveals important procedural variations. Japanese field practice focuses on revealing examinees' recognition of specific crime details rather than integrating responses across questions to determine guilt or innocence. This question-focused approach addresses practical challenges when sufficient unpublicized crime details are unavailable.

Encouragingly, the small false-positive rates in both field studies indicated that critical information leakage—a major validity threat—was not problematic in these Israeli cases. This finding has important implications for maintaining CIT integrity in field applications.

Limitations & Considerations

Several methodological constraints affected the reviewed field studies:

• Ground truth criteria were unavailable, requiring reliance on confessions which may depend on test outcomes—a problematic circularity
• The studies used suboptimal test parameters (too few questions, administration after CQT)
• Even in optimal studies, sensitivity (65-76% for guilty suspects) remained lower than specificity (94% for innocent suspects), with significant numbers of guilty suspects remaining undetected
• Sample sizes were modest, limiting generalizability

The chapter acknowledges that difficulties in conducting proper field studies may explain why so few have been published, particularly regarding ground truth establishment and the need for Japanese criminal investigation settings where CIT is standard practice.

Practical Applications

The review offers concrete guidance for improving field CIT validity. Examiners should use multiple CIT questions with repetition, combine multiple physiological measures, and modify investigation procedures so polygraph examiners can inspect crime scenes soon after crimes occur—as practiced by Japanese National Police. This early involvement ensures identification of salient crime features before memory fades or information leaks.

For practitioners, the asymmetric error pattern has profound implications: the CIT's strength lies in protecting the innocent (low false positives) while its weakness is missing some guilty individuals (higher false negatives). This error profile makes the CIT particularly valuable as an investigative tool to exonerate innocent suspects rather than as definitive proof of guilt. The chapter provides a roadmap for closing the laboratory-field validity gap through improved methodology and cross-national learning from successful implementations like Japan's systematic program.