Extracting concealed information from groups

This 2010 study by Meijer, Smulders, and Merckelbach introduced a group-level variant of the skin conductance-based Concealed Information Test (CIT) designed to extract critical concealed information from multiple individuals sharing the same knowledge. Using a mock terrorist attack scenario with 12 participants, the researchers demonstrated that aggregating standardized skin conductance responses across group members — without requiring prior knowledge of the correct answer — could reliably identify all three concealed critical items. The findings established a foundational proof-of-concept for applying the CIT as a collective intelligence tool in counterterrorism and criminal network investigations.

Twelve participants underwent a laboratory-based mock terrorist attack scenario in which they shared the same concealed information; skin conductance responses were recorded, within-person z-score standardized, and analyzed using repeated-measures ANOVA to identify which answer options were collectively recognized as critical across the group.

This research addressed a fundamental gap in forensic psychophysiology by developing a group-based variant of the Concealed Information Test, moving beyond the traditional single-suspect model to address modern security threats involving groups like terrorist networks. Twelve university participants were given shared information about a mock terrorist attack and tested using skin conductance measurements. The key innovation was aggregating individual physiological responses across all group members rather than analyzing each person separately. All three concealed critical items embedded in the scenario were successfully detected through this group-aggregated approach, demonstrating the potential for extracting information when multiple individuals share the same concealed knowledge.

This research opens new possibilities for polygraph applications in cases involving criminal conspiracies or terrorist groups where multiple suspects may share concealed information, potentially expanding the utility of CIT beyond traditional single-suspect scenarios.

Background & Context

The standard Concealed Information Test (CIT) — sometimes called the Guilty Knowledge Test — was developed as a psychophysiological method for determining whether an individual harbors specific crime-relevant memories. The CIT presents a series of multiple-choice questions while measuring psychophysiological or behavioral responses, and when a suspect consistently shows distinct responses to critical crime-related items compared to neutral control items, concealed knowledge is inferred. For decades, the test's design presumed a single suspect in isolation.

Lie detection procedures are typically aimed at determining guilt or innocence of a single suspect, yet serious security threats often involve groups, such as terrorist networks or criminal organizations. This fundamental mismatch between conventional CIT design and the realities of modern criminal conspiracies created an important gap in forensic psychophysiology — one that this study was designed to address.

An additional challenge motivating this work relates to the "searching" or open-ended variant of the CIT. Known as the Searching Peak of Tension Test (SPOT), this variant can be employed when the correct alternative is not known to investigative authorities but is the topic of investigation — a series of answer options is presented, and the option evoking the largest physiological response is considered to warrant further investigation, potentially revealing information such as the location of a body. Extending this logic to the group context was the central innovation of this research.

Research Design & Methodology

The study describes a variant of the skin conductance-based Concealed Information Test designed to allow for the extraction of critical information from groups. Twelve participants were given information about an upcoming mock terrorist attack, with specific instructions not to reveal this information to anyone. Participants were recruited at Maastricht University and each was briefed individually via a role-playing scenario.

Upon arrival in the laboratory, each participant was told that the experiment entailed role playing and that he had to pretend he was a member of a terrorist organization. The modified CIT was then used to test whether the location of the planned attack could be revealed from the pooled physiological data. The core innovation was aggregating individual physiological responses across all group members rather than evaluating each person individually.

Key design elements included:

• Physiological measure: Skin conductance responses (SCRs) were recorded using two Beckmann silver/silver-chloride (Ag-AgCl) electrodes placed on the medial phalanges of participants' nondominant hands, filled with isotonic electrode paste (0.9% NaCl), with SCRs computer-scored as the maximum positive deflection between 1 and 5 seconds following stimulus onset.
• Standardization: Within-question standardized scores were computed by subtracting the mean of all five responses (excluding a buffer option) from the response to each answer option and dividing by the standard deviation of all five responses.
• Statistical analysis: Three separate univariate repeated-measures ANOVAs were carried out on these z-scores, one for each question.
• Critical items: Three concealed items of information were embedded in the mock scenario, each tested across multiple answer options.
• Sample size: N = 12 participants, all sharing the same concealed information about the mock attack.

Results & Key Findings

All three concealed critical items were successfully detected using the group-aggregated physiological approach. Using skin resistance responses (SCRs) as the sole physiological measure and relying upon group averages, the researchers reported that all three critical items of information were correctly detected. This represented a 100% item-detection rate across the three embedded concealed questions.

Key findings from the study include:

• Group aggregation works: Pooling SCR data across 12 participants reliably surfaced the critical answer options, even without any single individual being assessed in isolation.
• Standardization was essential: Within-person z-score normalization removed individual baseline differences in skin conductance reactivity, making group-level comparison statistically valid.
• ANOVA sensitivity: Repeated-measures ANOVAs on the standardized z-scores successfully identified critical items as statistically distinct from control items across all three questions.
• Information extraction without prior knowledge: Unlike the classic CIT, investigators did not need to know the correct answer in advance — the group's collective physiological responses revealed it.
• Mock terrorism scenario ecological validity: The forensically realistic cover story (terrorist attack planning) enhanced the motivational stakes of participants concealing information.

The study's outcome is particularly striking given the small sample of only 12 participants. Prior attempts to examine group CIT procedures did not include a searching test for partial knowledge, and subsequent work building on this study has shown that the Group SCIT is quite accurate in reconstructing a comprehensive description of a future felony. The Meijer, Smulders, and Merckelbach paper thus served as a foundational proof-of-concept for an entirely new paradigm in forensic psychophysiology.

Discussion & Significance

The significance of this study lies in its reconceptualization of the CIT as a group-level intelligence tool, not merely an individual guilt-detection instrument. Traditional polygraph and CIT examinations assume sequential, isolated testing of one suspect at a time. This study demonstrated that if multiple individuals within a network share the same concealed information, their collective physiological data can be aggregated to expose that information — effectively turning the group itself into the "unit of analysis."

Building on the Meijer, Smulders, and Merckelbach (2010) framework, subsequent researchers have recognized that the group CIT provides an opportunity to detect and reveal information not yet known to investigators, which could be used to prevent malicious intents or find a hostage. This extends the practical scope of psychophysiological lie detection far beyond its historical use in post-hoc criminal investigations into proactive threat-prevention scenarios.

The study also makes a meaningful methodological contribution by demonstrating that within-individual standardization of SCRs is sufficient to enable cross-person aggregation. This validates a relatively simple statistical approach — z-score normalization followed by ANOVA — as an accessible framework that does not require sophisticated machine learning or complex signal-processing pipelines. For a field sometimes criticized for opaque or inconsistent scoring methods, this transparency is a notable strength.

Limitations & Considerations

While the findings are compelling, several important limitations should be acknowledged:

• Very small sample (N = 12): The proof-of-concept nature of the study means results must be replicated with larger groups before strong inferential conclusions can be drawn.
• Laboratory mock scenario: Participants role-played as terrorist operatives rather than being genuine criminal co-conspirators. The motivational stakes, stress levels, and countermeasure use in real-world scenarios may differ substantially.
• Uniform knowledge assumption: All 12 participants shared the same concealed information. Real criminal networks typically involve partial or compartmentalized knowledge — a more complex scenario not fully addressed here.
• Single physiological channel: Relying solely on skin conductance responses limits the robustness of detection; subsequent research has incorporated respiration, heart rate, and EEG measures for multi-channel approaches.
• No innocent control group: The study did not include participants who lacked the concealed information, limiting conclusions about false-positive rates at the group level.

Although available data confirm the validity of the CIT broadly, there is quite a gap between laboratory studies and realistic criminal investigations — a gap that group-level CIT research is beginning to bridge, but has not yet fully closed. Future work incorporating partial-knowledge designs and real-world field conditions will be critical to establishing the operational utility of this approach.

Practical Applications

This study opens a genuinely novel operational avenue for security and law enforcement agencies. In counterterrorism contexts, investigators may not have a single suspect but rather a network of individuals with shared operational knowledge. Serious security threats often involve groups such as terrorist networks or criminal organizations, and the group CIT variant described here allows for the extraction of critical information from such groups. Rather than requiring individual examinee decisions, investigators could screen a group and use the collective physiological signal to identify what the group knows — such as the location of a planned attack, the identity of a ringleader, or the timing of an operation.

For forensic examiners and policymakers, the key takeaway is that the Concealed Information Test is a flexible paradigm that can be adapted beyond its traditional one-examiner-one-examinee structure. The within-person standardization method described here is practically implementable with existing polygraph hardware and basic statistical software. As the study's own keyword framing connects to broader domains of forensic science, terrorism, lie detection, polygraph, concealed information test, and criminal networks — underscoring that this research sits at the intersection of psychophysiology, national security, and forensic intelligence gathering. Agencies investing in screening technology should consider this group-aggregation framework as a complement to traditional individual testing protocols.