Guilty on the Go: Uncovering Concealed Information by Assessing Response Preparation Processes in a Go-Nogo-Paradigm

This study developed a novel Go-Nogo paradigm for the Reaction Time-Based Concealed Information Test (RT-CIT) that detects both concealed crime-related knowledge and strategic response preparation processes used to defeat the test. Two experiments with 138 participants demonstrated that the paradigm successfully identified both familiarity effects (slower reactions to crime-related information) and response preparation effects (fewer errors for probes matching prime questions) in guilty participants, offering a potential solution to RT-CIT vulnerability to countermeasures.

Two online experiments (total N = 138) using mock crime scenarios where participants engaged in simulated chats concealing crime information, then completed a Go-Nogo task where crime-related prime questions preceded test items (targets, probes, or irrelevants) that either matched or mismatched the prime content, with reaction times and error rates measured.

Background & Context

The Reaction Time-Based Concealed Information Test (RT-CIT) has emerged as a promising alternative to traditional physiological polygraph measures, offering a cost-effective and accessible method for detecting concealed crime-related knowledge. The RT-CIT detects familiarity with crime-related information by comparing response latencies, providing a valid index of deception. Unlike physiological CIT methods requiring specialized equipment, the RT-CIT can be administered using standard computers.

However, a critical vulnerability has plagued RT-CIT validity: RT-CIT results can be manipulated by preparing innocent-looking responses for crime-related probes. Previous research showed that countermeasures like pressing a finger or imagining a slap when irrelevant items appeared could eliminate RT differences, causing test accuracy to plummet from 91% to 45%. This susceptibility to strategic response preparation threatened the test's real-world applicability.

This study introduces an innovative Go-Nogo paradigm designed to detect not only familiarity with concealed information but also the response preparation processes that guilty examinees use when attempting to defeat the test. By incorporating priming questions that match or mismatch crime-related details, the researchers developed a method to reveal when examinees have pre-prepared innocent responses—a countermeasure detection capability that could significantly enhance RT-CIT resilience.

Research Design & Methodology

Two experiments were conducted with a total sample of N = 138 participants, all tested using online mock crime scenarios. The paradigm represented a departure from traditional RT-CIT design by introducing a priming mechanism to assess response preparation.

Participants in the guilty condition engaged in a simulated chat conversation where they concealed information about a fictitious crime (such as stolen items). Each trial began with a fixation cross (400ms), followed by a crime-related prime question presented for 1000ms, then a test item—either a publicly known target, a probe (crime-related detail), or an irrelevant item. After 1500ms, the word automatically moved toward an on-screen "investigator," and participants had to press the spacebar (go response) before arrival if the answer was not publicly known, allowing measurement of response times when withholding both probes and irrelevant items.

The critical innovation was the manipulation of prime-probe matching: test items either matched the prime question in content or did not, creating a Go-Nogo task structure where participants had to decide whether to press a key if the answer was not publicly known. This design allowed researchers to detect two distinct effects:

• Familiarity effect: Slower reaction times and higher error rates for crime-related probes versus irrelevant items
• Response preparation effect: Reduced errors for probes when preceded by matching prime questions (indicating pre-prepared responses)

Results & Key Findings

The new paradigm successfully detected both familiarity and response preparation processes in guilty participants. Evidence was found for both familiarity (slower reactions toward crime-related than unrelated information) and response preparation effects (less errors for probes matching the prime question) in guilty participants, demonstrating the dual-detection capability of the method.

For guilty suspects, the familiarity effect manifested in increased response latencies and error rates for probes compared to irrelevant items. This replicated the standard RT-CIT effect, confirming that guilty examinees recognized crime-related details and struggled to inhibit truthful responses.

The breakthrough finding concerned response preparation detection: The retrieval effect was reflected in an interaction of prime question and probe type, where automatic activation of prepared go responses for probes by matching questions reduced reaction times and error rates for probes after matching compared to non-matching prime questions. This pattern revealed when guilty participants had pre-prepared innocent responses that were automatically triggered by relevant questions—precisely the countermeasure strategy that undermines traditional RT-CIT.

Key statistical patterns included:

• Slower reaction times for probes versus irrelevant items (standard familiarity effect)
• Significantly fewer errors for probes when preceded by matching versus non-matching prime questions
• Effects specific to guilty participants, with innocent participants showing no systematic patterns

Discussion & Significance

This research represents a meaningful advance in addressing the Achilles' heel of reaction time-based concealed information testing: vulnerability to strategic response preparation. The new paradigm was designed to detect a specific strategy to manipulate test outcomes by guilty participants—preparing innocent-looking responses to probes in order to conceal guilty knowledge, a strategy previously suggested to undermine RT-CIT validity and reliability.

The findings demonstrate that cognitive forensic measures can evolve beyond simple recognition detection to assess the metacognitive processes examinees employ when attempting deception. By measuring not just what examinees know but how they've prepared to conceal that knowledge, the paradigm opens new possibilities for countermeasure-resistant testing. The automatic retrieval of prepared responses when triggered by matching questions creates a detectable signature that guilty examinees cannot easily suppress.

The study builds on theoretical accounts of response preparation and episodic memory binding, where including questions as prime stimuli presented immediately before probes enabled measurement of retrieval processes for prepared responses triggered by respective prime questions. This theoretical foundation suggests the effect taps into automatic memory processes that are difficult to consciously control, potentially making the paradigm more robust than standard RT-CIT against sophisticated countermeasures.

Limitations & Considerations

A major limitation is that the paradigm requires multiple crime information categories (e.g., what was stolen, from whom, using what tool) and two crime-relevant items per category (one as target, one as critical probe)—information not always available in real investigations, limiting practical use. This data requirement is more demanding than standard CIT protocols.

The paradigm was validated only in mock crime contexts, raising questions about generalizability to real criminal investigations where stakes are higher and examinees may be more motivated and sophisticated in their deception strategies. The online administration, while demonstrating feasibility for remote testing, may not capture the psychological pressure of in-person forensic examinations.

Additionally, future development might combine this paradigm's rationale with the two-alternative forced choice task used in standard RT-CIT (measuring RTs for targets) to assess response preparation while maintaining conventional test structure. The current Go-Nogo format represents a substantial departure from established RT-CIT protocols, potentially complicating field adoption.

Practical Applications

For polygraph practitioners and forensic psychologists, this research offers a potential pathway toward countermeasure-resistant concealed information testing when sufficient crime details are available. The paradigm could be particularly valuable in cases involving complex crimes with multiple investigative questions (location, weapon, victim, stolen items), where the priming structure naturally fits the information structure. Law enforcement agencies with access to comprehensive case details could implement this approach to detect both guilty knowledge and deliberate attempts to appear innocent.

For consumers of polygraph services and those evaluating test validity, this study underscores an important reality: no deception detection method is immune to countermeasures, but methodological innovations can address specific vulnerabilities. While previous research showed response deadlines help prevent certain countermeasures, this paradigm takes a complementary approach by actively detecting response preparation rather than merely preventing it. Understanding that modern RT-CIT research is evolving toward countermeasure detection, not just concealed information detection, provides a more realistic framework for evaluating test utility in adversarial contexts.