No effect of moderate alcohol intake on the detection of concealed identity information in the laboratory

This laboratory study examined whether moderate alcohol consumption affects the ability to detect concealed identity information using the Concealed Information Test. The research investigated alcohol as a potential factor that might impair CIT detection accuracy.

Laboratory experiment using the Concealed Information Test to examine the effects of moderate alcohol consumption on concealed information detection performance.

This laboratory study examined whether moderate alcohol intake could serve as a countermeasure against the Concealed Information Test (CIT) by reducing detection accuracy for concealed identity information. Participants were administered alcohol or control substances and then underwent CIT testing. The results demonstrated that moderate alcohol consumption did not significantly impair the CIT's ability to detect concealed information. This finding suggests that alcohol consumption does not provide an effective means of defeating polygraph testing using the CIT format. The research contributes to understanding the robustness of the CIT against potential countermeasures.

These findings suggest that moderate alcohol consumption should not be considered a significant threat to CIT validity in real-world applications, though examiners should still be aware of substance use by examinees.

Background & Context

The relationship between alcohol consumption and deception detection has puzzled polygraph researchers for decades. Alcohol's effects on cognitive control and executive function suggest it might either impair or enhance the detectability of concealed information, making this a critical question for polygraph validity. The Concealed Information Test (CIT) detects concealed knowledge by measuring responses to crime-relevant information, and previous field research suggested alcohol intoxication might actually increase these detection effects.

This preregistered laboratory study aimed to investigate whether previously observed effects of alcohol intoxication on reaction time CIT (RT-CIT) performance in a field study would translate to a controlled laboratory environment. The research directly challenges the validity of earlier findings from a 2018 field study by the same research team, which found that higher blood alcohol levels correlated with larger CIT effects—a counterintuitive result that required rigorous experimental follow-up.

Research Design & Methodology

A total of 248 participants were recruited, with 62 excluded due to strict pre-registered inclusion criteria related to alcohol administration, and 9 additional participants excluded due to experimenter error or high error rates, resulting in a final sample of 177 participants (mean age 24.21 years, 112 female, 65 male). Participants were randomly assigned to either an alcohol group (n=88; receiving a drink with 3cl alcohol) or a sober control group (n=89; receiving a drink with just alcohol drops to mask group assignment). This double-blind design represented a crucial methodological improvement over the earlier field study.

After drink administration, participants completed an RT-CIT in which they were instructed to hide knowledge of their own identity. The test used a "secret agent" cover story where participants had to conceal their real identity information (name, birthday, birthplace, parents' names) while responding to a new "cover" identity. Blood alcohol concentration (BAC) was estimated via breath alcohol ratio using the Dräger Alcotest 3000. Key design features included:

• Power analysis determined 79 participants needed for 95% power to detect r=0.39 (the field study effect); the final sample of 182 provided power of 0.9998
• Trait impulsivity (BIS-11) and psychopathic traits (PPI-R) were assessed, along with self-reported lying frequency
• Testing occurred 25-40 minutes post-consumption to allow BAC to rise adequately; the CIT closely resembled the 2018 field study protocol

Results & Key Findings

In contrast to the previous field study, results revealed no differences in CIT-performance between intoxicated and sober participants. This null finding directly contradicted the earlier field research and represented a critical replication failure under controlled conditions.

Detailed statistical analyses revealed:

• Correlations between BAC and CIT-effects showed only small and non-significant relationships for reaction times (r=0.08, p=0.32) and error rates (r=0.01, p=0.95)
• Bayesian independent samples t-tests comparing CIT-effects in error rates and reaction times between the Alcohol and Control groups were performed to quantify evidence for the null hypothesis
• The large number of participants with a BAC of zero highlighted limitations of the correlative approach used in the field study
• No significant correlations were found between CIT-effects and measures of tension, boredom, fatigue, intoxication, test difficulty, impulsivity, substance use, or lying frequency

Discussion & Significance

The sample size—more than twice that required to detect an effect as large as the field study with 95% power—indicates the observed null-effect most likely reflects a genuine absence of such an effect, at least under laboratory conditions. This finding has profound implications for understanding the robustness of CIT testing across diverse testing circumstances and challenges simplistic assumptions about alcohol as a countermeasure.

Several theoretically interesting explanations for the discrepancy between the current results and the 2018 field study are possible, including the possibility that effects in the smaller field study (n=42) were due to chance. The randomized assignment in the laboratory study controls for potentially confounding effects of habitual alcohol use, though small non-significant correlations between habitual alcohol use and CIT-effects suggest this doesn't fully explain the diverging results. The research highlights how field settings may introduce uncontrolled variables—including self-selection biases, drinking motivations, and social context—that can produce spurious effects not replicable under experimental control.

Limitations & Considerations

Previous research by Bradley and Ainsworth found that alcohol intoxication during a mock crime reduced CIT effectiveness, likely due to poorer encoding, though another study failed to replicate this effect. The current study examined alcohol consumed before testing, not during the event-to-be-concealed, limiting generalizability to real-world scenarios where alcohol might be consumed during crime commission. Additionally, the moderate alcohol dose (3cl) may not reflect the higher intoxication levels seen in some forensic contexts.

The laboratory setting, while providing experimental control, sacrifices ecological validity. The consequences of detection failure were minimal compared to real criminal investigations. The university student sample may not generalize to forensic populations with different baseline cognitive abilities, substance use patterns, or motivations to conceal information.

Practical Applications

These findings provide reassuring evidence that moderate alcohol consumption by examinees does not substantially compromise RT-CIT validity. Examiners conducting field tests need not be overly concerned that examinees who have consumed moderate amounts of alcohol will evade detection. However, this should not lead to complacency—breath alcohol testing remains advisable as part of pre-test procedures, and cases involving heavy intoxication warrant careful consideration of whether to proceed with testing.

The study also demonstrates the critical importance of replication, particularly when field findings contradict theoretical expectations. The discrepancy between the 2018 field study and this controlled experiment underscores that correlational field research, despite its ecological validity, requires experimental confirmation before informing policy or practice. For polygraph research more broadly, these findings suggest that CIT robustness against potential countermeasures may be greater than previously feared, though continued vigilance and research remain essential.