
Below-chance performance on forced-choice performance validity tests (PVTs) is typically interpreted against a binomial model that assumes a .50 probability under uninformed responding. We tested whether this assumption holds for the Word Memory Test (WMT) and whether item-level lexical properties account for departures from the binomial expectation using a sham extrasensory perception (ESP) paradigm to elicit uninformed (i.e., “random”) responding. Fifty-five undergraduates completed the WMT Immediate Recognition (IR) trial without prior exposure to the study list. Total scores and item-level responses were tested against exact binomial expectations. Pair-level lexical contrasts (target minus foil) for log word frequency, age of acquisition, concreteness, and orthographic distance were derived from the English Lexicon Project and examined as predictors of item selection in ordinary least squares (OLS) and generalized estimating equation (GEE) models. Mean IR performance was 24.15 of 40 (60.4
Generative artificial intelligence (AI) has created a new source of information, coaching, and real-time assistance for examinees in high-stakes psychological and neuropsychological assessment. This conceptual review integrates established research on coaching and test exposure with emerging evidence on AI-assisted feigning, performance-validity test security, and high-stakes response manipulation. Available findings are heterogeneous: sophisticated AI-generated coaching has sometimes moderated symptom overreporting and cognitive underperformance, whereas less structured AI use has produced more obvious exaggeration and remained detectable by established validity measures. Prompt-based studies also show that chatbots can generate information that poses a variable threat to test security. We propose a five-mode taxonomy of AI involvement - information seeking and symptom organization, narrative construction, strategic response shaping, test-specific preparation or exposure, and unauthorized real-time assistance - to organize their potential evidentiary consequences. Within this taxonomy, the term AI-assisted response distortion (AARD) is reserved for AI use that materially alters, strategically shapes, or contaminates the response process; benign information seeking or organization does not constitute response distortion in itself. The taxonomy is not a new validity model or clinical algorithm. Current evidence does not support AI-specific performance- or symptom-validity test cutoffs, a distinct malingering construct, or routine adverse inference from disclosed AI use. The clearest implications remain measure-specific and grounded in established validity principles. The article concludes with a research agenda focused on coaching potency, prompt and model effects, test-security exposure, real-time assistance, positive impression management, warnings, fairness, and real-world forensic validation.
Mental health experts conducting civil forensic assessments are frequently asked to determine whether, and to what extent, an event at issue contributed to psychological injury and functional impairment. These determinations are particularly challenging in historical childhood abuse claims, which provide a useful exemplar of complex causal assessment because evaluators often work with incomplete or discrepant records while considering multiple alleged harms and competing contributors. This conceptual and practice-oriented paper proposes Structured Causation Analysis (SCA), a harm-specific approach designed to make causal reasoning more systematic, transparent, and reviewable. The approach distinguishes between whether a claimed harm is established and whether case-specific evidence supports attributing that harm to the event at issue, and it systematically addresses evidentiary validity and interpretability, alternative contributors, corroborating and disconfirming information, and the strength and specificity of causal attribution. We introduce the Causation Analysis Matrix (CAM) as a practical tool for implementing SCA. By requiring explicit consideration of plausible competing contributors and evidence that may weaken an initially favoured hypothesis, SCA also incorporates bias-mitigation principles into causal reasoning. A de-identified and modified historical childhood sexual abuse case illustrates the approach.
In legal contexts, psychiatric and psychological assessment decisions can have substantial consequences, so mental-health symptoms reported by evaluees must be assessed for credibility rather than taken at face value. To this end, the Inventory of Problems–29 (IOP-29) is particularly relevant because it was specifically developed to assess the credibility of self-reported symptoms and has been investigated in numerous language versions. Because no study had yet addressed the Russian-language IOP-29, this study replicated Akca et al.’s (2023) research paradigm in a Kazakhstan community sample. Overall, 154 volunteers completed the IOP-29 three times, yielding 462 administrations under honest responding, random responding, and instructed feigning conditions; the latter involved feigned schizophrenia, depression, or PTSD. The False Disorder Probability Score (FDS) discriminated feigned from honest responding, with specificity = 0.92 and sensitivity = 0.81 at the standard cutoff (FDS ≥ 0.50). The Random Responding Scale (RRS) differentiated random from non-random responding at the mean-score level, but classification accuracy was mixed. These findings provide support for the validity of the Russian IOP-29 as a measure of the credibility of reported mental health symptoms, whereas the RRS requires further study before it can be used routinely in applied settings.
Invalid responding can bias clinical judgement, particularly in adult attention-deficit/hyperactivity disorder (ADHD), where rates of symptom overreporting and cognitive underperformance are elevated. Performance validity tests (PVTs) and symptom validity tests (SVTs) are commonly used to assess response credibility, yet their relevance for long-term clinical outcomes in ADHD remains unclear. The present study examined whether PVT and SVT scores obtained during diagnostic assessment are associated with long-term clinical outcomes in adults with ADHD. Data were drawn from an outpatient cohort of adults diagnosed with ADHD (N = 181). Participants completed standardised diagnostic assessment including two memory-based PVTs, one attention-based PVT, and two embedded SVTs. Long-term clinical outcomes, including psychotherapy use, pharmacotherapy, and diagnostic re-evaluation, were assessed via self-report approximately one year later. Associations were examined using correlational analyses. No significant associations were observed between PVT and SVT scores and long-term clinical outcomes. In contrast, baseline ADHD symptom ratings showed small and domain-specific associations with treatment-related outcomes. These findings suggest that validity indicators may be more relevant to proximal clinical processes than to the prediction of longer-term treatment outcomes in adult ADHD. Future research should examine how validity indicators influence decision-making and treatment processes in adult ADHD assessment, using longitudinal and multi-method designs with more sensitive outcome measures.