Scientific Publications

Validation study of the Spanish version of the Psy-Flex (Ruiz et al., 2021 translation), a brief six-item scale measuring the presence of psychological flexibility across the six processes of the ACT Hexaflex, in a sample of 170 healthy individuals. CFA confirmed a unifactorial structure with good fit, and the scale showed good internal consistency (α = 0.839), high one-year temporal stability (ICC = 0.887), and evidence of divergent, discriminant, and criterion validity. The strongest negative correlation was with the AAQ-II (r = -0.617). A network analysis positioned defusion and self-as-context as central processes, with committed action in a peripheral position. The Psy-Flex is confirmed as a brief, robust measure of psychological flexibility.

Experimental study applying a contingency-reversal paradigm to examine rule-governed behavior in four frontier large language models (LLMs): GPT 5.2, Claude Opus 4.5, Grok 4.1 Fast, and Gemini 3 Flash. In a 2×2 factorial design (Rule × Reasoning) comprising 640 sessions, the models played a discrimination-learning task whose contingencies reversed without warning at trial 41. All four LLMs displayed rule-based insensitivity to contingencies—replicating the phenomenon documented in humans: providing a rule accelerated acquisition but reduced adaptation after reversal. The magnitude varied markedly (Claude retained higher sensitivity, 56%; GPT 5.2 and Grok barely adapted, 5-9%), differences the authors interpret in light of alignment procedures (pliance vs. tracking). Extended reasoning had a limited impact. This is the first study to demonstrate contingency insensitivity in LLMs.

Pilot feasibility study using a randomized, nonconcurrent multiple-baseline single-case experimental design (SCED) that evaluated a brief three-session Repetitive Negative Thinking (RNT)-focused ACT protocol, delivered individually via videoconference, in healthcare professionals experiencing burnout. Three participants completed the study. The intervention produced clinically meaningful improvements in burnout symptoms (emotional exhaustion and cynicism), emotional distress, and RNT, with large between-case effect sizes on daily measures (exhaustion g = -1.20; cynicism g = -0.91; emotional symptoms g = -0.89; RNT g = -0.78) and even larger effects on weekly measures. Professional efficacy was the least responsive variable. Findings support the feasibility of brief, process-based, RNT-focused interventions for healthcare professionals experiencing high occupational stress with limited time for traditional treatments.

Multinational longitudinal study (33 countries, n=586) examining psychosocial predictors of COVID-19 vaccination intention at two time points: pre-rollout (April-June 2020) and post-rollout (January-February 2021). Self-efficacy and perceived severity were consistent predictors of vaccination intention at both time points. Perceived susceptibility was negatively associated with intention only before rollout. After vaccine rollout, declines were observed in self-efficacy, prosociality, psychological flexibility, and positive affect. General psychological variables did not significantly predict vaccination intention.

Experimental study with 90 Colombian undergraduates selected for high generalized pliance or high generalized tracking. Participants completed a matching-to-sample task under two instruction conditions: General Instructions (GI) and General Instructions plus Responding Criteria (GI+RC). Results confirmed the rule-based insensitivity effect: participants with explicit responding criteria showed lower contingency sensitivity. About one-third of GI participants falsely believed they had received explicit criteria, behaving similarly to the instructed group. No differences in contingency sensitivity were found between high pliance and high tracking participants, but high pliance participants reported greater motivation to follow instructions and earn points, suggesting pliance may have competing effects on insensitivity to contingencies.

In older adults with MDD, we evaluated whether adding 4 ACT sessions to TBS (theta-burst rTMS) accelerates clinical response. In a double-blind, controlled RCT, ACT+TBS combination reduced depression (HDRS) faster at week 6 and showed higher remission than control with support sessions; differences did not persist at weeks 8–12. Integrating ACT could accelerate early TBS benefits in geriatric depression.

Multinational cross-sectional study using network analysis (Gaussian graphical models) to examine the structure of negative automatic thoughts measured by the ATQ-8 in 3,964 participants from six countries (Saudi Arabia, China, Spain, Georgia, Turkey, and Peru). CFA confirmed the unidimensional structure across all samples. The item 'I am worthless' (ATQ4) showed the highest strength centrality overall and in most countries, acting as a cognitive hub reinforcing other negative thoughts. Gender comparisons revealed stronger interconnections among automatic thoughts in women than men, particularly between worthlessness and future hopelessness, consistent with higher depression rates in women. The overall network structure was largely consistent across countries, with some cross-cultural variation in node centrality.

Consensus document (2.5 years) proposing an agenda for contextual behavioral science. It defines that research should be multilevel, process-based, multidimensional, prosocial, and pragmatic, and presents 33 recommendations on design, measurement, and dissemination (e.g., idiographic analysis, high temporal density measures, adaptive trials, and open science). The goal is to maximize the applied and scientific impact of CBS on public health and social problems.

Correlational study with Brazilian adults (n=225) examining the relationship between repetitive negative thinking (RNT), psychological inflexibility (PI), cognitive fusion (CF), and psychological distress. RNT was strongly associated with PI and CF. Mediation showed that CF explains a substantial part of the PI→RNT link, raising the explained variance from ~64% to ~74%. Correlations with depression, anxiety, and stress were weaker. Findings support ACT interventions focused on defusion to reduce RNT.

With samples from Colombia, Peru, Mexico, Chile, and Spain (N=3389), the study confirmed the unidimensional structure of the CFQ and its measurement invariance across countries. The instrument showed excellent internal consistency and discriminant validity regarding emotional symptoms (DASS-21), supporting that cognitive fusion is a distinct construct. Results endorse the use of CFQ in cross-cultural comparisons and its utility for assessing key ACT processes.

The PSWQ-C was translated and validated in 585 Colombian children (8–12 years). The 3 reverse-scored items were removed, resulting in an 11-item version with high internal consistency (α/ω = .88). The unidimensional model showed excellent fit and scalar invariance by sex, age, and school. It showed convergent and discriminant validity against measures of repetitive negative thinking and anxiety. Girls scored higher than boys. The Spanish version facilitates the detection of pathological worry in child populations.

Trial with multiple baseline design (17 children, 8–12 years) compared RNT-focused ACT vs. non-directive therapy (NDT), in 3 videoconference sessions. ACT showed greater improvements in emotional symptoms and worry, reported by parents and children. Worry reduction mediated treatment effects. Findings support the utility of a brief, RNT-focused intervention for childhood GAD, outperforming a common control intervention like NDT.