Spanish validation of the PBAT and cross-country mapping of process-outcome networks using link community analysis

@article{goicoechea2026spanish,
  author = {Goicoechea, Carmen and Wallman-Jones, Amie and Ciarrochi, Joseph and Hayes, Steven C. and Hofmann, Stefan G. and Perakakis, Pandelis},
  title = {Spanish validation of the PBAT and cross-country mapping of process-outcome networks using link community analysis},
  year = {2026},
  journal = {Journal of Contextual Behavioral Science},
  volume = {41},
  pages = {101021},
  publisher = {Elsevier BV},
  doi = {10.1016/j.jcbs.2026.101021}
}

Study type: Cross-sectional validation and network comparison.

In plain language

The Process-Based Assessment Tool (PBAT) is an item pool for assessing biopsychosocial processes relevant to psychological change. Goicoechea and colleagues examined a Spanish translation in 602 adults, following the original U.S. validation protocol. They studied relationships among PBAT items and their associations with distress, health, vitality, and psychological need satisfaction or frustration.

The researchers then compared process–outcome networks from Spain, the United States, Sweden, and Poland. They used the Boruta algorithm to identify predictive items and link community analysis to show how processes and outcomes formed overlapping groups. Some broad patterns recurred, but the most prominent predictors of distress differed across samples.

These networks describe relationships between people at one time. They provide evidence about the Spanish item pool and generate hypotheses for further research. They do not show which process causes distress in a particular person or establish that a country-level pattern should guide an individual’s treatment.

Key points

• The Spanish items showed expected associations. Positive behaviors generally related to need satisfaction and well-being, while negative behaviors related to need frustration and distress. The retention item showed a less straightforward pattern.
• Predictive patterns differed across samples. Difficulties finding appropriate emotional outlets were prominent in Spain; difficulty changing ineffective behavior was prominent in the U.S.; cognitive interference and attentional difficulties were prominent in Sweden. Poland showed a more distributed pattern across negative affect, attention, and variation.
• Networks made shared connections visible. Link communities grouped processes and outcomes by shared predictive connections and allowed a node to belong to more than one community.
• Comparability had limits. Analyses supported equivalent factor structure and loadings between Spain and Sweden, but five item intercepts were non-invariant. Item-level U.S. and Polish data were unavailable for those invariance tests.

What the paper does and does not establish

• Cross-sectional prediction is not evidence of within-person dynamics or causation. Testing the proposed personalized use of these networks requires repeated assessments within individuals.
• Country differences could reflect recruitment, sample characteristics, or translation as well as cultural context. The study does not isolate cultural causes.
• The Spanish sample covered ages 18–75, and the study’s clinical outcomes and need measures were self-reports. Further validation in clinical settings and across groups is needed.

Related work

• Development of the Process-Based Assessment Tool (PBAT)
• Toward empirical process-based case conceptualization
• Time-series machine learning for individual treatment planning
• All publications by Joseph Ciarrochi

Summary based on the full published article supplied by the author. The local PDF is the publisher version, published online on 1 August 2026. Summary prepared from the full paper on 29 September 2026. Please cite the published article linked above.