Analyzing Perceived Causal Networks with PECAN2: A Practical R Tutorial
Abstract
Perceived Causal Networks (PECANs) provide a structured representation of how individuals perceive causal relations among their problems or psychological processes. PECAN assessments can be used in different designs. A single individual may complete one PECAN at a single timepoint or complete PECANs repeatedly, for example once per day, allowing the assessments to be aggregated into a longitudinal (L-)PECAN summarizing the perceived network over a defined period. Alternatively, single PECAN assessments from multiple individuals can be aggregated to describe patterns across a patient population. Working with PECAN data requires decisions about data representation, visualization, descriptive analysis, and aggregation. This tutorial introduces PECAN2, an R package that provides a reproducible workflow for these tasks. Using 14 daily L-PECAN assessments from a hypothetical patient with primary low back pain, we demonstrate how a single PECAN can be extracted from assessment data, represented as node and directed-edge data frames, visualized, and summarized using descriptive centrality measures. We then show how repeated PECANs can be aggregated while distinguishing the occurrence of problems and perceived causal relations from their magnitude when reported. For edges, the workflow additionally distinguishes occurrence across all assessment occasions from occurrence conditional on both participating nodes being present. These alternative denominators address different substantive questions. We further demonstrate how aggregated statistics can be mapped onto graphical properties and describe optional composite indices combining occurrence and magnitude. PECAN2 provides a reproducible workflow for processing, visualizing, aggregating, and descriptively analyzing person-specific perceived causal network data, with potential applications in individualized assessment and case conceptualization.
Related
- Ensemble clustering: A practical tutorial
- Using qualitative data to build causal pathways on perceived associations between poor menstrual health and hygiene and harms among schoolgirls in western Kenya.
- Analyzing diagnostic modalities for malaria detection and surveillance in varied belts: Burning need
- Analyzing Test Items Using Difficulty Index and Discrimination Index for Evaluating Postgraduate Medical Residents
- Heavy Drinking in Young Adulthood Correlates With Earlier Genetic Cardiomyopathy Diagnosis: Analyzing a Retrospective Survey.
BibTeX
@misc{reichert2026analyzing,
title = {{Analyzing Perceived Causal Networks with PECAN2: A Practical R Tutorial}},
author = {Julian Reichert and Jonas Tesarz and Sven Schulz-Niethammer and Lars Klintwall and Felix Vogel},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/k4h2j_v1},
url = {https://osf.io/preprints/psyarxiv/k4h2j_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Analyzing Perceived Causal Networks with PECAN2: A Practical R Tutorial
AU - Julian Reichert
AU - Jonas Tesarz
AU - Sven Schulz-Niethammer
AU - Lars Klintwall
AU - Felix Vogel
PY - 2026
DA - 2026/09/06
PB - PsyArXiv
DO - 10.31234/osf.io/k4h2j_v1
UR - https://osf.io/preprints/psyarxiv/k4h2j_v1/
AB - Perceived Causal Networks (PECANs) provide a structured representation of how individuals perceive causal relations among their problems or psychological processes. PECAN assessments can be used in different designs. A single individual may complete one PECAN at a single timepoint or complete PECANs repeatedly, for example once per day, allowing the assessments to be aggregated into a longitudinal (L-)PECAN summarizing the perceived network over a defined period. Alternatively, single PECAN assessments from multiple individuals can be aggregated to describe patterns across a patient population. Working with PECAN data requires decisions about data representation, visualization, descriptive analysis, and aggregation. This tutorial introduces PECAN2, an R package that provides a reproducible workflow for these tasks. Using 14 daily L-PECAN assessments from a hypothetical patient with primary low back pain, we demonstrate how a single PECAN can be extracted from assessment data, represented as node and directed-edge data frames, visualized, and summarized using descriptive centrality measures. We then show how repeated PECANs can be aggregated while distinguishing the occurrence of problems and perceived causal relations from their magnitude when reported. For edges, the workflow additionally distinguishes occurrence across all assessment occasions from occurrence conditional on both participating nodes being present. These alternative denominators address different substantive questions. We further demonstrate how aggregated statistics can be mapped onto graphical properties and describe optional composite indices combining occurrence and magnitude. PECAN2 provides a reproducible workflow for processing, visualizing, aggregating, and descriptively analyzing person-specific perceived causal network data, with potential applications in individualized assessment and case conceptualization.
ER -