In plain language
Cognitive behavioral therapy works, but its success rates have not improved in decades. One likely reason is that treatments are matched to diagnostic labels (like "social anxiety disorder") rather than to the individual person. This methodological paper by Stefan Hofmann asks a different question, in the spirit of process-based therapy: how do we pick the right intervention for this person, with this problem, in this situation?
The paper borrows an idea from engineering called network control theory. Instead of viewing a mental health problem as a hidden disease, it treats it as a network of interacting thoughts, feelings, and behaviors that can settle into stable "attractor states." Therapy, on this view, is a way of pushing a maladaptive network past a tipping point so it reorganizes into a healthier state. The key practical question becomes: which nodes or connections in the person's network, if changed, would move the whole system the most?
Hofmann lays out a concrete five-step procedure and illustrates it with a composite case ("Bill," a 30-year-old with performance-type social anxiety). The therapist and client draw the client's problem network together, build matrices linking network nodes to change processes and processes to treatment strategies, then choose the strategies with the greatest expected "control energy." For Bill, this analysis pointed to exposure plus attention training rather than a generic protocol. The approach requires nothing more than paper and pencil, and it gives therapists a systematic, personalized alternative to trial-and-error treatment selection.
Key findings
- Network control theory, a framework from engineering and mathematics, can be translated into a practical guide for selecting and personalizing psychotherapy strategies within process-based therapy (PBT).
- Mental health problems are modeled as dynamic networks with attractor states and tipping points: highly connected networks resist change until a perturbation pushes them past a critical threshold into a new stable state.
- The paper proposes a five-step clinical procedure: (1) build the client's idiographic problem network, (2) develop a process matrix linking change processes to network nodes, (3) develop a strategy matrix linking treatment strategies to processes, (4) implement the strategies to perturb the system, and (5) re-examine the network in a closed feedback loop.
- Interventions are prioritized by their expected "control energy" — how strongly a strategy influences change processes and how strongly those processes influence central nodes of the network.
- In the composite case of "Bill" (social anxiety, performance subtype), the matrices identified distress tolerance and self-focused attention as the highest-leverage processes, pointing to exposure therapy plus attention training as the optimal personalized treatment.
- The method works as a simple paper-and-pencil proof of concept, without requiring intensive ecological momentary assessment data or computational network estimation.
How to cite
APA
Hofmann, S. G. (2025). A network control theory of dynamic systems approach to personalize therapy. Behavior Therapy, 56(1), 199-212. https://doi.org/10.1016/j.beth.2024.10.006
BibTeX
@article{hofmann2025network,
author = {Hofmann, Stefan G.},
title = {A Network Control Theory of Dynamic Systems Approach to Personalize Therapy},
journal = {Behavior Therapy},
year = {2025},
volume = {56},
number = {1},
pages = {199--212},
doi = {10.1016/j.beth.2024.10.006}
}
Related work
- All publications by Joseph Ciarrochi (searchable, with free PDFs)
- Process-Based Therapy & Idionomic Analysis
Author: Joseph Ciarrochi (ORCID 0000-0003-0471-8100). Free copy hosted with permission for scholarly use. Please cite the published version via the DOI above.