The Problem We Are Trying to Solve
A significant proportion of people who seek help for chronic psychological distress - anxiety, depression, burnout, existential stuckness - do not respond adequately to available treatments, or respond and relapse. We think part of the reason is that most treatments work on the contents of distress rather than its generative structure.
We have developed a brief relational intervention, the Odoki Method, that works differently. It does not address thoughts, beliefs, narratives, or behaviours directly. It works by directing attention to the somatic signal that underlies those contents, before conceptual capture occurs. We think this produces structural change rather than symptomatic management - and our preliminary data suggests it does so quickly.
We are looking for research collaborators who can help us study this properly.
The Mechanism Hypothesis
We propose that a significant class of psychological suffering is maintained by what we term the reification of the self-model: the process by which a dynamic biological state - the body’s ongoing allostatic response to perceived threat - becomes organised around the defence of a fixed identity structure.
In predictive processing terms: the brain assigns high precision weighting to identity-level priors, which causes incoming interoceptive signals to be interpreted through rigid conceptual categories rather than updated against them. The result is a self-reinforcing loop in which somatic alarm generates defensive identity constructs, which in turn sustain the alarm.
The Odoki Method interrupts this loop from the outside. A trained guide directs the participant’s attention to the somatic signal itself - prior to its conceptual packaging - while providing a stable relational container that allows the system’s precision weighting to shift without triggering further defensive closure. The guide works without knowledge of the participant’s narrative history, which prevents the guide’s own predictive hierarchy from flooding with top-down simulations of the participant’s situation.
If this account is correct, the intervention should produce broad-spectrum, transdiagnostic effects - improvements in general wellbeing more consistent than improvements in specific symptom measures - because it is targeting a generative structure rather than symptomatic content. This is what the pilot data shows.
Preliminary Findings
We have completed a pilot study (n=19, convenience sample, single guide, no control condition) using three validated instruments: WEMWBS, QIDS-SR, and GAD-7, administered pre- and post-intervention via our participant management platform.
| Measure | Mean improvement | Cohen’s d | Clinically meaningful improvement (MCID) |
|---|---|---|---|
| WEMWBS | 15.2 points | 1.24 (large) | 94.7% (18/19) |
| QIDS-SR | 3.1 points | 0.91 (large) | 36.8% (7/19) |
| GAD-7 | 3.7 points | 0.76 (medium) | 42.1% (8/19) |
The WEMWBS result is the most striking: 18 of 19 participants achieved clinically meaningful improvement, with a mean shift from the low-average range into flourishing. The QIDS-SR MCID figure is lower, but this is partly a floor effect - most participants entered with mild or minimal baseline scores, limiting room for MCID-threshold change. The GAD-7 picture is more complex: five participants did not improve, but for all five, WEMWBS improved without exception and four of five also improved on QIDS-SR, suggesting the GAD-7 may be registering transitional arousal rather than worsening condition.
Sessions average 30 minutes. Later cohort participants completed the full process in as few as 10 sessions (5 hours total contact time). The method is delivered without knowledge of the participant’s personal history or situational narrative.
Full findings, including individual trajectory charts and cohort flow diagrams, are available at odoki.com/research/.
What We Don’t Know
The limitations are significant and we state them plainly. The sample is small, recruited through personal networks, and assessed by the same individual who delivered the intervention. There was no control condition and no follow-up data. The single-guide design means observed effects cannot yet be distinguished from individual practitioner skill - though we now have additional trained guides with comparable early results.
We do not know what is actually happening mechanistically. We have a theoretical account, but we have not measured it. We do not know whether the effects persist, though the structural rather than symptomatic account of the mechanism would predict durability - a hypothesis we consider a priority for follow-up study.
These are the questions we want to study.
The Ask
We are seeking a conversation with researchers who have the methodological expertise to help us design studies that could answer these questions - starting with mechanistic work that could characterise what is actually occurring during the intervention, before moving toward controlled efficacy trials.
We are not looking for validation. We are looking for rigorous scrutiny and collaborative study design.
If this is of interest, I would welcome a brief call at your convenience.
Malcolm Holmes
[email protected]
The Odoki Method is delivered by Odoki Ltd, a UK registered company. We are actively exploring charitable structures to steward the method and receive research funding. Full theoretical background and pilot study write-up: odoki.com/research/