Practice · research august 2026 · published 2026-08-03 · v1 · 4 min read
The sandwich method
Deep work with a mirror holds when a person stands on both sides of it
What the evidence actually supports about wrapping machine-assisted self-examination in human contact, which parts are well measured, and where our own synthesis outruns the studies. The canonical treatment of the sandwich method.
Writing your way through something difficult works, modestly. Across more than a hundred studies of the expressive-writing paradigm James Pennebaker introduced in the 1980s, the average benefit to health and distress lands small, in the region of a sixth of a standard deviation, larger in narrower sets and stubbornly uneven to replicate. Small and real is still real. What is new is that the writing now has a fluent counterpart inside it, awake at any hour, with no closing time and no view of its own that costs anything, and our own essay on this fork prescribed a protective pattern for exactly that situation. Sandwich the deep work between human contact. An intention set with a person before, an integration with a person after. We attributed the pattern to self-discovery research and named no study, which is an assertion wearing a citation’s clothes, so this brick is the result of going and looking.
The closest thing to a direct test appeared in 2022, when a meta-analysis in Clinical Psychology Review pooled thirty-one trials and 6,215 people to ask whether text-based internet treatments do better with a human being attached to them than with the technology guiding alone. They do, on symptom reduction and on adherence both, and the authors are careful that the margin is slight rather than dramatic. The wider guided-versus-unguided literature agrees and adds an honest wrinkle. The advantage shows up clearest early and narrows at follow-up, and the arms labelled unguided frequently contain some human contact anyway, which makes the real contrast smaller than the labels suggest. Take that body at its most conservative and the direction survives. The same material, through the same text channel, does more when a person is somewhere in the loop.
For the two slices of bread the evidence is strong and slightly off-target, which is worth saying rather than smoothing. Before is the best measured of the three. Gollwitzer and Sheeran pooled ninety-four independent tests and found that spelling out in advance when, where, and how you will act moved goal attainment by a medium-to-large margin, d of 0.65, which is a large number by the standards of that field. It is a finding about the intention being specified, not about it being said to somebody, so the corpus’s version adds a social commitment the studies did not isolate.
After is nearly as well measured and carries a warning inside it. Tannenbaum and Cerasoli pooled the debriefing literature in 2013 and found that individuals and teams who conduct a structured review of what just happened outperform those who do not by roughly a quarter, which is the reflection half of the sandwich in an industrial register. The warning comes from Bernard Rimé’s long program on social sharing. People share almost every emotional experience they have, and mere sharing does not produce recovery; the sharing that helps is the kind that does cognitive work on the event rather than replaying the feeling. So the conversation afterwards earns its place by asking what you are going to do, not by being present while you talk.
Which locates the mechanism, and it is not about warmth. A machine counterpart has no stake in how your week turns out, so it can receive an intention and never ask about it again, and receive a conclusion and never notice that you did nothing with it, while a person on either side supplies precisely the two forces missing there: a commitment that is owed, and an account that will be called. What none of this evidence covers is the sandwich as a unit. No trial we could find has tested human intention-setting, machine-assisted self-examination, and human integration as one sequence, so the parts are cited and the assembly is ours, offered as a practice and gradeable the day somebody runs it properly. The honest version of the prescription is narrower than the confident one and more useful than the vague one. Do the hard looking with the machine, which will look with you at three in the morning and never tire, and put a human being at both ends. The mirror will show you the thing. A person is what makes you do something about it, and two short conversations, one before and one after, are the whole price of an insight that outlives the night you had it.
Evidence and lineage
Research trail
Follow the sources, inspect how the claims are graded, or propose a correction at the exact record it concerns.
Sources 7
-
Koelen and colleagues (2022). Man vs. machine. A meta-analysis on the added value of human support in text-based internet treatments (e-therapy) for mental disorders (Clinical Psychology Review) . Clinical Psychology Review, 96, 102179 . DOI 10.1016/j.cpr.2022.102179
The closest direct test of the sandwich's central premise. Thirty-one studies, 6,215 participants, comparing human guidance against technological guidance in text-delivered treatment; human guidance came out ahead on symptom reduction and on adherence, with the authors describing the margin as slight.
Comment on this source -
Eirini Karyotaki and colleagues (2021). Internet-Based Cognitive Behavioral Therapy for Depression: A Systematic Review and Individual Patient Data Network Meta-analysis . JAMA Psychiatry, 78(4), 361-371 . DOI 10.1001/jamapsychiatry.2020.4364
The wider guided-versus-unguided body. Guided conditions outperform unguided ones in the short term; the advantage attenuates by follow-up, and arms labelled unguided often contain basic support, which shrinks the true contrast. Cited for direction, not magnitude.
Comment on this source -
Peter M. Gollwitzer and Paschal Sheeran (2006). Implementation intentions and goal achievement. A meta-analysis of effects and processes (Advances in Experimental Social Psychology, 38, 69-119) . Advances in Experimental Social Psychology, 38, 69-119 . DOI 10.1016/S0065-2601(06)38002-1
The evidence for the front slice. Ninety-four independent tests; specifying in advance the when, where, and how of acting produced a medium-to-large effect on goal attainment (d = 0.65). Scope limit noted in the text: the studies isolate specification, not the presence of another person.
Comment on this source -
Scott I. Tannenbaum and Christopher P. Cerasoli (2013). Do team and individual debriefs enhance performance? A meta-analysis (Human Factors, 55, 231-245) . Human Factors, 55(1), 231-245 . DOI 10.1177/0018720812448394
The evidence for the back slice. Structured post-event review raised subsequent performance by roughly a quarter across teams and individuals. Setting is work and training performance rather than personal insight, which is the extrapolation the brick makes and flags.
Comment on this source -
Joanne Frattaroli (2006). Experimental disclosure and its moderators: A meta-analysis . Psychological Bulletin, 132(6), 823-865 . DOI 10.1037/0033-2909.132.6.823
The evidence for the filling. Solitary written disclosure about a stressful experience yields small average benefits, around d = 0.16 across the accumulated literature and larger in narrower sets such as Smyth's thirteen-study analysis at d = 0.47, with replication acknowledged as uneven.
Comment on this source -
Emmanuelle Zech and Bernard Rimé (2005). Is talking about an emotional experience helpful? Effects on emotional recovery and perceived benefits . Clinical Psychology & Psychotherapy, 12(4), 270-287; later cognitive-work result: Nils and Rimé 2012, DOI 10.1002/ejsp.1895 . DOI 10.1002/cpp.460
The warning inside the back slice. Emotional experiences are shared at very high rates, yet sharing alone does not produce emotional recovery; the sharing associated with recovery is the kind doing cognitive work on the event rather than socio-affective replay. Grounds the brick's insistence that the after-conversation asks what happens next.
Comment on this source -
Ozlem Ayduk and Ethan Kross (2010). From a distance: Implications of spontaneous self-distancing for adaptive self-reflection . Journal of Personality and Social Psychology, 98(5), 809-829 . DOI 10.1037/a0019205
Adjacent support for why the filling can go wrong on its own. Reflection from a self-distanced perspective yields meaning and closure where self-immersed reflection yields rumination. Consulted during the research pass and not load-bearing in the brick, which does not claim the machine counterpart induces distancing.
Comment on this source
Claims and confidence 7
- verified
Human guidance outperforms technological guidance in text-based internet treatments on both symptom reduction and adherence, across thirty-one studies and 6,215 participants.
Koelen et al. 2022, Clinical Psychology Review. Peer-reviewed meta-analysis; the authors characterise the advantage as slight, and the trials are clinical populations rather than general self-examination.
Sources: [1]
Respond to this claim - directional
Guided internet-based self-help outperforms unguided self-help in the short term, with the advantage attenuating by follow-up.
Karyotaki and colleagues' individual-participant-data meta-analyses and successor network analyses. Direction consistent; magnitude unstable because arms labelled unguided frequently include basic human support.
Sources: [2]
Respond to this claim - verified
Forming an explicit implementation intention raises goal attainment by a medium-to-large margin, d = 0.65 across ninety-four independent tests.
Gollwitzer and Sheeran 2006. Well replicated. Scope: the manipulation is specification of when, where, and how, not a commitment stated to another person, so the corpus's social version is an extension.
Sources: [3]
Respond to this claim - verified
Structured debriefs raise subsequent performance by approximately 25 percent for individuals and teams.
Tannenbaum and Cerasoli 2013, Human Factors, published and unpublished studies pooled. Setting is work and training performance; transfer to personal reflection is our inference.
Sources: [4]
Respond to this claim - directional
Expressive writing about a stressful experience produces small average benefits with uneven replication.
The Pennebaker paradigm's accumulated meta-analyses. Around d = 0.16 across the full literature, up to d = 0.47 in narrower early analyses; boundary conditions for the effect remain unsettled.
Sources: [5]
Respond to this claim - directional
Talking about an emotional experience does not by itself produce emotional recovery; the sharing associated with recovery does cognitive work on the event.
Zech and Rimé and the two-mode account of social sharing. The null for mere sharing is robust within that programme; the two-mode explanation is a theory with supporting but not decisive evidence.
Sources: [6]
Respond to this claim - directional
No published trial tests human intention-setting, machine-assisted self-examination, and human integration as a single sequence.
Our literature search of 2026-08-03 across the guided self-help, expressive-writing, implementation-intention, debriefing, and social-sharing literatures. Absence found in searching, which is weaker than an absence established by systematic review.
Sources: [1] , [2] , [3] , [4] , [5] , [6]
Respond to this claim
Read next
You have walked Practice end to end, from the refusal that starts the work to the two conversations that make it survive the week. Every step here was built to be run on something you actually know, and none of it waits on anyone's permission.
Practice Take one thing you have been circling for months and cannot yet say cleanly. Tell one person, out loud, that you are going to work it out this week, and let them expect an answer from you. Then take it to the machine and refuse the first draft, and the second, until what comes back is a sentence you could not have written on Monday. Fix it, with your name and the date on it. Then go back to the person who heard the intention, tell them what you found, and let them ask what you are going to do about it.