Proposed follow-up study and survey instrument
Kevin Ellerton | Meditation University | building on the NYU pilot study by Dr. Aronson and Dr. Cohen, Making a Habit of Daily Meditation for Student Mental Well-being
Adherence, not efficacy, is the bottleneck. People know meditation helps; most never make it a habit, and the benefits require sustained practice. The NYU pilot study by Dr. Aronson and Dr. Cohen tested whether teaching the architecture of habit formation improves adherence over simply describing meditation's benefits.
The signal was encouraging but underpowered. The habit group meditated roughly 50% more and completed at a higher rate (about 71% vs 53%), though the small sample (48 enrolled, 23 completers) left the primary comparison short of significance. Two findings matter most for the next step: whoever kept practicing improved on mindfulness and stress regardless of group (consistency is the active ingredient), and about a quarter of enrollees finished the baseline but never started, which is the single biggest leak to design against.
The recent Singapore classroom study (Siew & Yu, 2026) points the same way. Intrinsic motivation was the strongest predictor of both engagement and perceived usefulness, while trait mindfulness and conscientiousness predicted finding practice easier; attachment style, conflict style, grit, and attention control showed no significant link. The convergent message: motivation is the lever, and adherence is the prize.
The levers below are ranked by best estimate of impact. The grading separates what the literature already supports, what is mixed, and what is genuinely novel, so we invest where the payoff is highest.
| Lever | Evidence | Draws on | Note |
|---|---|---|---|
| 1. Induce intrinsic motivation early via a short guided audio that helps a person meet a real source of their own pain (anxiety, racing mind, low mood) with awareness and acceptance, so they feel relief firsthand | Novel | Van Cappellen et al. (2020); calls in the national-sample motivation literature | Intrinsic motivation is the top predictor but is almost always measured, not created. The primary bet here, and fully scalable as audio. |
| 2. Group / spiritual friendship (small cohort or buddy) | Supported | Canby et al. (2021); Borgers/Karremans buddy pilot; Kok & Singer (2017); Upaddha Sutta | Group atmosphere and peer accountability predict adherence. Ancient pedigree (sangha) and modern data agree. |
| 3. Start absurdly small (kaizen: 1 min, then build) | Supported | Fogg, Tiny Habits; gradual-training tradition | Lowers the activation cost that likely drove early dropout. |
| 4. Anchor to an existing cue, ideally a fixed morning anchor, with a night bookend | Supported | Stecher et al. (2021, 2023) | Anchoring works; fixed morning anchors are the most durable. Waking and sleep are 100% reliable daily cues. |
| 5. Temptation bundling with the morning tea or coffee people already drink, plus sensory ritual (cushion, incense, music) | Supported (exercise) / Novel (meditation) | Milkman et al.; Zen tea tradition (chazen ichimi) | Bundling raised gym visits ~51%. The beverage doubles as an anchor. Untested for meditation. |
| 6. Mala beads worn as a visible cue and used as a tactile anchor and session counter | Novel / traditional | Japa tradition; connects to Dr. Cohen's Japa research | One inexpensive object as cue + anchor + tracker + identity signal. Untested as an adherence aid. |
| 7. "Waiting moment" informal practice (lines, red lights) as present-moment cues | Novel | Eckhart Tolle's teaching | Turns ubiquitous dead time into practice; harder to measure objectively. |
| Financial incentives / streaks (alone) | Mixed | Stecher et al. (2023) | Can initiate but rarely maintain behavior after withdrawal. Use as scaffolding, not the core manipulation. |
Because recruitment size is the main constraint, here are three designs of increasing ambition. We pick the tier our recruitment can actually power.
| Tier | Recruited N | Design | What it tests |
|---|---|---|---|
| 1. Focused | ~60 to 120 (audience only) | Clean 2-arm RCT: motivation-induction audio + standard onboarding vs standard onboarding. Both arms share an anchor + tiny-start scaffold. | Does inducing intrinsic motivation early raise 8-week adherence? Powered for a medium effect. |
| 2. Factorial | ~150 to 300 (audience + modest ads) | 2×2: motivation induction (yes/no) × social support (solo vs buddy/cohort), on a shared anchor + kaizen base. | The two strongest levers and whether they interact. |
| 3. Optimization | 500+ (Meta ads at scale) | A MOST / fractional-factorial screen (Collins' Multiphase Optimization Strategy) varying several components at once: motivation audio, buddy, bundling, mala cue, reminders. | Which components carry their weight, and the best scalable package. The rigorous way to test many ideas efficiently. |
This needs no new data. The NYU pilot already collected baseline trait mindfulness (MAAS), baseline stress (PSS), and pre/post change, so we can test a hypothesis I find genuinely important for tailoring how we teach different people.
Hypothesis. People higher in baseline trait mindfulness adopt the practice more easily and may show solid outcomes. But the largest life changes occur in people who start low in trait mindfulness yet carry enough pain (anxiety, stress) to power through the hard early phase on willpower; over time they build trait mindfulness and show the biggest transformation. In short, I expect a baseline mindfulness × baseline distress interaction on the magnitude of change, not a simple main effect.
How to test. Enter baseline MAAS and PSS (and their interaction) as moderators of pre/post change in well-being, stress, and mindfulness, and as predictors of adherence. The NYU pilot is underpowered (about 23 completers), so this is hypothesis-generating only; we pre-register and power it properly in the next study.
Where the field stands. A 2024 meta-analysis of 177 RCTs (Sieder et al.) found that lower baseline trait mindfulness predicted larger improvement, a small effect consistent with a room-to-grow story. But an earlier meta-analysis (Buric et al., 2022) found higher baseline mindfulness linked to greater well-being gains and more session attendance. The question is open and outcome-dependent, which is exactly why a clean, pre-registered test is worthwhile. The practical payoff: tailoring delivery, a lighter touch for high-trait individuals, and more scaffolding plus motivation induction for low-trait, high-distress individuals.
Below is the working survey, updated to add a section on group and partnered practice per Dr. Aronson's suggestion, so the replication can test the social-support hypothesis. It is a draft pending IRB review; the consent statement is a placeholder.
Consent statement to be supplied and approved per IRB before launch.
| None | A little | A lot | |
|---|---|---|---|
| Anxiety | ( ) | ( ) | ( ) |
| Focus / attention | ( ) | ( ) | ( ) |
| Relationships | ( ) | ( ) | ( ) |
| Mood | ( ) | ( ) | ( ) |
| Sleep | ( ) | ( ) | ( ) |