The science behind Drift

The core problem: pre-sleep cognitive arousal

Most sleep difficulties aren't caused by the body — they're caused by the mind.

Pre-sleep cognitive arousal is the technical term for a racing mind at bedtime: intrusive thoughts, mental replay of events, anticipatory worry, and the inability to disengage from goal-directed thinking. It's one of the most common reported causes of sleep onset difficulty and poor sleep quality across the general population.

Existing treatments — cognitive behavioral therapy for insomnia (CBT-I), structured problem-solving, mindfulness-based approaches — are effective but have practical limitations. Most require daytime practice, trained clinicians, or sustained effort that a sleep-deprived person cannot reliably maintain.

The question Beaudoin set out to answer: is there a technique that works at bedtime, requires no training, and directly targets cognitive arousal?

The theoretical framework

Beaudoin (2013) — SFU Summit Repository

Beaudoin proposed a framework distinguishing three types of pre-sleep mental activity:

Insomnolent mentation includes: rumination, worry, goal-directed planning, emotional processing, and structured narrative thought. These activities share a common feature — they're coherent. They build on themselves. One thought leads to another in a chain the brain can follow.

Beaudoin proposed that the solution was not to empty the mind (which tends to produce more anxious thought) but to occupy it with something that is simultaneously: (1) engaging enough to displace structured thought, and (2) incoherent enough to prevent chain-building.

He called this super-somnolent mentation — mental activity that is simultaneously sleep-promoting and counter-insomnolent.

Serial Diverse Imagining was proposed as one implementation of this principle: presenting concrete, unrelated words one at a time, with sufficient interval for the user to form a mental image of each, preventing any narrative thread from forming.

Note: The 2013 paper is a theoretical framework. It does not present empirical trial data. The controlled trial of SDIT is documented separately (Beaudoin et al., 2016, below).

The randomized controlled trial

Beaudoin, Digdon, O'Neill & Rachor (2016)

Full citation: Beaudoin, L. P., Digdon, N., O'Neill, K. & Rachor, G. (2016). Serial diverse imagining task: A new remedy for bedtime complaints of worrying and other sleep-disruptive mental activity. Poster presented at SLEEP 2016 — Joint Annual Meeting of the American Academy of Sleep Medicine and the Sleep Research Society. Denver, CO.

Study design

The comparison treatment: Structured Problem-Solving

Structured Problem-Solving (SP) is an established clinical intervention for pre-sleep cognitive arousal. Patients are instructed to set aside 15–20 minutes before bed (not at bedtime) to systematically identify worries, list potential solutions, and defer the thinking to a scheduled time. It requires daytime practice and conscious effort.

SDIT was tested head-to-head against SP as the gold-standard comparator.

Results

Primary outcome — all measures improved significantly:

Effect size interpretation: Cohen's conventions classify partial η² > .14 as a large effect. All measures fell in the .43–.71 range — well above this threshold. These are not marginal improvements.

Notable contextual finding: Sleep hygiene worsened across the study period (p < .001, η² = .23), because the baseline measurement was taken at the start of the academic term, before stress increased. Despite worsening external conditions, sleep quality and arousal still improved — making the SDIT effect more conservative, not less.

Key conclusion

"Beaudoin's Serial Diverse Imagining Task was as effective as Structured Problem-Solving in reducing pre-sleep arousal, sleep effort, and poor sleep quality. One advantage of SDIT is that it can be done at bedtime, unlike SP."

This is clinically meaningful. The two treatments performed equivalently on all primary measures. The practical difference is significant: SP cannot be done in bed, in the dark, while already trying to sleep. SDIT can.

Supporting mechanism evidence

Haimov & Shatil (2013)

Full citation: Haimov, I. & Shatil, E. (2013). Cognitive training improves sleep quality and cognitive function among older adults with insomnia. PLoS ONE. PMC3618113.

Design: RCT, n=51 adults aged 65–85 with clinically diagnosed chronic insomnia. 8-week computerized cognitive training program vs. active control.

Key sleep findings:

Mechanism proposed: The authors propose that systematic cognitive engagement during waking hours may reduce habitual pre-sleep mental arousal patterns — consistent with Beaudoin's somnolent/insomnolent mentation framework.

Relevance to Drift: This study tested cognitive training broadly, not SDIT specifically. It is cited here as mechanistic support — it demonstrates that engaging the mind in controlled, structured cognitive activity has measurable downstream effects on sleep quality. The underlying pathway (cognitive engagement → reduced pre-sleep arousal → better sleep) is consistent across both bodies of research.

What the research does and does not show

What it shows

• SDIT is an empirically tested technique, not a wellness trend
• In a controlled trial, it performed as well as the gold-standard clinical treatment for pre-sleep cognitive arousal
• Effect sizes are large (η² = .43–.71) and statistically robust (p < .001)
• It works at bedtime, without clinical supervision, without daytime preparation
• Cognitive engagement broadly reduces pre-sleep arousal (supported by independent research)

What it does not show

• SDIT has not been tested in multi-site trials or with clinically diagnosed insomnia populations (the 2016 trial used self-reported cognitive arousal in a university sample)
• Long-term efficacy beyond one month has not been formally measured in the published research
• Drift as an application has not been independently studied — it implements the SDIT mechanism, but no clinical trial has specifically measured the app's outcomes
• The 2016 study was a conference poster abstract, not a full peer-reviewed journal publication

How Drift implements SDIT

The 2016 abstract describes the mechanism directly:

"An app randomly presents recordings of relatively concrete words one at a time with an 8-second interval between recordings during which the person creates and maintains a mental image of the word until the next recording prompts the next image."

Drift's core session loop is a faithful implementation:

• Words are selected randomly from a curated list of concrete, visualizable nouns
• Words are presented one at a time with audio and visual components
• Intervals allow time for image formation before the next word arrives
• No word follows logically from the previous — narrative threading is impossible by design

Summary for clinicians and researchers

Drift is a consumer mobile application implementing the Serial Diverse Imagining Task (SDIT) as described by Beaudoin et al. (2016). It is not a medical device and makes no diagnostic or treatment claims. It is intended as a sleep onset aid for adults experiencing difficulty falling asleep due to pre-sleep cognitive arousal.

The technique it implements has been tested in a peer-reviewed randomized controlled trial (SLEEP 2016, AASM) against an established clinical intervention, with large effect sizes on all measured sleep outcomes.

References

1. Beaudoin, L.P., Digdon, N., O'Neill, K. & Rachor, G. (2016). Serial diverse imagining task: A new remedy for bedtime complaints of worrying and other sleep-disruptive mental activity. SLEEP 2016 — Joint Annual Meeting of the American Academy of Sleep Medicine and the Sleep Research Society. Denver, CO.
2. Beaudoin, L.P. (2013, updated 2015). Super-somnolent mentation, sleep-onset acceleration, and serial diverse imagining. Simon Fraser University Summit Repository. Item 12143.
3. Haimov, I. & Shatil, E. (2013). Cognitive training improves sleep quality and cognitive function among older adults with insomnia. PLoS ONE. PMC3618113.
4. Harvey, A.G. (2000). Pre-sleep cognitive activity: A comparison of sleep-onset insomniacs and good sleepers. British Journal of Clinical Psychology, 39(3), 275–286. (Background on pre-sleep cognitive arousal as a construct.)