Within-person observational study: what this kind of study can — and can't — tell parents

The same people are tracked repeatedly — here, night by night, with devices logging behaviour and sleep — and each person is compared only to themselves.

What it can show

Comparing a teen to their own usual nights rules out stable differences between families, which makes the pattern stronger than a one-off survey.

What it can't

It is still observational: heavy-use nights and poor-sleep nights moved together, which we phrase as linked, not proof.

1 episode in the launch series uses this design:

Episode 2⏱ 3:12 video · 📖 1 min read

Within-person observational study Video coming soon

Teens on Phones After 10pm Fall Asleep an Hour Later

Zhao, Hale & Christakis, JAMA Pediatrics, 2026

Try tonight: Set a 20-minute phone-free buffer before bed — phones down, ideally charging outside the bedroom, for the last 20 minutes of the night.

Questions about this design

What is a within-person observational study?

The same people are tracked repeatedly — here, night by night, with devices logging behaviour and sleep — and each person is compared only to themselves.

Can a within-person observational study prove cause and effect?

Comparing a teen to their own usual nights rules out stable differences between families, which makes the pattern stronger than a one-off survey. It is still observational: heavy-use nights and poor-sleep nights moved together, which we phrase as linked, not proof.

Which Parenting by Science episodes use this design?

Episode 2 — “Teens on Phones After 10pm Fall Asleep an Hour Later” (Zhao, Hale & Christakis, JAMA Pediatrics, 2026).

Meta-analysis (2)Systematic review & meta-analysis (2)Observational survey (1)Longitudinal cohort study (1)Survey + classic experiments (1)Cross-sectional survey (1)Observational cohort study (1)Cross-sectional study (1)Randomized clinical trial (1)