Is Your Baby Tracker Measuring What You Think It Is?

Explanation

"She slept 9 hours" sounds like a fact about a baby. It is really a fact about a measurement, and the method changes the answer.

The Rise & SHINE study is the clearest demonstration, because it ran three methods over the same infants on the same nights: ankle actigraphy, parent diaries, and parent questionnaires, across seven days in 314 infants averaging 6.4 months.

Rise & SHINE

Supporting evidence

Against the actigraphy measurement, both parent-reported methods came out higher:

Method	Night sleep	Daytime sleep
Parent questionnaire	+67 min	+29 min
Parent diary	+43 min	+31 min

And the actigraphy figures themselves came with a wide spread across infants: night sleep averaged 526 minutes with a standard deviation of 67 minutes, and the longest night sleep period averaged 460 minutes with a standard deviation of 100 minutes.

The 2025 review of 35 infant-sleep studies published between 2000 and 2024 reported the same structural issue: systematic differences between measurement methods, alongside substantial variability between and within studies.

2025 mini-scoping review

What the numbers mean

Put the two findings together and something specific follows.

The gap between measurement methods (43 to 67 minutes) is a large fraction of the ordinary spread between healthy babies (a standard deviation of 67 minutes on night sleep). In other words: the difference caused by how you measured can be about the same size as the difference between one baby and another.

That has a direct consequence for the most common thing parents do with these numbers, which is compare them:

• Comparing your logged number with another family's logged number compares two methods as much as two babies.
• Comparing your logged number with a published research average compares a recalled figure with a device-measured one, in a direction the study above tells us is not random.
• Comparing your baby's number this week with the same baby's number last week, recorded the same way by the same people, removes almost all of that, because the method error is roughly constant and cancels.

The third comparison is the only one of the three that is mostly about the baby.

What parents can observe

If a number is going to be used at all, these are what make it usable:

• Consistency of method. Whatever your household counts as "asleep", count it the same way every time. An inconsistent definition is worse than a slightly wrong one.
• When the entry was made. The diary beat the questionnaire in the study above by roughly 24 minutes on night sleep, and the only difference was how soon it was written.
• Change over time in your own log, rather than the absolute value.
• What else changed in the same period, as context for the change, never as its cause.

What may explain a pattern

A number that moves can be a real change in the baby, or a change in how the household is measuring: a new caregiver on shift with a slightly different definition of settled, entries made later in the evening than they used to be, a week where someone was logging from memory.

Both are worth knowing about, and they look identical in the data. The only way to tell them apart is to know how the log was kept, which is why when and who are worth recording alongside what.

What parents can do

Treat a tracked number as your household's own instrument: useful for comparing this week to last week, unreliable for comparing your baby to someone else's. Keep the method stable, record at the time rather than at the end of the day, and bring trends rather than single figures to appointments.

And when a number surprises you, check the measurement before drawing a conclusion about the baby. That is the same discipline the researchers above applied to their own data.