Problematic smartphone use

Problematic smartphone use is the research literature's name for phone use that the person experiences as out of hand and that is costing them something. It is not a diagnosis. No classification system contains it, and the clumsy phrase is deliberate: it is what the field settled on after deciding it could not defend the word addiction.

That retreat is the interesting part, and it is not a quarrel about vocabulary. A name arrives with a checklist attached, the checklist becomes the questionnaire items, and the items are the measurement. Changing what this is called changed what gets counted.

Behavioural addiction is a real category with a short membership list. ICD-11 admits gambling disorder and gaming disorder and stops there; nothing about phones is in it, or in DSM-5, or in any other manual.

The reason is not timidity. When the concept was carried over from substance dependence it came with the six-component template Griffiths set out — salience, mood modification, tolerance, withdrawal, conflict and relapse — and two of those six have no clear referent here. Tolerance means needing more of the thing to get the same effect, and nobody has said what "more phone" would be. Withdrawal, in the phone studies, is operationalised as discomfort at being separated from the device, which is also what being separated from your keys or your wallet produces. Panova and Carbonell went through the case criterion by criterion and concluded the evidence did not support addiction as the right word, with maladaptive coping being the better description of most of what the scales pick up.

Billieux and colleagues pressed a sharper objection, and it is about method rather than phones. If you assume a substance template and write items to confirm it, you will find an addiction in any behaviour common enough to supply variance. They call this overpathologising everyday life, and the warning generalises: by that procedure, work, exercise, sex, dancing and sun-tanning have all been proposed as addictions.

There is also a problem with the object. The phone is a channel, not a reinforcer. Nobody is attached to the glass. The things people cannot leave are specific — a feed, a game, a group chat, one person's replies — and they have different mechanics, which an instrument that scores "smartphone use" averages into one number.

The two most used instruments are Bianchi and Phillips's Mobile Phone Problem Use Scale and Kwon's Smartphone Addiction Scale. Read their items and the construct is unmistakable, and it is not quantity.

They ask about preoccupation (it is the first thing in the morning and the last at night, it intrudes while you are doing other things), loss of control (longer than intended, repeated failed attempts to cut down), withdrawal-like unease (restless, impatient or anxious without it), and functional impairment (missed sleep, missed work, people complaining, money).

Almost nothing on them asks how many hours. Where duration appears it is self-estimated and peripheral. So the thing being measured, by design, is the experience of use being out of hand and the damage attached to it.

Which creates a specific difficulty. "I feel impatient when I cannot check my phone" and "it is the first thing I reach for in the morning" describe most adults alive in 2026, so where the line gets drawn decides everything downstream. Published prevalence estimates for problematic use range from a few per cent to over a third of a sample, and the spread comes mostly from which scale and which cut-off, not from which population. The cut-off problem is the general version of this, and this is one of its cleaner examples.

Ellis and colleagues asked the obvious question of these instruments: do the scores predict what the phone records? Largely they did not. Scale scores had little relationship to objectively logged use, which is an unflattering result for a questionnaire with the word use in its name.

Meanwhile the same scales relate consistently to depression and anxiety. Elhai's conceptual overview is the standard reference for that half, and it is also where the field's resolution sits: the questionnaires that predict how somebody is doing are the ones least connected to how long the phone was on.

One honest caveat, because it cuts against the finding. These scales contain items about distress and impairment, so some of their correlation with distress is built in by construction rather than discovered. What survives that objection is narrower but still there.

The separate argument — that the hour count in the settings app is the wrong number for ordinary readers to be worrying about — is made in full, with its own evidence, in the piece on screen time, and is not repeated here. This page is about the construct and its name.

Kardefelt-Winther's compensatory model is the other reading on the table, and it reverses the arrows: heavy use is often doing a job, most often relieving something the person has no better way to relieve. The candidate studied most is loneliness. On that account, a high score is a symptom and treating the phone is treating the symptom.

A description of two recent weeks, on two dimensions: how much control you felt, and what it cost. Not a kind of person, and not a condition, because no condition exists to be had.

The cost usually shows up in three places that have entries of their own, and one sentence each is the right amount here. Evening light and what the screen contains belong to screens and sleep. Being unable to go to bed while exhausted is its own phenomenon with its own causes, covered in revenge bedtime procrastination. And the reason interrupted work stays expensive after the interruption ends is attention residue.

The phone check is built on the construct above rather than on a total, and reads three facets separately instead of producing one severity number.

The scope note is unusually easy to write, because the field has already written it. There is no diagnosis here, so nobody can be diagnosed with this and nobody should accept being told they have it. What is real is distress and cost, and those are worth taking seriously on their own terms: use that has cost sleep, work or a relationship over months and has survived genuine attempts to change it is worth talking to someone about, and in a decent conversation the first question will be what the use is doing for you. Gambling and gaming are the exceptions with recognised diagnoses behind them, so if the activity is one of those, the honest framing changes. And the field's own position is best read as an admission rather than a conclusion: "problematic" is a placeholder held in place on purpose, protecting the possibility that this turns out to be a symptom rather than a disorder.