The standard advice offered to a young woman experiencing something unpleasant online is to disengage. Delete the app, close the account, stop looking.

It is offered as though it were free. It is not, and understanding what it costs explains why it is so rarely taken.

What is actually being given up

For most adolescents, online platforms are not an entertainment supplement to social life. They are where a substantial portion of social life takes place.

Group coordination happens there. Invitations circulate there. Shared jokes, ongoing conversations and the ambient awareness of what a friendship group is doing all live there.

Leaving does not remove a person from a diversion. It removes her from the channel through which her social group operates, and the group does not relocate to accommodate her.

Research on adolescent social media use consistently finds that maintaining existing friendships is among the primary reported uses, and that adolescents describe non-participation as socially costly.

The asymmetry

The advice also allocates the cost to the wrong person.

Where harassment is the reason for withdrawal, the outcome is that the harasser retains access to the space and the person harassed does not. This is a solution that resolves the problem by removing the victim, which would be recognised as inadequate in any other setting.

The same asymmetry appears in the workplace version of this advice — that a woman experiencing harassment online should reduce her professional visibility — with the same consequence, which is that participation becomes conditional on tolerance of abuse.

What adolescents actually do instead

Qualitative research on how young people manage unwanted contact and harassment finds a much wider repertoire than adults generally assume.

Account segmentation is common — separate accounts for different audiences, with different privacy settings and different content.

Selective visibility features are used heavily, including restricting who can see specific posts.

Blocking is used readily, and there is a well-developed set of social conventions around it.

Temporary withdrawal is used tactically rather than permanently.

The picture that emerges is of considerable practical sophistication, and it sits oddly with advice premised on adolescents not understanding the risks.

Where the sophistication runs out

Three areas where the research suggests young people manage less well.

Persistence. Adolescents are reasonably good at managing current audience and less good at anticipating that content persists and may be seen by a future audience — employers, institutions, future partners — with different standards.

Aggregation. Individual pieces of information disclosed separately can be combined to identify location, school and routine. This is the mechanism behind most location-based risk, and it is not intuitive.

Coercion within relationships. Pressure from a partner to share passwords, to enable location sharing, or to send images is one of the more common experiences reported and one of the least well addressed by safety education, which is oriented towards stranger risk.

That last point is important. Most safety education is built around a threat model — the unknown adult — that accounts for a minority of actual harm. The majority involves known people, frequently peers, and frequently within relationships.

Better advice than withdrawal

Segmentation rather than abstinence. Separate accounts with different audiences give control without exit.

Configure once, properly. Most platforms have restrictive options that are rarely used because they are buried. An hour spent on settings is worth considerably more than repeated situational management.

Teach aggregation directly. Showing how separate innocuous details combine into an identification is more effective than telling people not to share things.

Address coercion in relationships explicitly, because the majority of pressure to share credentials and images comes from within them and safety education mostly does not mention it.

And treat exit as a tactical option rather than the answer. Temporary withdrawal during a specific incident is reasonable; permanent withdrawal as a general policy imposes a cost most adolescents will not accept, which is why the advice is ignored.

For adults giving the advice

The main thing worth understanding is that the reluctance to log off is not stubbornness or addiction. It is a correct assessment that the proposed remedy has a substantial cost which the person proposing it has not counted.

Advice that acknowledges the cost is taken more seriously than advice that pretends there is not one, and it opens the conversation about what would actually help — which is generally a set of specific configuration changes and a plan for a specific person, not a withdrawal from social life.