How Should Researchers Handle Deleted Posts or Content That Was Public When It Was Collected?

What happens when research data outlive the original post?

A researcher collects a publicly accessible social media post. At the time of collection, anyone could see it. Months later, the author deletes the post, makes the account private, removes the entire profile, or asks a platform to erase the content.

The research team still has its copy.

Must the researcher delete it too? Can it still be analyzed? Can it be quoted in a paper even though readers can no longer see the original? Does deletion amount to withdrawal from research when the author never knew about the study in the first place?

There is no universal rule that makes every deleted online item unusable. But deletion changes information that matters ethically. A defensible decision should consider both the conditions under which the material was originally collected and what its subsequent disappearance may mean for continued retention and publication.

Deletion does not automatically invalidate collection, but it should trigger reassessment

Online data can change ethical status after collection

Researchers should separate collection from continued use

A post's status at collection matters. If information was intentionally available to an unrestricted public, that fact may support its treatment as public information under an applicable research framework. UKRI, for example, states that information intentionally made public in internet spaces may be considered in the public domain, while still requiring researchers to examine privacy and identifiability critically.

But an ethically defensible collection decision does not necessarily settle every later use of the material.

The answers can be related without being identical.

Deletion can communicate something without communicating everything

People delete online material for many reasons. They may regret the post, correct an error, reduce their digital footprint, leave a platform, resolve a dispute, protect themselves from harassment, change privacy preferences, or simply tidy an account.

Researchers should therefore be cautious about assigning a single motive to deletion.

At the same time, deletion is evidence that the author no longer wishes the platform to display that content in the same way. That changed context can matter when researchers are deciding whether to reproduce the material in a publication, particularly if it concerns sensitive experiences or identifiable individuals.

The ethically useful position lies between two extremes: deletion is neither meaningless nor automatically equivalent to formal withdrawal from every research use.

Deletion is not necessarily research withdrawal

Withdrawal from research ordinarily occurs within a research relationship governed by a consent process or another established mechanism. A person who deletes a public post may have no idea that a researcher previously collected it.

It would therefore be misleading to claim automatically that deletion constitutes a formal withdrawal request from a study the person never knew existed.

Conversely, researchers should not use the absence of formal withdrawal as a reason to ignore the author's changed privacy behavior. Ethical reflection can extend beyond the minimum formal mechanism available to participants.

Consent documents and protocols may already determine what happens

If participants were recruited and informed that their online material would be collected, the consent process may specify withdrawal rights, data-retention periods, and whether already-analyzed data can be removed.

Researchers should follow those commitments.

The situation is less straightforward when researchers collected public social media posts as existing data without contacting their authors. In that case, the ethics protocol should ideally have anticipated changes in availability rather than leaving researchers to invent a rule during manuscript preparation.

Deleted content may become more identifying when researchers publish it

This sounds counterintuitive. If the original disappears, shouldn't identification become harder?

Sometimes it does. But publication can also become the only durable public reproduction of material that its author attempted to remove. A research paper may preserve a statement long after the original context has vanished.

Archived copies, screenshots, search-engine caches, reposts, or quotation databases may also retain the original text. A distinctive quotation can still potentially reconnect the publication to the author.

The question of whether online quotations can identify their authors therefore remains relevant even after the source is deleted.

Exact quotation deserves particular scrutiny

There is an ethical difference between retaining deleted material internally for an aggregate analysis and reproducing that material verbatim in a publication.

Researchers should avoid treating all these uses as ethically equivalent simply because the original collection event was the same.

Paraphrasing may sometimes preserve evidence without republishing deleted wording

If the analytical contribution lies in the meaning of a post rather than its exact linguistic form, researchers may consider paraphrasing the online material.

This can reduce direct searchability and avoid restoring the author's exact deleted words to public circulation. It is not appropriate when exact wording is itself the phenomenon under analysis, and it does not eliminate identification through contextual details.

Deleting a post does not necessarily delete all copies

Online content can persist in screenshots, archives, backups, caches, reposts, quoted replies, downloaded datasets, and research files. Researchers should not promise that deletion from the original platform causes every copy to disappear unless they can actually ensure that result.

This persistence is one reason online research protocols should explain retention clearly where participants are recruited directly.

It also demonstrates the broader distinction between technical accessibility and ethically appropriate reuse. Researchers may possess a copy that is no longer publicly accessible elsewhere.

Data minimization still applies after collection

Researchers sometimes retain complete datasets indefinitely because storage is inexpensive and future analyses are imaginable.

A better question is whether continued possession remains necessary for the defined research purpose, research integrity, verification, contractual obligations, or legitimate future use.

The Association of Internet Researchers' ethics guidance encourages researchers to consider whether collected data are necessary and proportionate to the research aim and whether unnecessary information should be deleted. This lifecycle perspective is particularly relevant when the original source has disappeared.

Deletion can itself be research data, but only when justified

Some studies legitimately examine deletion. Researchers may study content moderation, self-censorship, misinformation correction, account closure, or how public discourse changes over time.

In those studies, the disappearance of a post may be analytically important.

That does not eliminate privacy concerns. Researchers should establish whether tracking deleted content forms part of the approved methodology and whether retaining the original material is necessary to answer the question.

Changed privacy settings create a similar problem

A user may not delete anything but may convert an unrestricted account into a private one.

The research team may already possess material collected when it was public. Again, the later change does not automatically rewrite the conditions that existed at collection. It does provide new information about the user's current access preferences.

Researchers should consider that change when deciding whether identifiable or verbatim material needs to be republished.

Law and platform rights may impose requirements beyond research ethics

Data-protection laws, erasure rights, platform agreements, contractual conditions, copyright, archival obligations, and institutional retention rules differ among jurisdictions and studies.

Researchers should not assume that an ethics argument alone establishes a legal right to retain deleted content indefinitely. Nor should they assume that a platform's deletion function automatically imposes a legal duty on every independent research copy.

A post disappears before the paper is published

Another study might reasonably reach a different decision if the exact wording or deletion event itself were the object of research. The relevant point is that the changed context prompted reassessment rather than an automatic response.

Deleted content is neither automatically forbidden nor permanently fair game

Build changing online availability into the protocol from the beginning

Online datasets are not static. Posts disappear, accounts become private, usernames change, and communities close. Plan for that before collection rather than treating every change as an unexpected ethics emergency.

The question is not merely what you were allowed to collect yesterday. It is what you can justify retaining and exposing today.

When online content disappears, reassess the research copy

Questions about deleted online research data

Public once does not necessarily mean unrestricted forever

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.