Replication suggestion: C2PA provenance labels and trust in digital news
Work to replicate
Trattner et al. (2026), C2PA Provenance Labels Increase Trust in Digital News
Platforms Across Western Countries, Proceedings of the International AAAI
Conference on Web and Social Media (ICWSM 2026).
Why this is worth an independent replication
The paper reports a practically important, cross-country result about whether
C2PA-style provenance labels affect trust in news platforms. Independent
replication would be valuable because provenance interfaces are increasingly
used as a trust intervention, while small changes in label wording, detail,
country, platform context, and participant interpretation could change the
effect. A clean reimplementation would also test whether the main effect can be
recovered without relying on the original experiment code.
Materials currently visible
As of 2026-08-02, the public OSF component exposes the article preprint, but I
could not find public analysis code, participant-level data, or the full
experiment implementation there. The first step should therefore be to ask the
authors whether the remaining materials can be shared. If they cannot, a
transparent reconstruction from the paper should document every inferred or
newly implemented choice.
Proposed confirmatory scope
- independently implement the authentic-content study described in the paper;
- retain a no-label baseline and the original label/detail conditions as far as
the published materials permit;
- preregister hypotheses, exclusions, sample size, stopping rule, outcomes,
contrasts, and the exact reconstruction decisions before data collection;
- recruit a new sample and report the original effect estimate alongside
uncertainty and a predeclared equivalence or smallest-effect-size analysis;
- publish the instrument, stimuli where licensing permits, anonymized data,
analysis code, environment lockfile, and a reproducible report;
- keep exploratory extensions separate from the confirmatory replication.
Feasibility and interest
I am interested in leading or contributing to this replication, contingent on
ethics approval, access to legally reusable materials, and a feasible sampling
budget. The implementation and analysis would be open source. This issue is a
replication suggestion, not a claim that recruitment or data collection has
already begun.
Scope check requested
The reported result is behavioural, although an independent replication would
require a new open experiment implementation and computational analysis. Before
any ReScience C manuscript work, I would welcome an editorial view on whether
that combination qualifies as “computational research” under the journal's
scope. If not, the suggestion can still identify a valuable open replication,
but it should be routed to an HCI/open-science venue rather than represented as
a ReScience C submission.
Replication suggestion: C2PA provenance labels and trust in digital news
Work to replicate
Trattner et al. (2026), C2PA Provenance Labels Increase Trust in Digital News
Platforms Across Western Countries, Proceedings of the International AAAI
Conference on Web and Social Media (ICWSM 2026).
Why this is worth an independent replication
The paper reports a practically important, cross-country result about whether
C2PA-style provenance labels affect trust in news platforms. Independent
replication would be valuable because provenance interfaces are increasingly
used as a trust intervention, while small changes in label wording, detail,
country, platform context, and participant interpretation could change the
effect. A clean reimplementation would also test whether the main effect can be
recovered without relying on the original experiment code.
Materials currently visible
As of 2026-08-02, the public OSF component exposes the article preprint, but I
could not find public analysis code, participant-level data, or the full
experiment implementation there. The first step should therefore be to ask the
authors whether the remaining materials can be shared. If they cannot, a
transparent reconstruction from the paper should document every inferred or
newly implemented choice.
Proposed confirmatory scope
the published materials permit;
contrasts, and the exact reconstruction decisions before data collection;
uncertainty and a predeclared equivalence or smallest-effect-size analysis;
analysis code, environment lockfile, and a reproducible report;
Feasibility and interest
I am interested in leading or contributing to this replication, contingent on
ethics approval, access to legally reusable materials, and a feasible sampling
budget. The implementation and analysis would be open source. This issue is a
replication suggestion, not a claim that recruitment or data collection has
already begun.
Scope check requested
The reported result is behavioural, although an independent replication would
require a new open experiment implementation and computational analysis. Before
any ReScience C manuscript work, I would welcome an editorial view on whether
that combination qualifies as “computational research” under the journal's
scope. If not, the suggestion can still identify a valuable open replication,
but it should be routed to an HCI/open-science venue rather than represented as
a ReScience C submission.