# The Permanence Paradox: Research Prospectus and National Research Program **The Archive Ethics Project** · Version 1.0 · 2026-10-11 **Status:** Working paper, not yet reviewed outside the project **Source:** https://archiveethics.org/documents/permanence-paradox-prospectus.md **Retrieved:** 2026-10-12 ## Abstract Sets out the central and supporting research questions, competing perspectives to investigate, intended scholarly outputs, and the design of three studies: a systematic national content analysis, an institutional transparency survey, and a longitudinal archival practice analysis. --- ## Research questions ### Working study title **The Permanence Paradox: Historical Accuracy, Digital Discoverability, and Editorial Accountability in American Journalism** ### Central research question To what extent do American news organizations maintain transparent, consistent, and ethically defensible policies governing historical reporting, subsequent factual developments, and continuing search-engine discoverability? ### Supporting research questions **RQ1. Policy prevalence:** What proportion of eligible organizations have a publicly documented policy, an internal policy, an informal practice, no policy, or an undetermined status? **RQ2. Available remedies:** Which organizations expressly permit, prohibit, or evaluate individually corrections, contextual updates, anonymization, de-indexing, and removal? **RQ3. Governing authority:** Who makes the final decisions, and how do these arrangements differ between professional newsrooms and student-run outlets? **RQ4. Subsequent outcomes:** How do written policies treat dismissals, acquittals, vacaturs, expungements, institutional findings, and other later developments? **RQ5. Archival practice:** How often can observed changes to historical content be documented, and what can the available evidence legitimately establish about why changes occurred? **RQ6. Transparency and procedure:** How commonly do publications describe intake procedures, decision criteria, response timelines, appeal paths, and dated policy histories? **RQ7. Ethical theory:** How do preservation, accountability, privacy, compassion, rehabilitation, and continued public interest interact when the discoverability of old reporting is assessed? ### Competing perspectives to investigate AEP must fairly present both the case for maintaining search accessibility and the case for individualized reconsideration. Counterarguments include the risk of obscuring truthful reporting, creating privileges for influential or persistent requesters, weakening public accountability, treating similarly situated individuals inconsistently, and relying on technically imperfect search controls. Arguments for individualized review include later legal developments, changed public interest, the permanence of name-based search, disproportionate harm, and the ability to preserve the underlying archive without identical external discoverability. The initial question is **how policies and practices operate**, not whether the dataset can be assembled into a favorable argument for a particular person. ### Intended scholarly outputs - A formal research prospectus and analytical literature review. - A preregistered protocol for *future* collection, with sampling and coding instruments. - A pilot content-analysis study and reliability report. - A national institutional survey with full response-rate disclosure. - A longitudinal archive-practice report with methodological uncertainty. - Reproducible research datasets, codebooks, and analysis scripts. - An annual *State of News Archive Ethics* report. - Independently reviewed institutional case studies. - A model policy accompanied by scholarly commentary justifying each major provision. - A substantive counterargument paper: **The Case Against De-indexing: Historical Preservation, Public Accountability, and the Risks of Editorial Revision**. **Publication-status convention:** Label internal working papers, externally reviewed reports, and articles actually accepted through academic peer review distinctly. Calling a website report a “thesis” does not confer university approval, a degree, or peer-reviewed status. ## National research program ### Study I: Systematic national content analysis **Population:** Active U.S.-based student and professional news publications meeting predetermined inclusion criteria. Define the institutional sampling frame, unit of analysis, treatment of chains and jointly governed outlets, and exclusions before estimating prevalence. **Design:** Use a defensible probability sample if making national prevalence estimates; otherwise use a transparently described census attempt or nonprobability sample and avoid extrapolating beyond observed institutions. Stratify or report separately by professional versus student newsroom and potentially size, geography, governance, and medium when reliable metadata exist. **Coding:** Publish a codebook defining policy visibility, allowable remedies, decision authority, response procedures, later legal outcomes, and the difference between explicit prohibitions and unknown positions. Use at least two independent coders on a meaningful overlapping sample. Document disagreements and adjudication. Report intercoder reliability using a suitable metric, such as Krippendorff's alpha where appropriate. **Outputs:** Denominator-aware descriptive findings, subgroup comparisons with uncertainty measures when justified, reproducible summaries, and a documented set of classifications. ### Study II: Institutional transparency and practice survey Send a standardized, neutral questionnaire to the appropriate newsroom contact. Inquire about policy existence, publication, remedies, procedures, authority, and policy history. Do not imply that nonparticipation is a negative result. Maintain outreach evidence and response timestamps. An initial response window of roughly 10 business days, one reminder, and an additional 10 business days is a proposed process, not a scientific requirement. Publish response and nonresponse counts, contact limitations, selection criteria, question wording, and relevant biases. Evaluate whether the research requires human-subjects or institutional ethics review based on its design, participants, affiliations, data collected, and applicable rules. Do not promise confidentiality that the project cannot technically or legally protect. ### Study III: Longitudinal archival practice analysis Use available website captures, archived versions, metadata, canonical URLs, and technical signals to identify *candidate* changes to historical articles and policies. The existing NSMPA tool and the founder's earlier discovery tools can support discovery but should not be treated as independent verification. For each potentially meaningful change, assess whether differences might arise from website migration, capture errors, templating, broken links, a URL redirect, content-management redesign, syndication, or actual editorial revision. Maintain source snapshots or preservation references, observation dates, and a reviewer rationale. A de-indexed page, deleted URL, and inaccessible archived capture are different phenomena. A comparison between two archived versions can show an observed difference. It cannot by itself prove the original editor's intent, the adoption date of an unpublished policy, or a policy violation. ### Preregistration and exploratory legacy research The project's past exploratory discovery and analysis should be labeled **exploratory**. New confirmatory protocols may be prospectively registered at the Open Science Framework (OSF) before collecting the relevant data. Do not retrospectively claim preregistration for existing findings.