Corrections & Updates.

We will be wrong about something. When we are, you will see it — publicly, severity-classified, timestamped, and linked to the page it touched. No silent edits. No retroactive rewrites. No statute of limitations.

ScreenExtendersHub publishes buying-decision content across four authority pillars — Laptop Screen Extenders, Portable Monitors, Desktop Screen Extenders, and Docking Stations. Readers act on what we say. Wrong port counts, misstated OS compatibility, or outdated power-delivery numbers translate directly into wasted money and broken workflows. That responsibility is not abstract. The policy below is how we handle it.

It is built on three commitments that are not negotiable, regardless of commercial pressure or operational inconvenience:

One — Public visibility. Every correction at Tier 1, Tier 2, or Tier 3 is logged in the public ledger below, with the affected URL, the original error, the corrected information, and a timestamp. Tier 4 cosmetic fixes are silent because logging them would obscure rather than clarify.

Two — Severity over volume. A single Tier 1 spec error matters more than fifty typo fixes. The taxonomy in §03 is how we keep proportion. The taxonomy is also how we resist the affiliate-publishing temptation to either deflect serious errors as "clarifications" or inflate trivial fixes into apparent vigilance.

Three — The policy is recursive. This document is itself versioned. When we change how we handle corrections, that change is logged in §09, with the same severity treatment as any content change. The policy that governs the work is held to the same standard as the work.

Our broader editorial framework — testing protocols, scoring rubrics, sourcing standards — sits on the Methodology page. Our financial relationships are documented in the Affiliate Disclosure. Editorial accountability and authorship are addressed in About Us and About the Founder. This policy completes that system by addressing what happens after publication — the part most affiliate sites either omit entirely or treat as a footer afterthought.

Most affiliate sites wait for readers to find errors. We don't. Corrections enter our queue from five distinct sources, four of which are internal. The pipeline below is the system in operation:

The queue itself is not public — operational triage is internal — but every output that meets Tier 1, 2, or 3 thresholds resolves into a ledger entry below. Cross-reference: testing methodology and the sourcing rubric live on the Methodology page; technical terms used here (DisplayLink, DP Alt Mode, PD profiles) are defined in the Glossary.

Theory matters less than operation. The six panels below trace a single hypothetical Tier 1 correction from the moment a reader email arrives to the moment a permanent ledger entry exists. Each panel includes a specimen of the actual artifact generated at that stage. This is what no other affiliate-publishing corrections page surfaces — the system shown working.

The trace above is what every Tier 1 correction looks like in operation. Tier 2 traces follow the same six-panel structure with a 7-day SLA and a blue update banner. Tier 3 traces typically skip the on-page banner step. Tier 4 cosmetic fixes do not generate a ticket — they are corrected silently as discovered.

Every correction is classified into one of four tiers. The tier determines remediation speed, visibility on the affected page, and whether a public ledger entry is generated. Click a tier to see its full criteria, SLA, and ledger treatment.

The taxonomy is clean in theory. In operation, the work happens in the borderline cases. Five recurring edge cases are named below, each with an explicit ruling. These are how we resist the affiliate-publishing temptation to either inflate trivial fixes or deflect serious errors.

1. The manufacturer's own documentation is wrong.

   A reader reports an error in our review. The "error" is actually an inconsistency with the manufacturer's spec sheet — but the spec sheet itself is wrong. We tested the product directly and our measurement is correct.

   RulingThe review stands. We add a Tier 3 clarification noting the discrepancy with manufacturer documentation, and where feasible we contact the manufacturer to flag the spec-sheet error. The clarification protects readers who are comparing our review against the spec sheet — but we do not "correct" a measurement to match a wrong spec.

2. A reader-reported error is actually a disagreement.

   A reader reports an "error" in a verdict. The facts are correct; the reader simply reaches a different conclusion from the same evidence. They want our recommendation changed.

   RulingThis routes to Path B (§04 Disagreement), not the corrections queue. The feedback is logged internally for editorial review. No ledger entry. No banner. We document the disagreement openly when we respond to the reader, and we revisit the verdict in the next quarterly audit if the disagreement reflects a pattern across multiple reports.

3. Firmware ships that improves a product after a negative review.

   We published a critical review. Months later, the manufacturer ships a firmware update that addresses the specific issues we cited. The product is now meaningfully better than what we reviewed.

   RulingTier 2 update. The original review remains intact with original measurements preserved. A dated update notice appears at the top noting the firmware change and what it addressed. We do not retroactively rewrite the verdict; we add a new dated section reflecting current state. The reader sees both — the product as we found it then, and the product as it operates now.

4. An Amazon listing changes a product variant under the same URL.

   The affiliate link points to an Amazon listing. The listing has been silently edited to point to a different variant — different panel, different ports, different cable kit — but the URL and primary photo are unchanged.

   RulingTier 2 update if the new variant is comparable; Tier 1 retraction if the new variant is materially different from what we reviewed. The affiliate link is not silently swapped (see §07 item 08). Either we update the review to reflect the new variant explicitly, or we remove the link entirely. The reader who clicks must reach the product we evaluated, or be told plainly that the listing has changed.

5. A competing publication contradicts our findings.

   Another reviewer publishes findings that contradict ours on a measurable point — they report different brightness, different latency, different OS compatibility. Their methodology is documented; ours is documented. Both are plausible.

   RulingThis is not automatically a correction. We re-examine our test conditions, re-test where feasible, and either confirm our original measurement (no change), update with new measurements (Tier 2 if the world has shifted, or a footnoted note), or issue a Tier 1 correction if our original test was demonstrably flawed. We do not silently conform to the other publication's numbers to avoid embarrassment.

These rulings are not exhaustive. New edge cases will surface, and when they do, the ruling and its reasoning will be added here in a future policy version (see §09). The list is itself versioned.

Reader feedback arrives in three structurally different forms. Conflating them is the most common failure mode in affiliate-site corrections policies — readers end up with their disagreements treated as errors, or their reported errors dismissed as opinion. We route each separately:

The distinction matters because affiliate sites that treat all reader feedback as "corrections" inflate their ledgers in ways that obscure actual errors. The distinction also matters because manufacturers occasionally pressure publishers to reclassify Path A as Path B — a legitimate factual error gets recharacterized as a "disagreement about emphasis" so no public correction is logged. We don't.

Every Tier 1, Tier 2, and Tier 3 correction resolves here. Entries are permanent — never edited, never removed. The ledger filter below lets you isolate by severity. The first entry below files itself in real time on initial view as a small demonstration of how new entries arrive.

When a correction is published, this is what readers see at the top of the affected page. The visual treatment varies by tier so severity is communicated at a glance — before the reader has parsed a single word.

Tier 4 cosmetic fixes generate no notice at all — that is intentional. A page covered in fix-banners for typos would obscure the meaningful corrections. The threshold for visible notice is whether the change could affect a reader's understanding or decision.

The patterns below are common in affiliate publishing. They are dark patterns. We commit publicly that we do not engage in any of them, and we invite readers to flag suspected violations to the corrections inbox. Each item is followed by the defense — the commercial pressure that produces the pattern, and the editorial cost of refusing it.

• Silent score changes.

  A product's score will never be revised after publication without a logged Tier 1 or Tier 2 entry naming the change and the reason.

  Why this mattersSilent score changes are how affiliate sites quietly improve recommendations after manufacturer outreach or commission negotiations. The commercial pressure is direct: a higher score produces more clicks. Refusing this pattern means we sometimes carry a score lower than the latest manufacturer revision warrants until we publicly log the update. That is the cost of the bright line — and it is correct.
• Ghost deletions.

  Published reviews are not deleted. If a review is retracted, the URL remains live with a public retraction notice and a permanent ledger entry; the original content is preserved.

  Why this mattersThe temptation to delete a review the manufacturer dislikes is constant — especially when commission relationships are at stake. Ghost deletion erases the evidence trail and lets sites pretend coverage never happened. We keep the URL live, even for retractions, because readers who linked to or remembered the review deserve to see what changed and why. The cost is permanent visibility of work we may later regret. The benefit is permanent accountability.
• Retroactive ledger edits.

  Ledger entries, once published, are append-only. Subsequent clarifications are added as new dated notes within the entry — earlier text is never overwritten.

  Why this mattersAn editable ledger is not a ledger — it is a press release archive. The whole point of a public corrections record is that it cannot be quietly cleaned up later. Append-only enforcement means we live with the wording of past entries even when we would phrase them differently today. That is the price of credibility.
• Manufacturer-driven rating inflation.

  Scores are not increased in response to manufacturer outreach, threats, withheld review samples, or commercial pressure. The methodology drives the score; commercial relationships do not.

  Why this mattersThis is the most common form of editorial capture in affiliate publishing. Manufacturers withhold review samples from publications that score products poorly — a soft form of coercion that scales. Refusing means we sometimes lose access to early units. We accept the tradeoff. Scores reflect what we found, not what manufacturers prefer we found, full stop.
• Removal of "who should not buy this" sections.

  Negative-fit disclosures stay. Manufacturers occasionally request their removal; those requests are documented internally and declined.

  Why this mattersThe "who should not buy this" section is the single most reader-protective element of a buying-decision review. It is also the section manufacturers most often ask to be softened or removed. The pressure is real: removing it improves perceived sentiment, which improves commission. We do not remove it. Where the section is wrong, it gets corrected through the standard taxonomy. Where it is uncomfortable for the manufacturer, that is not our problem.
• Recommendation changes driven by commission rate.

  Affiliate commission structure does not influence which products we recommend. If the best answer is "don't buy this," we say so — see Disclosure.

  Why this mattersCommission rates vary across products and across time. The temptation to nudge readers toward higher-commission recommendations is structural — every affiliate site faces it. Our defense is procedural: recommendation language and "best for" tags are determined before commission rates are checked, not after. The Disclosure page documents the financial relationship; this bright line documents that the relationship does not flow back into editorial.
• Backdating corrections.

  Every ledger entry carries the actual date of correction publication, not the date the error was introduced. Timestamps are real.

  Why this mattersBackdating is how slow corrections get cosmetically reframed as fast ones. A page edited two months after an error was reported gets a "corrected on the same day" timestamp, and the reader has no way to verify the gap. We publish actual remediation dates — including the inconvenient ones where we missed the SLA. Logged late corrections are honest; backdated fast corrections are theater.
• Hidden affiliate-link substitution.

  A link will never be silently swapped from one product to another (e.g., from a discontinued model to a newer "equivalent") without a logged Tier 2 entry. The product the reader clicks must be the product the review evaluated.

  Why this mattersLink substitution is technically trivial — change one URL, the page looks identical. It is also how readers end up buying products that were never reviewed under the impression that they were. We treat the affiliate link as a load-bearing claim: clicking it must take the reader to the unit we tested, or the link must be removed and the substitution must be logged. There is no third option.
• Stealth de-recommendation.

  A product will never be quietly removed from a roundup or "best of" list without a logged ledger entry naming the removal and the reason. Roundup composition is editorially accountable.

  Why this mattersRoundups change. New products launch, old products age out, manufacturers fall out of favor. The temptation is to update the roundup silently and let the new composition speak for itself. We log removals openly because readers may have made decisions based on prior compositions, and because silent roundup churn is how affiliate sites swap recommendations to chase commission without ever publicly explaining the change.
• Manufacturer-supplied measurements presented as our own.

  Test data, brightness numbers, latency figures, and battery-life claims are either independently measured or attributed to their source. Manufacturer specs will never be passed off as our findings.

  Why this mattersThe single fastest way to produce review content is to copy manufacturer spec sheets and frame them as testing results. The reader cannot tell the difference unless attribution is explicit. We mark every measurement: independently measured, or sourced — and we never blur the line. Our Methodology page documents which measurements we own.
• AI-generated content presented as expert review.

  Reviews and buying-decision content are produced under named human editorial judgment. AI tools may assist with drafting, research synthesis, or copy editing, but no published recommendation is generated without human verification of every load-bearing claim. We will never publish a fully AI-authored review under an editorial byline.

  Why this mattersThe economic incentive to publish AI-authored reviews under human bylines is enormous and growing. The reader cost is invisible until the AI confidently asserts something false — a port count, a compatibility claim, a feature that does not exist. We anchor the bright line at human verification of every load-bearing claim, regardless of which tools assisted the drafting. The byline names the human who is accountable.
• Reverse-engineering trust through "neutral" framing.

  A review will never be written to appear neutral while being structured to funnel readers toward a higher-commission product. Recommendations, "best for" tags, and verdict language are determined before commission rates are checked, not after.

  Why this mattersThis is the most sophisticated form of affiliate manipulation: the language and structure of editorial neutrality, used to disguise commercially-motivated steering. The pattern is hard to detect because every individual sentence reads as fair. The defense is procedural: editorial decisions sequence before commission lookup, and we document that sequence on this page so that readers can hold us to it. If a "best for" tag ever lands on a higher-commission product against the methodology's findings, that is a reportable violation.

Reader reports are Source 01 in the detection pipeline (§02). To keep them from being lost in the general inbox, corrections have a dedicated channel with a stated review cadence:

This policy is itself versioned. When the document changes — taxonomy revisions, SLA adjustments, new trigger sources, scope changes — those revisions are logged below. The corrections policy is held to the same accountability standard as the content it governs. Toggle the hypothetical preview below to see what a future v2.0 entry would look like.

Future versions will appear above v1.0 in reverse-chronological order. Each version increment notes what changed and why. Material policy changes (taxonomy revisions, SLA adjustments) are themselves classified through this taxonomy — a Tier 2 banner runs on the policy page for 60 days following each material change. The policy that governs the work is held to the same standard as the work.

The corrections policy classifies itself. Below is this document's own entry in its own version history, signed and dated using the same fields any ledger entry carries. If a future revision changes the severity taxonomy, the SLAs, or any of the bright lines, that change generates an entry in §09 and a banner on this page — same treatment any review on the site would receive.

If you find this document itself in error — a contradiction between sections, an unstated commitment, an undocumented edge case — that is a Tier 1 candidate for the policy on the policy. Report through the standard corrections channel. The recursive accountability is not rhetorical.