Source-expiry review for fitness research publishing

Research question and scope

How should an editorial team decide whether a research source is still fit for a bounded article claim? The unit of analysis is one source-claim pair, including the source version, access date, population, period, and wording used. The method basis is a source register with a review interval, a change flag, and a decision record that distinguishes correction, update, supersession, and simple link failure.

Supported findings

A source can remain authoritative while a particular statistic becomes outdated or a page changes its definitions. Review evidence as a claim-specific record, not as a promise that a URL will remain unchanged.

Evidence note 1

A new survey year can change a percentage without making the old observation false.

Evidence note 2

An updated guidance page may keep the recommendation while changing explanatory context.

Evidence note 3

A projection is tied to its base year and forecast period.

Evidence note 4

A broken link is a discoverability problem, not evidence that a claim has been disproved.

Evidence note 5

A source register makes uncertainty visible before an editor republishes a number.

Evidence note 6

A new survey year can change a percentage without making the old observation false. For this question, note 1 should be recorded before any summary is written. The reviewer should preserve the original observation, identify who or what was eligible to appear in it, and state which part of the claim the note can support. A later reader should not have to infer whether the item was measured directly, reported by a participant, copied from a schedule, or supplied by an external publication.

Evidence note 7

An updated guidance page may keep the recommendation while changing explanatory context. For this question, note 2 should be recorded before any summary is written. The reviewer should preserve the original observation, identify who or what was eligible to appear in it, and state which part of the claim the note can support. A later reader should not have to infer whether the item was measured directly, reported by a participant, copied from a schedule, or supplied by an external publication.

Evidence note 8

A projection is tied to its base year and forecast period. For this question, note 3 should be recorded before any summary is written. The reviewer should preserve the original observation, identify who or what was eligible to appear in it, and state which part of the claim the note can support. A later reader should not have to infer whether the item was measured directly, reported by a participant, copied from a schedule, or supplied by an external publication.

Evidence note 9

A broken link is a discoverability problem, not evidence that a claim has been disproved. For this question, note 4 should be recorded before any summary is written. The reviewer should preserve the original observation, identify who or what was eligible to appear in it, and state which part of the claim the note can support. A later reader should not have to infer whether the item was measured directly, reported by a participant, copied from a schedule, or supplied by an external publication.

Evidence note 10

A source register makes uncertainty visible before an editor republishes a number. For this question, note 5 should be recorded before any summary is written. The reviewer should preserve the original observation, identify who or what was eligible to appear in it, and state which part of the claim the note can support. A later reader should not have to infer whether the item was measured directly, reported by a participant, copied from a schedule, or supplied by an external publication.

Evidence note 11

A second-pass check for research-question-content-source-expiry asks a different question about item 1: A new survey year can change a percentage without making the old observation false. If this condition changes the denominator, timing, or meaning of the record, annotate the break rather than smoothing it away. If it does not change the stated measure, explain why that judgment is defensible. This makes the evidence auditable without implying that the local record represents everyone who was absent from it.

Evidence note 12

A second-pass check for research-question-content-source-expiry asks a different question about item 2: An updated guidance page may keep the recommendation while changing explanatory context. If this condition changes the denominator, timing, or meaning of the record, annotate the break rather than smoothing it away. If it does not change the stated measure, explain why that judgment is defensible. This makes the evidence auditable without implying that the local record represents everyone who was absent from it.

Evidence note 13

A second-pass check for research-question-content-source-expiry asks a different question about item 3: A projection is tied to its base year and forecast period. If this condition changes the denominator, timing, or meaning of the record, annotate the break rather than smoothing it away. If it does not change the stated measure, explain why that judgment is defensible. This makes the evidence auditable without implying that the local record represents everyone who was absent from it.

Evidence note 14

A second-pass check for research-question-content-source-expiry asks a different question about item 4: A broken link is a discoverability problem, not evidence that a claim has been disproved. If this condition changes the denominator, timing, or meaning of the record, annotate the break rather than smoothing it away. If it does not change the stated measure, explain why that judgment is defensible. This makes the evidence auditable without implying that the local record represents everyone who was absent from it.

Evidence note 15

A second-pass check for research-question-content-source-expiry asks a different question about item 5: A source register makes uncertainty visible before an editor republishes a number. If this condition changes the denominator, timing, or meaning of the record, annotate the break rather than smoothing it away. If it does not change the stated measure, explain why that judgment is defensible. This makes the evidence auditable without implying that the local record represents everyone who was absent from it.

Evidence note 16

Interpretive memo 1: A new survey year can change a percentage without making the old observation false. Compare this observation only with records collected under the same rule. A changed instrument, revised question, different class or service boundary, or new follow-up interval can create a discontinuity. Mark that discontinuity in the table and describe the likely effect on interpretation. The honest result may be a smaller claim, an explicit unknown, or a request for a qualified reviewer.

Evidence note 17

Interpretive memo 2: An updated guidance page may keep the recommendation while changing explanatory context. Compare this observation only with records collected under the same rule. A changed instrument, revised question, different class or service boundary, or new follow-up interval can create a discontinuity. Mark that discontinuity in the table and describe the likely effect on interpretation. The honest result may be a smaller claim, an explicit unknown, or a request for a qualified reviewer.

Evidence note 18

Interpretive memo 3: A projection is tied to its base year and forecast period. Compare this observation only with records collected under the same rule. A changed instrument, revised question, different class or service boundary, or new follow-up interval can create a discontinuity. Mark that discontinuity in the table and describe the likely effect on interpretation. The honest result may be a smaller claim, an explicit unknown, or a request for a qualified reviewer.

Evidence note 19

Interpretive memo 4: A broken link is a discoverability problem, not evidence that a claim has been disproved. Compare this observation only with records collected under the same rule. A changed instrument, revised question, different class or service boundary, or new follow-up interval can create a discontinuity. Mark that discontinuity in the table and describe the likely effect on interpretation. The honest result may be a smaller claim, an explicit unknown, or a request for a qualified reviewer.

Evidence note 20

Interpretive memo 5: A source register makes uncertainty visible before an editor republishes a number. Compare this observation only with records collected under the same rule. A changed instrument, revised question, different class or service boundary, or new follow-up interval can create a discontinuity. Mark that discontinuity in the table and describe the likely effect on interpretation. The honest result may be a smaller claim, an explicit unknown, or a request for a qualified reviewer.

Bounded conclusion

How should an editorial team decide whether a research source is still fit for a bounded article claim? The defensible endpoint is a documented observation with its source context and uncertainty attached; any stronger claim would exceed the stated unit or method.

Sources and verification

1. https://www.who.int/publications/i/item/9789240015128
2. https://www.cdc.gov/physical-activity/php/about/index.html
3. https://www.nist.gov/privacy-framework