To appraise records for selection, judge their enduring value — evidential, informational, legal/fiscal, and research — and keep only what justifies the long-term cost of preservation. Work in aggregate (series or file level, not item by item), apply explicit criteria consistently, and document every decision, because destruction is irreversible and the record of your reasoning is what makes the choice defensible. Appraisal is the most consequential judgement an archivist makes; treat it as a deliberate, documented process, not weeding.

What values am I actually weighing?

Schellenberg's framework still anchors practice. You assess two broad value types:

• Primary value — usefulness to the creator (administrative, legal, fiscal). Usually time-limited.
• Secondary value — usefulness to others over time, split into:
  • Evidential value: what the records prove about the creator's own organisation and functions.
  • Informational value: what they reveal about people, places, events beyond the creator.

Add research value (likely use by future researchers) and any continuing legal or fiscal obligations. A record high in informational value but routine to the creator can still merit permanent retention.

How do I run an appraisal workflow end to end?

1. Understand the creator and functions. Read the administrative history; the most significant functions usually generate the most valuable records (this is the macro-appraisal insight).
2. Survey the records in aggregate. Box-level and series-level review first; sample within series.
3. Apply explicit criteria (below) to each series, not each item.
4. Identify duplication and ephemera — drafts, duplicates, routine transactional records with no evidential weight.
5. Make and record the decision — retain, destroy, or sample.
6. Apply a retention/destruction authority where one exists.
7. Document everything before anything is destroyed.

What criteria should I apply, and how do I weight them?

A practical scoring grid keeps decisions consistent across a large accession:

Criterion	Question	Weight in practice
Evidential value	Does this prove how the creator functioned?	High
Informational value	Does it inform on wider people/events?	High
Uniqueness	Is the information held nowhere else?	High
Research demand	Are these records likely to be used?	Medium
Legal/fiscal need	Is retention legally required?	Variable (can be decisive)
Cost/condition	Is preservation feasible and proportionate?	Medium

retain if (evidential OR informational) AND unique
sample  if bulky AND repetitive AND value is statistical
destroy if duplicate OR transactional-ephemera AND no legal hold

When should I sample rather than keep or destroy everything?

Sampling fits bulky, repetitive series where the value is statistical rather than in any single file — case files, claim forms, registration records. Options:

• Systematic (every nth file) — simple, defensible, preserves a representative spread.
• Random — statistically sound for quantitative research.
• Purposive — keep exemplars or notable cases plus a baseline sample.

Document the sampling method and rate so future researchers understand what survives and why.

How do I document a decision so it holds up later?

Appraisal documentation is not optional; it is the accountability trail. Capture, per accession:

• What was reviewed (extent, series, dates).
• Criteria applied and the reasoning.
• What is retained, destroyed, or sampled, with quantities.
• The destruction authority and any legal holds checked.
• Who decided, who authorised, and the date.

Store this with the accession record. Without it, a future reappraisal or audit cannot reconstruct why material was destroyed — and destruction cannot be undone.

What are the common appraisal mistakes?

• Item-level appraisal by default, which is slow and rarely changes the outcome.
• Keeping everything "to be safe" — this just defers the cost and buries value in bulk.
• Destroying without documentation, leaving no defensible record.
• Ignoring legal/fiscal holds before destruction.
• Treating duplicates as originals, inflating retention needlessly.

Key Takeaways

• Weigh evidential, informational, legal/fiscal and research value to judge enduring worth.
• Appraise in aggregate (series/file), not item by item, except for small high-stakes material.
• Use macro-appraisal: analyse the creator's functions first, then target the records that document the most significant ones.
• Sample bulky, repetitive series with a documented method and rate.
• Document every decision — what, why, how much, who, when, and the destruction authority — before destroying anything.
• Avoid both "keep everything" and undocumented weeding; destruction is irreversible.

Frequently Asked Questions

What is archival appraisal?

Appraisal is the process of judging which records have enough enduring value — evidential, informational, legal or research — to justify permanent retention, and which can be destroyed. It is a core, irreversible archival decision.

What is the difference between evidential and informational value?

Evidential value is what records prove about the creator's own functions and actions. Informational value is what records reveal about people, places and events beyond the creator. Schellenberg distinguished the two.

Should I appraise at item level or in aggregate?

Appraise in aggregate wherever possible — at series or file level — because item-by-item appraisal is slow and rarely improves the decision. Reserve item-level appraisal for small, high-stakes or very mixed material.

How do I document an appraisal decision?

Record what you looked at, the criteria applied, what you kept and destroyed, who decided and when, and the authority for destruction. This documentation is essential for accountability and future audit.

Can I destroy records once they are in the archive?

Reappraisal and deaccessioning are legitimate but must be done deliberately, against policy, with documentation and appropriate authorisation. They are not routine and should never be casual weeding.

What is macro-appraisal?

Macro-appraisal, associated with Terry Cook and the Canadian tradition, appraises records by analysing the functions and significance of the creating body first, then targeting the records that document the most important functions.

Related reading

• Beginner's Guide to Accession records
• Troubleshooting: Apply More Product Less Process
• Best Practices to Arrange a fonds into series
• Appraise born-digital records