In re Class Action Settlement Administration Litigation, 2026 WL 2444936 (D.D.C. Aug. 19, 2026), the court rejected a request for "three-strata validation" in discovery production. This decision clarifies which validation steps meet Rule 26(g) obligations and which lead to unnecessary costs and duplication. If you're designing validation protocols for complex litigation, this checklist translates the court's reasoning into actionable steps.
What This Checklist Covers
This checklist focuses on validation requirements for document productions in federal civil litigation. It distinguishes between mandatory completeness testing (identifying false negatives) and optional precision testing (identifying false positives). Each item reflects current case law on proportionality and Rule 26(g) certification requirements.
Use this when finalizing a production protocol, responding to opposing counsel's validation demands, or defending your methodology before a magistrate judge.
Prerequisites
Before starting validation, ensure you have:
- A documented collection scope. Define custodians, date ranges, and data sources in writing, preferably in a joint ESI protocol.
- Search methodology on record. Whether you used keywords, predictive coding, or manual review, document the approach and rationale.
- Sampling capability. Your review platform should generate statistically valid random samples from designated document sets.
- Baseline metrics. Know the size of your collected universe, the number of documents coded responsive, and the number excluded pre-review.
Good looks like: A protocol signed by both parties that identifies which document sets will be sampled, at what confidence level, and at what stage of the review process.
Validation Checklist
1. Conduct Null Set Testing on Non-Responsive Documents
Review a statistically valid random sample of documents coded as non-responsive. This "elusion testing" ensures responsive documents aren't erroneously omitted from your production.
Victor Stanley, Inc. v. Creative Pipe, Inc., 250 F.R.D. 251, 261 (D. Md. 2008), established that search validation requires more than attorney assertions. The In re Class Action court called null set testing standard practice.
Good looks like: A written report stating sample size, confidence interval, number of false negatives found, and whether the false-negative rate falls within acceptable limits (typically under 5% at 95% confidence). If the rate is too high, revise search terms or review additional documents.
2. Validate the Excluded Set Before Merits Review Begins
If you excluded documents from the review universe (by date range, file type, or custodian), sample that Excluded Set before coding documents as responsive or non-responsive. Confirm the exclusion criteria didn't inadvertently filter out responsive material.
The In re Class Action court permitted this early-stage validation but rejected requests to re-validate the Excluded Set after production. The court found re-validation "duplicative and unnecessary" when parties had already agreed to statistical sampling at collection.
Good looks like: Sampling results documented in a meet-and-confer letter or joint status report, showing that excluded documents contain negligible responsive content. If sampling reveals problems, expand the collection scope before review begins.
3. Certify Completeness Under Fed.R.Civ.P. 26(g)
When you produce documents, your counsel signs a certification that the response is complete and correct to the best of their knowledge. This certification focuses on false negatives (missing responsive documents), not false positives (overproduction).
The court in In re Class Action noted that Rule 26(g) does not prohibit overproduction. Validation of documents already coded responsive "reveals only the production's false-positive rate" and doesn't identify whether documents are missing.
Good looks like: A privilege log and production letter that references your validation testing and states the basis for your completeness certification. Be prepared to defend the methodology if challenged, but you're not required to re-review responsive documents to weed out marginal hits.
4. Document Your Validation Methodology in Writing
Prepare a validation report or declaration that explains sample size, confidence levels, findings, and any corrective actions. If opposing counsel challenges your production, you'll need to show the court that your process was defensible.
"Ipse dixit pronouncements from lawyers unsupported by an affidavit or other showing that the search methodology was effective" carry little weight, per Victor Stanley.
Good looks like: A declaration from your eDiscovery vendor or senior litigation support analyst stating: "We reviewed a random sample of 385 documents from the null set (95% confidence, ±5% margin of error). We identified 12 false negatives (3.1% error rate), which falls within acceptable limits. No further collection is warranted."
5. Resist Duplicative Validation Demands
If opposing counsel requests validation of responsive documents (to check for false positives) or re-validation of the Excluded Set after production, cite the In re Class Action decision. The Federal Rules don't require you to prove you haven't overproduced.
However, evaluate whether the request is truly duplicative. If you didn't validate the Excluded Set at collection, you may have to do it now.
Good looks like: A meet-and-confer response that distinguishes between validation you've already completed (null set testing, early Excluded Set sampling) and validation that falls outside Rule 26(g) obligations (responsive-document precision checks). Offer to provide your existing validation reports rather than conducting new sampling.
Common Mistakes
Skipping null set testing entirely. Some teams assume keyword search or predictive coding is self-validating. It's not. Every search must be tested against the non-responsive set.
Validating only the responsive set. Checking for false positives doesn't prove your production is complete. Courts care about false negatives.
Waiting until after production to validate the Excluded Set. If you discover responsive documents in the Excluded Set post-production, you face a costly re-collection and potential sanctions for incomplete discovery. Validate exclusions early.
Failing to document sample size and confidence intervals. "We reviewed some documents and they looked fine" won't survive a challenge. Use statistical sampling software and record the parameters.
Agreeing to "substantial completeness validation" without defining it. The term is ambiguous. If opposing counsel proposes it, clarify whether they mean null set testing (which you'll do) or re-validation of excluded documents (which you'll resist as duplicative).
Next Steps
After completing this checklist, prepare a validation summary for your litigation team. Include sample sizes, error rates, and any adjustments you made to search terms or collection scope. If opposing counsel raises validation disputes, you'll have a record that shows your process was proportional, defensible, and aligned with current case law.
If you're negotiating an ESI protocol, propose null set testing and early Excluded Set validation as the baseline. Resist open-ended "substantial completeness" language that could require redundant sampling after production is complete.
Validation isn't about perfection. It's about demonstrating that your false-negative rate is acceptable and that your production meets Rule 26(g) certification standards. The In re Class Action decision gives you a framework to push back on validation requests that cross the line from quality assurance into cost-shifting.
