Quality Control Fundamentals, Review Fundamentals Series Part 6

6 / 6

A multi-part series on the fundamentals eDiscovery practitioners need to know about document review planning and execution

In “The Main Event,” we discussed the costs and significance of review, as well as the question of what gets reviewed.  In “For What It Gets Reviewed,” we discussed the range of determinations that you might want reviewers to be making.  In “Who Does the Reviewing,” we discussed your options for review staffing.  In “Workflow Design Considerations,” we discussed document flow and tagging palette considerations.  In “More Workflow Design Considerations,” we discussed batch creation and tracking, reporting, and documentation considerations.  In this final Part, we review fundamentals of quality control.

The final and most important fundamental of review to understand is quality control.  No matter what you’re reviewing, what you’re reviewing it for, who’s reviewing it, or how you’re reviewing it, you will need to take proactive steps to ensure the overall quality and consistency of that work.  Perfection isn’t possible (and isn’t required), but reasonable efforts to meet your obligations of completeness, accuracy, and privilege protection are both.

The Myth of the Gold Standard

The Sedona Conference’s Best Practices Commentary on the Use of Search and Information Retrieval Methods in E-Discovery describes a persistent myth in eDiscovery:

It is not possible to discuss this issue without noting that there appears to be a myth that manual review by humans of large amounts of information is as accurate and complete as possible – perhaps even perfect – and constitutes the gold standard by which all searches
should be measured

The reality is quite different from this myth.  In reality, even the best reviewers make numerous mistakes due to simple human fallibility, and reviewers frequently come to different conclusions regarding questions of relevance, privilege, and more.  Studies have shown surprisingly low consistency between the independent results of equivalent review teams (“Assessor Overlap”).

Because of this reality, it is critical that your document review project include some steps to ensure an acceptable minimum level of quality, consistency, and completeness.

Traditional Methods

The most traditional method of quality control is second level (or second pass) review.  In this method, some portion of the material reviewed by first level (or first pass) review is re-reviewed by more senior reviewers to check the accuracy and consistency of the work.  The volume re-reviewed and the focus can vary widely depending on the needs of the project:

  • In a smaller project, you might re-review everything deemed relevant and non-privileged to make sure nothing irrelevant or privileged is produced
  • In a larger project, you might re-review a random 10% of the first level review to look for recurring mistakes to address or problem reviewers to retrain
  • In a project utilizing a technology-assisted review workflow, you might focus more of your efforts on evaluating the materials deemed irrelevant to be sure nothing important has been missed

In some projects you may establish more than two levels of review.  For example, you might add a third level in which case team members re-review certain materials prior to production.

The other traditional quality control method is targeted searching.  Targeted searching is the practice of running searches against the reviewed materials for key terms that would likely indicate clear relevance, irrelevance, or privilege and then double-checking that the results are coded correctly.  For example, you might search for key attorneys’ names and email addresses and double-check the privilege tagging applied to the results.


Sampling broadly comes in two categories: judgmental sampling and formal sampling.  Judgmental sampling is the informal process of looking at some randomly selected materials to get an anecdotal sense of what they contain.  The random 10% second-level review and targeted searching described above are examples of judgmental sampling.  The goal of these efforts is to get an impression and make an intuitive assessment rather than to take a specific measurement.

Formal sampling is just the opposite: you are reviewing a specified number of randomly-selected documents with the goal of taking a defined measurement with a particular strength.  Typically that measurement is either being taken to test classifiers or estimate prevalence:

  • Testing classifiers is the process of seeing how effective and efficient a particular classifier actually is, be it a search, a TAR process, or a human reviewer. Using this technique, you can quantify the accuracy and error rate of individual reviewers and teams (or quantify the recall and precision of searches or TAR processes).  These measurements can be used to identify problem reviewers, to measure overall review effectiveness, or to implement lot acceptance sampling.
  • Estimating prevalence is the process of reviewing a simple random sample of a given collection of materials to estimate how much of a given kind of thing is present. In the context of quality control, this is used most often to measure how much relevant material may have been left behind in the pools of material that were not reviewed (or reviewed and deemed irrelevant).  You might also use this method to create a yardstick for your review before you begin by estimating in advance how much relevant and privileged material you expect to find.

For a deeper dive on these two sampling techniques, please see our series, practice guides, or webinar on Estimating Prevalence and Testing Classifiers.

Feedback Loops

Regardless of the specific quality control methods you choose to employ on your project, it is critical that effective feedback loops are established.  In most document review projects, you will be engaged in ongoing quality control throughout first-level review, giving you the opportunity not just to catch and correct errors, but to identify issues and address them with first-level reviewers to improve the rest of their work.  Effective feedback loops make this possible.

The most important feedback loop is the one between the review managers and the reviewers, This is about feeding the insights gleaned from quality control efforts back to the reviewers though additional instruction and clarification.  For larger projects, it is common to have weekly review team meetings to review issues and answer questions.  It is also common to have one-on-one sessions with individual reviewers identified as requiring additional guidance, and it is a good idea to maintain a shared list of reviewer questions and review manager answers for everyone’s reference.  It is also important that there is a feedback loop between the case team and the review managers, so that review managers can request guidance and clarification as needed to share with their team and so that the case team can share any issues they have identified during any quality control review them have performed.

The Importance of Privilege Protection

As we conclude this series, it’s worth emphasizing the particular importance of engaging in quality control for the purpose of preventing the inadvertent disclosure of privileged materials.  As we discussed previously, Federal Rule of Evidence 502(b) establishes that inadvertent disclosures can lead to privilege waiver if reasonable steps to prevent the disclosure weren’t taken.

The Committee’s Explanatory Note on Rule of Evidence 502 makes clear that “reasonable steps” is a case-by-case determination that can depend on factors such as the total number of documents to be reviewed, the time constraints for production, how records were managed, what tools were used, and more.  Consequently, taking steps to ensure the quality of your privilege review approach is at least as important as what approach you take:

The implementation of the methodology selected should be tested for quality assurance; and the party selecting the methodology must be prepared to explain the rationale for the method chosen to the court, demonstrate that it is appropriate for the task, and show that it was properly implemented.  [emphasis added]

Victor Stanley Inc. v. Creative Pipe Inc., 250 F.R.D. 251, 262 (D. Md. 2008).

For Assistance or More Information

Xact Data Discovery (XDD) is a leading international provider of eDiscovery, data management and managed review services for law firms and corporations.  XDD helps clients optimize their eDiscovery matters by orchestrating precision communication between people, processes, technology and data.  XDD services include forensicseDiscovery processingRelativity hosting and managed review.

XDD offers exceptional customer service with a commitment to responsive, transparent and timely communication to ensure clients remain informed throughout the entire discovery life cycle.  At XDD, communication is everything – because you need to know.  Engage with XDD, we’re ready to listen.

About the Author

Matthew Verga

Director of Education

Matthew Verga is an electronic discovery expert proficient at leveraging his legal experience as an attorney, his technical knowledge as a practitioner, and his skills as a communicator to make complex eDiscovery topics accessible to diverse audiences. A fourteen-year industry veteran, Matthew has worked across every phase of the EDRM and at every level from the project trenches to enterprise program design. He leverages this background to produce engaging educational content to empower practitioners at all levels with knowledge they can use to improve their projects, their careers, and their organizations.

Whether you prefer email, text or carrier pigeons, we’re always available.

Discovery starts with listening.

(877) 545-XACT / or / Email Us