The Critical Role of Sampling in Discovery

  • Published on Jul 27, 2017
Written by: Philip Favro

Philip Favro is a leading expert on issues relating to electronically stored information. Phil serves as a court-appointed special master, expert witness, and trusted advisor to law firms and organizations on matters involving ESI and electronic discovery. He is a nationally recognized scholar on electronic discovery, with courts and academic journals citing his articles. Phil also regularly provides training to judges on electronic discovery and ESI. He is a licensed attorney who in private practice represented organizations and individuals in litigation across the spectrum of business disputes. In addition to handling a range of complex and other discovery issues, Phil has extensive experience in the courtroom including summary judgment, preliminary injunction, and discovery motion practice, together with trial and arbitration experience.

One of the most important aims of the 2015 amendments to the Federal Rules of Civil Procedure was the elimination of “unnecessary or wasteful discovery.” While a worthy objective, this goal is often elusive. Counsel, clients, and even the courts often get caught up in ensuring that every responsive document has been produced or otherwise accounted for in discovery. Though understandable, this approach is an unproductive and exhausting discovery trap. It squanders resources while failing to accomplish the very purposes for which discovery has been undertaken.

Sampling as a Solution

As discussed in an article recently published by Legaltech News, there are other methods and tools that can simplify the discovery process while reducing expenses and wasted opportunity costs. One of the best ways available for doing so is sampling.

Sampling encapsulates a very simple concept: that “a few” can adequately represent “the many.” Sampling has been deemed a sufficient measure in a variety of different circumstances. For example, sampling is used to measure patient health, to determine how voters are leaning on a particular candidate or ballot measure, and even to gauge the coordination of air traffic.

In like manner, a properly designed sample of electronically stored information has been considered sufficient for discovery purposes. In particular, sampling can be used to depict information from a broader set of data among dozens or hundreds of custodians spanning a range of years. This is confirmed by the Manual for Complex Litigation. Published by the Federal Judicial Center “as a resource for managing complex cases,” the Manual provides that sampling is an “acceptable” method for responding to discovery:

Acceptable sampling techniques, in lieu of discovery and presentation of voluminous data from the entire population, can save substantial time and expense, and in some cases provide the only practicable means to collect and present relevant data.[1]

Whether the methods chosen for creating a sample are acceptable depends on a variety of factors. According to the Manual, those factors include whether:

  • the population was properly chosen and defined;
  • the sample chosen was representative of that population;
  • the data gathered were accurately reported; and
  • the data were analyzed in accordance with accepted statistical principles.”[2]

Case Law and Sampling

Courts have followed the Manual’s guidance and used sampling to proportionally balance the burdens of discovery against its benefits. For example, the court in Duffy v. Lawrence Memorial Hospital recently ordered that sampling be used both to alleviate the defendant’s production burdens and to ensure that responsive information was actually produced in discovery.[3]

The Duffy court had initially ordered the defendant to produce certain patient records relating to the plaintiff’s claims. However, it reversed course once the burdens associated with the order became apparent. Those burdens included the deployment of hospital staff to individually review 15,574 electronic patient files so as to identify particular patient visit information. Not only would such a method be costly (nearly $200,000), it “would cause significant strain on Defendant’s access to and use of its electronic production database, slowing the system and potentially impacting patient care.”

Given these burdens, the court agreed to explore the defendant’s proposal that it instead produce a random sample of 257 patient records. It ultimately the adopted defendant’s sampling approach for three reasons. First, the sampling methodology was deemed acceptable since it complied with the criteria from the Manual on Complex Litigation. Second, sampling enhanced the accuracy of the information produced in discovery by reducing “the element of human error,” which generally occurs when large volumes of data are manually reviewed in a short period of time. Finally, sampling satisfied the tripartite mandate from Federal Rule of Civil Procedure 1 by providing the parties with a substantively better, more efficient, and less expensive method for producing relevant information.

Duffy is a quintessential example of the value that sampling yields in the discovery process. And Duffy is not an isolated example; several other decisions spotlight the benefits of sampling in achieving proportional results in discovery. Many of those decisions, along with some practical considerations and specific use cases regarding sampling, are discussed in the recently published Sedona Conference Commentary on Proportionality in Electronic Discovery. All of which demonstrates how sampling can provide lawyers and litigants with a more efficient and cost-effective way for obtaining crucial evidence in litigation.

[1] Manual for Complex Litigation, Fourth, § 11.493.

[2] Id.

[3] Duffy v. Lawrence Memorial Hospital, No. 2:14-cv-2256-SAC-TJJ, 2017 WL 1277808 (D. Kan. Mar. 31, 2017).