Statistical Sampling

Properly designed and implemented, statistical sampling is a valuable litigation tool. We are often involved in the design and implementation of sampling strategies, including the critical evaluation of opposing parties’ samples. When it is impractical or expensive to evaluate every observation in a population, a properly selected sample can yield robust inferences about the characteristics of the underlying population. 

We have applied sampling methods in cases involving data spoliation, fraud, product defects, financial instruments, medical claims, intellectual property, consumer products, real estate, and employment discrimination. For example, clients have requested statistical sampling projects to evaluate the quality and reliability of electronic data compiled from source documents, select records for more detailed investigation, and extrapolate results from records with adequate information to those with missing data.

This has included applying statistical analysis to questions related to statistical power and determination of sample sizes. Appropriate sample sizes are contingent, for example, on the variable of interest. Samples that are appropriately powered for one variable may not contain enough observations to draw reliable inferences on other variables of interest.

Key Questions in Sampling

Many of the challenges in statistical analysis are associated with selecting a representative sample and using that sample to estimate characteristics of the population from which the sample was drawn. To properly select a sample or evaluate the quality of an existing sample, our analyses often focus on key questions about sample design and implementation, including:

  • What question needs to be answered?
  • What is the population of interest?
  • What level of accuracy is required, and what size sample is required to achieve that desired accuracy level?
  • Is there a benefit to selecting a stratified sample?
  • Is the sample representative? Is it biased?
  • Were the sampling plan and data collection properly implemented?
  • Is the selected sample sufficient to answer the question of interest?
  • Which method should be used to extrapolate from the sample to the population of interest?
Read more