Accurate analysis sometimes depends on obtaining additional information. Care and skill is needed to obtain this additional data cost-effectively, while at the same time not biasing the results. Sampling and surveys are tools used to dramatically reduce required work while still obtaining useful data.
Often, data/observations that are not part of contemporaneous business records must be gathered. We can perform the data collection ourselves and/or we can perform the statistical analyses on surveys and data/observations conducted/collected by others.
A recurring need for this arises in class action disputes in which the Court needs to evaluate the existence and the similarity of alleged conduct(s) among class members. Statistics are an accepted way of providing “representative evidence” in this situation.
On Numerous Occasions:
- Collected and consolidated transactional databases. Conducted statistical analyses of the underlying data.
- Developed surveys measuring buyer motivations and values. The work included consideration of bias, sample selection, statistical methodology, and interpretation of results.
- Estimated overcharges in a population through statistical sampling and auditing of such samples.
Examples of Past Projects Include:
- In an audit of federal wildfire suppression costs, designed and implemented a statistically valid sample of the fire suppression costs. Estimated the total amount of billing overcharges based on proper statistical sampling methods.
- Designed and implemented a survey instrument to determine what effect(s) of one of the world’s largest booksellers opening a retail store has on nearby condominium residents. Based on the survey results, assessed the likelihood that the store’s presence had influence on the condominium owner’s decision to purchase the condominium.
- Designed a random sample for a project analyzing data of alleged billing overcharges at an insurance company. Estimated the total amount of billing overcharges based on proper statistical sampling methods.
- Designed a random sample and estimated the alleged average error rate (i.e., rent overcharge) in a dispute over rent overcharges. Extrapolated sample results to the entire population using proper statistical sampling methods.