Statistical Analyses
Statistical services include:
The design or review of field survey designs, sampling plans and Quality Assurance Project Plans (QAPPs) to ensure proposed activities meet project objectives. If your project involves field sampling, the first critical step before developing the sampling plan is to clearly define the project objectives in a way that allows the specific data to be collected that will answer your questions. Using EPA’s Data Quality Objectives (DQO) process as a guide, the first step is to make sure the right questions are asked, and then work backwards to articulate how decisions will be made, what data are needed to make those decisions, and how best to collect that data. Our statistical expertise can guide you through this process from sampling plan development through the decision-making process. We can identify the most appropriate statistical tools to use to answer your project questions, and evaluate study designs using statistical power analyses to identify the most efficient design for collecting the data necessary to meet project goals.
The tools to batch process and evaluate your environmental data. Exa specializes in compiling datasets from multiple sources, which often requires re-calculating interpretative summary statistics to provide consistency for further analysis. Exa can integrate scripting in R (a statistical programming language and environment) with your database to batch calculate results. Results from multiple data types may be combined and analyzed within Access queries or pulled directly from Access into R. Exa has generated scripts to standardize toxicity significance within compiled datasets, investigated relationships between sediment chemistry and toxicity, and evaluated the predictive performance of sediment quality guidelines on independent datasets. Diagnostic and evaluation tools can be customized for your application.
The analysis of environmental data using stochastic modeling, and descriptive and inferential statistics to evaluate spatial and/or temporal patterns in your data or construct confidence or Bayesian credible intervals for observed changes. Exa can help you evaluate the significance of observed changes both statistically and in the context of what constitutes an ecologically meaningful difference. A combination of our expertise with environmental statistical models and your intimate knowledge of your subject will produce the most meaningful results possible from your data.