In many of the situations where human health risk is a consideration, different types of data, from different sources and various methods of calculations must be organized and analyzed. Statistics is the major component in understanding the significance of these wide ranging data.

Human and ecological risk assessment, such as the risk of cancer or nuclear contamination, is a function of toxicity and other variables that exist over space and time. The concentration of heavy metals in groundwater is one example of a space-time variable. Another is the speed at which groundwater moves. Both apply in the situation where risk is associated with a nuclear-waste repository. The experts at The Sapphire Group can model space-time variables, and conduct simulations so that risk can be evaluated. This leads to a statistical distribution that quantifies a range of likely levels for risk, a significant aid to decision-making.

The Sapphire Group applies the value of this area of expertise in conducting both exposure and dose assessments. We also assist our clients, through the use of space-time modeling, in generating strategies for management of natural resources, such as fish populations, forests, and air and water quality prior to risk of ecological injury.

• Statistics represents an invaluable array of tools for:
• Quantifying exposure and concomitant risk at hazardous waste sites
• Analyzing censored data, e.g., non-detected chemical concentrations
• Predicting toxicity from spatially distributed data
• Estimating abundance and survival from mark-recapture data
• Predicting ozone from atmospheric conditions
• Accounting for spatial and temporal variability
• Predicting suitability or abundance from measurements of physical habitat
• Combining data from epidemiological studies using meta analysis
• Using Global 86 to analyze data from animal studies
• Exposure to environmental tobacco smoke
• As a complex mixture in air, both particulate-phase and vapor phase components of the mixture

Our professionals have experience in analyzing al types of complex numerical data including:

• Descriptive statistics and graphics Linear and nonlinear statistical models
• Nonparametric and semiparametric statistics
• Modeling exposure and dose from complex mixtures in air and in groundwater.

Our approaches include:

• Study design Design of experiments and surveys
• Sampling, including sequential sampling
• Sample size and power calculations
• Linear and nonlinear regression
• Time series analysis
• Analysis of variance and analysis of covariance
• Monte Carlo (probabilistic) and bootstrap analyses
• Sensitivity and uncertainty analyses
• Bayesian simulation
• Maximum likelihood estimation
• Decision theory
• Stochastic processes
• Data validation

RETURN TO TOP OF PAGE