First, consideration of the statistical method of data analysis is important at the beginning of a project, not when the risk assessment starts. Utilizing the 95% Upper Confidence Limit of the arithmetic mean (95UCL) as a conservative concentration includes many site-specific considerations that should be discussed prior to collection of data, including:
- locations of samples (spatial variability),
- depths of samples (vertical variability),
- when should samples be collected (temporal variability),
- type of sampling done (i.e.,composite, discrete or incremental), and
- number of samples collected (data adequacy).
Understanding the site data concentrations and data trends are critical for a risk assessment to be an effective tool in managing sites. It can also identify metals or other analytes where an understanding of background conditions will require a targeted data collection approach. Additionally, as data is collected within an investigation, risk assessment can be utilized to statistically evaluate the site data in real time, indicating where more data may be needed for example to understand migration or where specific hot spots might exist that will require more investigation.