You are here:
CAUSAL ANALYSIS AND PROBABILITY DATA: EXAMPLES FOR IMPAIRED AQUATIC CONDITION
Citation:
PAUL, J. F., H. A. WALKER, AND M. E. MCDONALD. CAUSAL ANALYSIS AND PROBABILITY DATA: EXAMPLES FOR IMPAIRED AQUATIC CONDITION. Presented at EMAP 2007 Symposium, Washington, DC, April 10 - 11, 2007.
Description:
Causal analysis is plausible reasoning applied to diagnosing observed effect(s), for example, diagnosing
cause of biological impairment in a stream. Sir Bradford Hill basically defined the application of causal
analysis when he enumerated the elements of causality for associating cigarette smoking with lung
cancer. The identification of the true cause for an effect can be extremely difficult, if not impossible, to
prove. However, by putting together a series of elements that must be satisfied, a plausible case can be
made on the possible cause of the observed effect(s). The primary objective of EMAP has been to focus
on condition of the aquatic resource, then to acquire a suite of exposure and habitat indicators, which can
be used for diagnostics. The site selection for EMAP sampling is probability based, ensuring unbiased
estimation of condition and of the stressors associated with condition. EMAP data are thus well suited for
use in plausible reasoning for identifying potential cause(s) of impaired aquatic resources. Possible
criteria for the EMAP data to be appropriate for a specific causal analysis are (1) a response threshold
must exist which identifies impairment of the aquatic resource, (2) the indicator of condition must be
responsive over the range of the potential indicators of cause, and (3) the spatial density of the data must
be adequate for the specific problem. Analysis tools used in the diagnostics of causal analysis include
scatter plots, cumulative distribution functions, least-squares regressions, quantile regressions, and
conditional probability analysis. Examples of applying theses tools with data for streams and estuaries are
used to illustrate how probability surveys can contribute to diagnostics.