||Factor Interactions and Aquatic Toxicity Testing.
Voyer, R. A. ;
Heltshe, J. F. ;
||Environmental Research Lab., Narragansett, RI.
Aquatic animals ;
Water pollution ;
Mathematical models ;
Experimental design ;
||Most EPA libraries have a fiche copy filed under the call number shown. Check with individual libraries about paper copy.
Hypothesizing that experimental variables constituting an exposure situation act independently when in combination, the authors have reviewed two published data sets dealing with effects of metal mixtures on aquatic animals in order to assess the potential practical significance of factor interactions and their implication to the design of aquatic toxicity tests. Both data sets were re-analyzed using three predictive models. The relative effectiveness of each model in predicting joint effects of independent test variables was evaluated in terms of calculated mean-square error and goodness-of-fit values, as well as by how well predicted treatment effects compared with responses observed by original investigators. A sequential testing protocol is presented which would permit an evaluation of the existence of factor interactions.