Statistical Evaluation of Arsenic Adsorption Data Using Linear‐Plateau Regression Analysis

Abstract

Abstract Few studies have utilized a statistical approach for comparisons of adsorption data. Our objective was to evaluate the suitability of linear‐plateau regression analysis to identify the As adsorption plateau and facilitate statistical comparison of As‐adsorption rates and maxima. Three northern Idaho soils exhibiting a range in As‐adsorption capacity were tested. Batch techniques with ion chromatographic determination of arsenite and arsenate were used. The linear‐plateau regression model provided a good fit for the adsorption‐rate data. The time required to attain an adsorption plateau ranged from 2 to 8 h, depending on soil characteristics and As species. The model allowed statistical identification of the As‐adsorption plateau and subsequent comparison of As‐adsorption rates and maxima through use of analysis of covariance and least significant difference procedures.