Recovering incomplete data using Statistical Multiple Imputations (SMI): a case study in environmental chemistry

Area of Science:

• Environmental Chemistry
• Statistical Analysis
• Environmental Science

Background:

• Environmental chemistry data often contains missing values and values below the limit of detection.
• These data limitations hinder traditional statistical analysis, particularly in environmental leaching studies.
• Analyzing leached contaminants like arsenic, chromium, and copper requires robust statistical methods.

Purpose of the Study:

• To present a statistical technique for analyzing environmental chemistry data with missing values and detection limit issues.
• To demonstrate the application of Statistical Multiple Imputation (SMI) on environmental leaching data.
• To assess the impact of SMI on data variance and analytical feasibility.

Main Methods:

• Application of Statistical Multiple Imputation (SMI) to address missing data points.
• Analysis of leachate from lysimeters containing treated and untreated wood waste.
• Quantification of arsenic (As), chromium (Cr), and copper (Cu) concentrations using ICP-OES.

Main Results:

• SMI successfully recovered an incomplete dataset from an environmental leaching study.
• The re-analyzed complete dataset allowed for successful statistical assessment.
• It was demonstrated that SMI did not significantly affect the data's variance.

Conclusions:

• Statistical Multiple Imputation (SMI) is an effective method for handling missing data in environmental chemistry.
• SMI facilitates the analysis of complex environmental datasets previously hindered by data gaps.
• This technique enhances the reliability and scope of environmental contaminant studies.