Many experimental observations of electrochemical nature suggest a sigmoid shape association existing between independent (explanatory) and dependent variables. A sigmoidal relationship can be advantageously analyzed in terms of probit (probability unit) and logit (logistic unit) functions, serving as mathematical models for fitting binary and binomial data. Well known in biological, toxicological and medical statistics, this approach has not hitherto been exploited in electrochemical science and engineering. This paper intends to illustrate quantitatively the potential usefulness of probit- and logit-based analysis to these disciplines by assessing current efficiency measurements in a hypothetical catalyst-assisted electrolytic copper plating process. Only a basic knowledge of probability theory and statistics is required.
Buy this Article