Abstract:
Analysis and discussion are presented for the various schemes that may be used in evaluating results such as the current efficiency (CE) obtained from aluminum smelting cell performance trials. A conventional parametric (requiring normal distribution of data) and a nonparametric (not requiring normal distribution) statistical method are given. The analysis allows the results obtained to be expressed in statistical terms with a known level of significance, thus providing a more scientific basis for the interpretation of the results. A scheme for experimental design that takes into account inherent variations and the use of a nonparametric statistical test which does not stipulate normal distribution of data are both suggested.
