The Importance of Identifying and Accommodating E-Resource Usage Data for the Presence of Outliers. The Negative Impacts of Inaccurate E-Journal Usage Data.

Abstract

This article presents the results of a quantitative analysis examining the effects of abnormal and extreme values on e-journal usage statistics. Detailed are the step-by-step procedures designed specifically to identify and remove these values, termed outliers. By greatly deviating from other values in a sample, outliers distort and contaminate that data. Between 2010 and 2011, e-journal usage at the J.N. Desmarais Library spiked as a result of illegal downloading. The identification and removal of outliers had a noticeable effect on e-journal usage levels. They represented over 100,000 erroneous articles downloaded in 2010 and nearly 200,000 erroneous downloading in 2011.