Pre-testing erroneous text-based documents: Logging end-user activities

Abstract

An extremely high number of erroneous text-based documents are in circulation, being multiplied, serving as samples both in the office world and in education. Creating, editing, and modifying these documents generates a serious financial loss when both human and machine resources are considered. Our research team developed an application and a testing method for building up an objective measurement system, in order to see the rate of loss in the handling of erroneous documents. In the pre-testing period, the error factor of a sample text was set up based on the logged activities of an expert researcher. It was found that first level modifications require about five times more human and machine resources in a short, one-paragraph text burdened with layout errors than in its properly formatted version. Further testing is required to find out how demanding longer texts and more serious modifications are, but it is already obvious that erroneous text-editing, and the lack of fundamental computational thinking skills involve unnecessary use of our resources.