This research addresses the need for models that guide data quality design and resource allocation decisions. Broadly, our research problem is: Given an information system that utilizes a set of data sources for producing required information, how can we determine the gain in information accuracy and, subsequently, the economic return if the accuracy of a chosen data source is improved? An earlier paper by the author approaches this problem through a construct and a model. The construct, named damage, is defined as the change in information accuracy that results from improving the accuracy of a chosen data source. The model that is provided together with this construct enables its quantification as well as a simple ranking of inputs according to the damage that errors in each inflict. The model suits environments in which the data are applied mostly by satisficing, multi-criteria decisions, such as databases. This paper reports on a series of Monte Carlo Simulations that validate the ranking component of the model under conjunctive decisions, and, in addition, explore and characterize special conditions in which a predicted ranking is not assured to be correct.
Askira Gelman, Irit, "Simulations of Error Propagation for Prioritizing Data Accuracy Improvements in Multi-Criteria Satisficing Decision Making Scenarios" (2009). ICIS 2009 Proceedings. 185.