Is our data credible?

Date of publication: June 15, 1994


Many organisations have difficulty in making effective use of market research information, particularly in business to business markets. There are various reasons for this of which perhaps the two most important are: a) The calendar time taken from fieldwork to reporting. b) The variability of the data, at least at a sample size consistent with client budgets. This paper firstly reviews some of the problems associated with the business to business research and then, through the medium of a (partially artificial) case study describes one system for automatically improving the value of research data within the context of a continuous monitor. The system uses the Microsoft Excel spreadsheet package to take raw data through to summary charting and reporting within a four-hour time frame. The process is based upon an expert system macro including two core elements which we call rationalization and remobilization, each of which is described in the paper. We conclude that since the system makes adjustments as the expert would and since computers are more rigorous than human beings, the result is less arbitrary than the alternatives such as adjusting for outliers by hand.

  • PDF
  • This could also be of interest