Estimating efforts required upgrading technical infrastructure of a global project using randomized response techniques

Ghosh. Ghosh (2003) presented an approach to estimate efforts required to upgrade technical infrastructure while implementing a multinational complex software solution. Practical experiences suggest that often individuals cannot express themselves properly being afraid of going against the sentiment stream of the group or corporate directions. When it happens, decisions made are biased and inaccurate. A mathematical approach is presented to measure efforts. Randomized response (RR) technique was introduced by Warner (1965), are used extensively by the statistical community [A., Chaudhuri et al., 1987] to collect sensitive data but preserving identity of the individual. In this paper, we propose use of RR to gather sensitive information so that identity of the individual is not scarified. Such a method adds certain degree of randomness to the answers to prevent the data collector from learning the true information. Measurement of complexity is subjective and may vary depending on number of known and unknown factors. Also complexity and efforts required to achieve the upgrade are not linear either. A questionnaire approach is proposed to gather data and then classified into a multivariate decision process for analysis and reporting. We also propose a way to build a multi-variate decision tree to classify collected data. Using this systematic approach facilitates the decision process based on a data set with a quantified degree of uncertainty. This also helps identifying the critical success factors and where executive sponsorship is most required.