Case
Abstract
Conducting primary research using financial crimes’ databases can be difficult, particularly in areas where documentation does not only involve instances of financial misconduct but also big data and cases of multiple wrongdoings. The present case study will give readers insights into some of the problems that researchers’ experienced when coding and calibrating big data from financial crime cases for further empirical analysis. In this case study, the coding procedures used to collate and arrange the data for further analysis are described as well as the modifications that were made so that the data could be analyzed and inferences drawn from a multiple regression model. A common feature with financial crimes data is that predictor variables are typically serially correlated and can lead researchers to wander beyond the data to come up with robust statistical results. This problem is addressed in this case study with reference to the assumptions and conditions of a multiple regression model. Areas of future research are also highlighted.
