ABSTRACT
Analysis is at the core of our use of data. There are innumerable texts on the performance of data analysis. Predictive analysis uses many different techniques including methods from descriptive statistics, regression, and large-scale data mining and machine learning. Inferential analysis uses quantitative data and is usually focused on hypothesis testing. The most insightful and potentially productive analytic result is essentially meaningless unless it can be communicated successfully, that is, in a way that it can understand and acted upon. Communication of strategic results must be sponsored and, in the most successful cases, carried out by senior executives, including the C-staff. There are two types of assumptions relevant to the planning and execution of analysis. The first type covers those assumptions inherent in the mathematics of the analytic tests themselves. The second set of assumptions regarding the planning and execution of analysis is discussed much less regularly, although violation of these assumptions also will invalidate the analysis and its interpretation.
