Management Science: The Art of Modeling with Spreadsheets, 4th Edition

5 Data Exploration and Preparation

5.1 INTRODUCTION

Well-prepared business analysts need to know how to use data to derive business insights and to improve business decisions in a variety of contexts. In one context, the analyst is called upon to describe the current situation. For example, the task might be to discover which products contributed most to profit over the past year. In another context, the analyst may be asked to predict the future situation. In this case, it may be useful to project sales by product line over the next year. In yet another context, the analyst might be called upon to prescribe specific actions the organization should take to achieve its goals. In this case the task might be to recommend a pricing strategy for each product line in order to increase profits. The first of these tasks, describing the past, is simply a matter of extracting and summarizing the appropriate data; it involves no modeling at all. At the other extreme, the pricing task may involve sophisticated management science modeling supported, of course, by relevant data. Whether the task at hand is primarily descriptive, predictive, or prescriptive, an analyst needs to command a range of data-related skills.

In this chapter, we focus on the basic skills needed to understand a data set, to explore individual variables and groups of variables for insights, and to prepare data for more complex analysis. Most of these skills relate to the descriptive tasks mentioned above. However, these ...