The Art of Data Science

Description

Data analysis is a difficult process largely because few people can describe exactly how to do it. It's not that there aren't any people doing data analysis on a regular basis. It's that the process by which we state a question, explore data, conduct formal modeling, interpret results, and communicate findings, is a difficult process to generalize and abstract. Fundamentally, data analysis is an art. It is not yet something that we can easily automate. Data analysts have many tools at their disposal, from linear regression to classification trees to random forests, and these tools have all been carefully implemented on computers. But ultimately, it takes a data analyst—a person—to find a way to assemble all of the tools and apply them to data to answer a question of interest to people.  This book writes down the process of data analysis with a minimum of technical detail. What we describe is not a specific "formula" for data analysis, but rather is a general process that can be applied in a variety of situations. Through our extensive experience both managing data analysts and conducting our own data analyses, we have carefully observed what produces coherent results and what fails to produce useful insights into data. This book is a distillation of our experience in a format that is applicable to both practitioners and managers in data science.  Table of Contents Data Analysis as Art Epicycles of Analysis Setting the Scene Epicycle of Analysis Setting Expectations Collecting Information Comparing Expectations to Data Applying the Epicycle of Analysis Process Stating and Refining the Question Types of Questions Applying the Epicycle to Stating and Refining Your Question Characteristics of a Good Question Translating a Question into a Data Problem Case Study Concluding Thoughts Exploratory Data Analysis Exploratory Data Analysis Checklist: A Case Study Formulate your question Read in your data Check the Packaging Look at the Top and the Bottom of your Data ABC: Always be Checking Your ``n''s Validate With at Least One External Data Source Make a Plot Try the Easy Solution First Follow-up Questions Using Models to Explore Your Data Models as Expectations Comparing Model Expectations to Reality Reacting to Data: Refining Our Expectations Examining Linear Relationships When Do We Stop? Summary Inference: A Primer Identify the population Describe the sampling process Describe a model for the population A Quick Example Factors Affecting the Quality of Inference Example: Apple Music Usage Populations Come in Many Forms Formal Modeling What Are the Goals of Formal Modeling? General Framework Associational Analyses Prediction Analyses Summary Inference vs. Prediction: Implications for Modeling Strategy Air Pollution and Mortality in New York City Inferring an Association Predicting the Outcome Summary Interpreting Your Results Principles of Interpretation Case Study: Non-diet Soda Consumption and Body Mass Index Communication Routine communication The Audience Content Style Attitude Concluding Thoughts About the Authors