about communicating effectively within a team and that effective collaboration is enabled by standardization and simplification of data tasks. This book provides a narrative outline of the data workflow at each stage of an empirical research project, from design to publication, as visualized in figure I.1. Chapters 1 and 2 contextualize the workflow and set the stage for the hands-on data tasks described in detail in chapters 3 to 7.
Do the data already exist?
Yes
Conduct sampling and randomization
Prepare data license agreement
Transfer data securely
Prepare data collection instruments and protocols
Collect and transfer data securely
Processing
No
Create a data map
Remove direct identifiers
Tidy data
Monitor data quality
Clean data
Analysis
Acquisition
Identify data needs
Construct indicators
Conduct exploratory analysis
Polish results
Export outputs
Publication
Reproducibility, transparency, and credibility
Design
FIGURE I.1 Overview of the tasks involved in development research data work
Prepare final research output
Assess risk of disclosure
Publish data
Release reproducibility package
Source: DIME (Development Impact Evaluation), World Bank.
Chapter 1 outlines the principles and practices that help consumers of research to be confident in the conclusions reached and describes the three pillars of a high-quality empirical research project: credibility, transparency, and reproducibility. It presents three popular methods 2
DEVELOPMENT RESEARCH IN PRACTICE: THE DIME ANALYTICS DATA HANDBOOK