How many times have you opened a dataset someone else has worked on and wondered “what happened here?!” For that matter, how many times have you opened your own dataset months after last working on it and forgotten how you cleaned your data? As you clean errors of transposition, copying, coding, routing, consistency, range, etc., it is vital you systematically document your progress.
Documenting data preparation may not be the most fun aspect of analysis, but it’s foundational. Skipping over documentation is like shoving your mess into your closet and under your bed – your mom will find it!
HELP! I don’t how these responses got recoded! What’s 999!?!?
We implement a systematic process of documentation in each project to ensure data quality. Not only does documentation support accountability, it also supports replicability. Our documentation tools include:
- Codebook – aligns your questions, response options, variable names, and values
- Syntax – outlines how data was processed
- Data Tracker – lists known issues from data collection