Start with a question
Analysis is more useful when it answers a decision question. Examples include which service step takes longest or whether a target was reached. Write who will use the answer and what decision they may make with it.
Check what the data represents
Find out when the data was collected, what one row represents, how fields were defined, and whether any records are missing. A spreadsheet can contain precise-looking numbers that measure the wrong thing or cover only part of the intended group.
Try it
Choose a small table you are allowed to use. Write its source, date range, row meaning, and one question it can reasonably answer. Note one question it cannot answer.
Key takeaways
- Begin with a decision question.
- Understand row meaning, dates, and definitions.
- State what the dataset cannot answer.