"Full Cycle" Data Research
The non-spatial sibling of library/investigations/_protocol/full-cycle-map-research.md. Same five steps, aimed at a chart, a table, or a dataset instead of a map. Reusable against anything with numbers in it.
It is deliberately the same shape as the map protocol. A student who has run one should hear the echo when they run the other, and should be told out loud that it is the same cycle. The method is the constant; only the subject changes.
The three standing questions
Before the cycle starts, and against every chart for the rest of the four years:
1. What is being counted? 2. Compared to what? 3. Who made this, and what did they want me to think?
A student who cannot answer 1 has nothing to analyse. A student who cannot answer 2 is about to mistake a big number for a large one. Question 3 is not cynicism — see the bucket rule in 01-read-the-chart.md.
The cycle
1) Name the source and the claim. The dataset or chart, its publisher, and — in one sentence — what it appears to be saying. Not what it is. What it appears to say to someone glancing at it.
2) What are the outliers?
- a value unlike the other values
- a category with far more or far less than the rest
- a value that is missing, or a category that is not on the chart at all
- a number that is suspiciously round, or repeated, or exactly at the edge of the scale
3) Try to identify possible patterns, which may be like:
- a) I think this value is related to when it was measured
- b) I think this value is related to who was asked
- c) I think this pattern is related to another dataset I have seen
- d) I think this pattern is related to something I already know about this subject
4) What questions do these make you think of?
- Why is the data shaped like this?
- What is the reason I see this here, but not there?
- What is it about this group that is different from the others?
- What would this chart look like if it counted something slightly different?
5) RESEARCH your question and come up with ANSWERS, or EVEN MORE QUESTIONS as a result of your RESEARCH.
Per-item worksheet: Source · Apparent claim · Outliers · Patterns · Questions · Research findings.
The two carried rules
Absence is data too. Item 2's third bullet is not optional. The default student behaviour is to chase the tallest bar; the skill is asking why a category is short, or why it is not on the chart. This is the same rule the New York investigation states and it originates in the map protocol.
Ending with more questions is a success. Item 5 says so on purpose. An investigation that opens new questions has worked. Say it out loud; almost no other school assignment does.