Connect the ideas
Data may name a group or count objects.
Record every response once and check the total.
Label displays and use a clear scale or key.
Support comparisons with actual counts.
Conclusions describe the surveyed group, not automatically everyone.
