Module 06 Activity
Scenario
Your charts assert differences between groups. This activity checks whether the data supports them.
What you build
A distribution set that shows spread rather than asserting it, with sample sizes visible.
Steps
- For every mean you chart, also compute the median, the range and the spread. Flag any group where the mean describes nobody.
- Rebuild one histogram at four bin widths and say whether the conclusion is stable across them.
- Take one box plot and overlay the points. Report anything the box hid.
- Replace one two-bar comparison with a plot of the points, and report the overlap between the groups.
- Compute the confidence interval for every category and add it, or add the sample size where you cannot.
- Identify the category with the fewest observations and say whether any decision rests on it.
Evidence to hand in
- Mean, median, range and spread per group, with any flagged.
- The four bin widths and your stability judgement.
- The box plot with points overlaid, and what was hidden.
- The points plot with the overlap reported.
- Intervals or sample sizes on every category.
- The smallest category and what rests on it.
Review checklist
- Every charted mean is accompanied by a spread measure.
- The bin-width conclusion was checked across several widths.
- At least one box plot has its points overlaid.
- Every category carries a sample size or an interval.
- The smallest category is identified and what rests on it is stated.
