← Use statistical methods to make a formal inference

5 Getting to Excellence

What lifts this report

  • Describe the samples properly first, shape, centre, spread, unusual features, and always in context with units.
  • Compare the two distributions, not just the medians: overlap, spread, skew, outliers. A big overlap is worth commenting on.
  • Explain the bootstrap in your own words rather than describing button clicks.
  • Justify the call explicitly from the interval’s position relative to zero.
  • Discuss sampling limitations: how the sample was taken, whether it is representative, and what that means for how far you can generalise.
  • Comment on practical significance: a statistically detectable difference can still be too small to matter in the real world. Saying so is a strong Excellence move.
✓

Sample size and interval width

Larger samples give narrower intervals (less sampling variation), which makes a call more likely. If your interval is very wide and straddles zero, note that a larger sample might have resolved it. That is good evaluative writing.

Practise just this