2.12 Optional: Designing and planning studies

This question is optional; e.g., if you need more practice, or you are studying for the exam.

(Answers are available in Sect. A.2)

To study the progress of an imported species of beetle, a research group is studying the research question:

Is the mean number of species of beetle found on trees different between local coniferous and broad leaf forests?

Ten possible forests of each type have been found, and the research budget allows for \(100\) trees to be sampled. Three designs are being considered:

  • A. Randomly sample \(50\) trees from one randomly selected coniferous forest, and \(50\) trees from one randomly selected broad leaf forest.
  • B. Randomly sample \(5\) trees from ten randomly sampled coniferous forests, and \(5\) trees from ten randomly selected broad leaf forests.
  • C. Randomly sample \(10\) trees from five randomly sampled coniferous forests, and \(10\) trees from five randomly selected broad leaf

The researchers wish to know which design they should use.

  1. Identify (with justification) which of these designs will be best for estimating the variation within a forest.
  2. Identify (with justification) which of these designs will be best for estimating the variation that exists between the forest (that is, from one forest to the next).
  3. Which design do you think would be best for answering the RQ? Explain your answer.
  4. Which design do you think would be the easiest design for data collection?