Assignment for ‘Sampling, DOE and Measuring’

 

  1. Present examples for each of the six types of errors which threaten the representativity of a sample. Preferably ‘real’, e.g. from a newspaper (cf. ‘Verknipte Statistiek’).

 

  1. Describe a situation where it would be a good idea to use

 

·        stratified sampling

·        cluster sampling

 

  1. Present three examples of experiments that suffer from confounding.

 

  1. Give a (rather precise) description of setting up an experiment to check the claim that praying increases the harvest. (Example derived from an experiment with soy beans in Jamestown, Ohio.)

 

  1. Present an example where
    • the realistic content is questionable
    • a CRD is not advisable and a RCBD is to be preferred

 

  1. Give an example of a measure which is

·        valid but unreliable

·        reliable but invalid

 

  1. Give an example of an experiment where sample mean and variance are not very sensible choices.