Question

1) A dietician divides his patients into males and females and then randomly samples five of each and records their change in weight after a diet. What kind of sampling is being used? How would this type of sampling compare to a simple random sample?

2) An English professor chooses all the term papers from a few randomly selected sections that she teaches and records the number of grammatical errors. Determine which of the four levels of measurement is most appropriate and explain.

Answer 1

1) a) Stratified Sampling is used by the dietician.

b)A simple random sample would not control for differences in sex. The dietician would not first divide his patients into males and females. Rather he would sample 10 patients (randomly) from the whole set without stratifying his data on the basis of sex.

2) The ratio scale is the best level of measurement for this case as :

i) The number of grammatical errors can range from 0 to n ( and the ratio scale is the only scale where absolute zero is meaningful). Further there is also an inherent ordinality to the parameter being measured.

ii) It will allow her to treat and analyse her data in more significant ways than using an ordinal scale.