Doctors and nurses Nurse-practitioners are nurses with advanced qualifications who often act much like primary-care physicians. Are they as effective as doctors at treating patients with chronic conditions? An experiment was conducted with $1316$patients who had been diagnosed with asthma, diabetes, or high blood pressure. Within each condition, patients were randomly assigned to either a doctor or a nurse-practitioner. The response variables included measures of the patients’ health and of their satisfaction with their medical care after six months.

(a) Which are the blocks in this experiment: the different diagnoses (asthma, etc.) or the type of care (nurse or doctor)? Why?

(b) Explain why a randomized block design is preferable to a completely randomized design in this setting.

$1316$patients were diagnosed with asthma, diabetes, high blood pressure. They were randomly assigned to nurse or doctors. They were put on medication and observed after six months.

The given patients were randomly assigned to a doctor or a nurse. The corresponding treatments are given. The blocks have to be the characteristics of patients, but they cannot be the treatments. Diagnoses have to be different blocks.

The blocks are diagnoses, asthma, diabetes, high blood pressure.

For the given above situation, the blocks are diagnoses, asthma, high blood pressure. It is not easy to compare results within a completely randomized experiment because individual health and satisfaction is related to diagnoses.

Randomized block design is preferred as there will be less variability in diagnoses block. Also, it would be very easy for the comparison of the effect of the care on response variables.