The blood bank at a hospital has 1,200 units of blood, out of which 37% units are of blood group B+. A clinical researcher randomly selects 300 units of blood and finds that 33% of those are of blood group B+. To test his result, he randomly selects 200 units of blood and finds that 40% of those are of blood group B+. Which of the following is the reason there is a difference between the two percentages selected by the researcher? a. The samples were not random samples.
b. Random error; the numbers were different due to variability inherent in sampling.
c. Both samples suffered from non-response bias.
d. The sample sizes were both too small.