"Why did they take my favorite show off the air?" The answer lies, no doubt, in low Nielsen ratings. What is the powerful rating system, anyway? Of the 92.1 million households in America, Nielsen Media Research randomly samples 4,000 on which to base their ratings. The sampling design is rather complex, but at the last two stages it involves randomly selecting city blocks (or equivalent units in rural areas) and the randomly selecting one household per block to be the Nielsen household. The rating for a program is the percentage of the sampled household that have a TV set on and turned to the program. The share for a program is the percentage of the viewing households that have a TV set turned to the program, where a viewing household is a household that has at least one TV set turned on.

a How many household are equivalent to one rating point?

b Is a share going to be larger or smaller than a rating point?

c For the week of April 19, 1992, "60 Minutes" was the top-rated show, with a rating of 21.7. Explain what this rating means.

d Discuss potential biases in the Nielsen ratings, even with the randomization on the selection of households carefully built in.