The idea of sampling is easily absorbed. Discuss the vocabulary, especially

"population," carefully. An opinion poll such as the Gallup Poll asks individ-

ual opinions. So the units are individual persons and the variables measured

are the answers to the questions asked. But the population involves many

choices: What lower age limit? American citizens only, or all U.S. residents?

The pervasiveness of sampling is worth emphasizing by example.

Example. Everyone is familiar with those maps of world sea-routes which

indicate the volume of traffic on each route by the width of the arrow flowing

(say) out of New York across the Atlantic and dividing to reach London,

Antwerp, etc. The Oxford Economic Atlas has such a map. It was compiled

by selecting a sample of all ships listed at Lloyd's of London, the insurer of

last resort. The location on a chosen date of each ship in the sample was

ascertained and the arrows on the map were drawn accordingly.

Example. The J. D. Power company conducts two sample surveys that are

widely quoted in automobile ads. The Initial Quality Survey concerns me-

chanical problems in individual models of new cars and light utility vehicles,

such as the Chevrolet Beretta. Power randomly selects about 90,000 new car

registrations within the first 90 days of ownership soon after the new model

year begins, with results published in February. About 35% of the people

selected respond.