ABSTRACT
By the mid-thirties several methods of measuring the size of radio audiences were in use in the U.S.A., but they had grown up in response to a need which did not exist in Great Britain. Virtually all radio was commercial there, American stations deriving all their income from selling time to advertisers who would fill it with 'sponsored' programmes in the course of which they would promote their wares. Naturally the stations' space-salesmen had to be armed with information about the audiences they could 'deliver' to the sponsors. It was soon found that advertisers were not satisfied with vague statements about the number of people who habitually listened to a station. They demanded information about the number of people - and the kinds of people - who actually listened to it on specific occasions. And they wanted this not only because it was a better guide to the value of what they were asked to buy but because, once they had bought time, it would show them whether in fact they had bought what they hoped they had. These measurements of audiences became known as 'ratings' - rather unfortunately for the word 'rating' could equally well describe measurements of other kinds, such as of audience enjoyment or comprehension.
