In 1994-5, we undertook an ambitious project for the 50-odd local radio stations of the Australian Broadcasting Corporation. These stations provided a diet of news, current affairs, and commentary. The questions posed to us were "Is the mix of topics covered what listeners want? Are we boring listeners by repeating the same types of story too often? Are we losing listeners by omitting whole categories of topics?"
We began by drawing up a list of every type of issue we could think of - some covered on ABC radio, some not. This list was gathered by combining existing program topics, interviewing listeners, and looking at various kinds of news database classification. In the end we had over 100 topics, grouped into 11 main areas:
1. Finance (12 topics)
2. Social Issues (15)
3. Education (5)
4. Crime and justice (9)
5. Health (14)
6. The environment (10)
7. Government (10)
8. Media and entertainment (16)
9. Leisure and sport (20)
10. Science and technology (6)
11. Work and industry (9)
Of course, having 126 topics in 11 groups was completely arbitrary. We could have had 20 topics, or 1000 - but 126 provided quite a lot of detail without being overwhelming to respondents.
We then held some focus group discussions with ABC radio listeners, added a few more topics which they mentioned but we hadn't thought of, and also cut out some near-duplicated topics. We ended up with 126 topics, still in the 11 groups.
The next step was to do a series of 9 surveys, one in each major Australian city. This was a combined phone/mail survey. We rang people up at random and asked them a few screening questions. If they were ABC radio listeners and had at least some interest in listening to talk programs, we asked them if we could mail them a questionnaire about the topics covered on radio.
80% said Yes. We made no attempt to force them into saying Yes - it wouldn't have been difficult, but many of the mail questionnaires wouldn't have been returned, and this would have added greatly to the cost of the survey. We used data from the other screening questions (age group, gender, stations listened to, talk/music preference, etc) to check whether those who refused questionnaires were consistently different in any way from those who agreed to accept them. We found no consistent differences.
Listeners who agreed to take part in the survey were sent a 4-page questionnaire. Half the first page was the covering note and instructions. Answering the questions was very simple. Because of the large number of questions (126) we simplified the answers down to 3 choices:
Never want to listen to this type of content
Sometimes want to listen
Often want to listen
All the respondents had to do was tick one box on each line. At the end of the questionnaire, on the fourth page, was a space for them to write in any other topic they wanted to hear about on radio.
We grouped the 126 issues into their 11 categories in a logical order. We considered having different versions of the questionnaire, maybe in a scrambled order, but decided that this might create as many problems as it solved.
The response rate was good - about 80%. However this was 80% of the 80% who accepted a questionnaire, so about 64% of those first asked. (This is a problem phone/mail surveys - people get two chances not to participate.)
And the results?
We found that the most popular topics were the ones that concerned people directly. In Adelaide, where I live, the tap water tastes foul (well, it's better now in 2012). So it was hardly surprising that one of the most popular topics in Adelaide was the quality of drinking water. In Melbourne, where the water tastes fine and doesn't run short, this wasn't an important issue.
The issues of least interest were those far removed from Australians' lives - e.g. "events and trends in Africa".
But the purpose of the study was not to find which topics were least popular and drop them from radio. It was to check the balance of public opinion against the balance of programming. Every topic needs to be included, on a radio station with any claims to comprehension - and even "events and trends in Africa" was of high interest to 10% of listeners. However we found that some of the 11 broad topic areas had little coverage. Topics in education and health had some of the highest interest levels, yet little radio time was given to these. So these surveys enabled the balance of topics to better match public preferences.
You may also be interested in findings about radio audiences.