Listener surveys

[davideduardo]

...any sort of public opinion research can be terribly skewed by how the questions are asked.

And that is why questions are very carefully crafted to remove bias when done by reputable and responsible research companies.

"Don't you think that Buicks are and old people's car?" invites a particular response.

But "Are Buicks for: people much younger than you, a little younger than you, people your age, people older than you, people much older than you" will not only get a more accurate response as the question and the answers do not lead the respondent.

And referring specifically to music radio, the more or less standard question for song research is "how much would your like to hear this song on the radio today?" 1. Not at all, I dislike it a lot. 2. Probably not. I dislike it a little. 3. I don't care, I neither like nor dislike it 4. I'd like to hear it. It's a song I like and 5 I'd really like to hear it: it is one of my favorites.

Properly asked questions are very neutral and deliver accurate responses.

 

[Avid Listener]

And that is why questions are very carefully crafted to remove bias when done by reputable and responsible research companies.

I believe the carefully crafted part. I don't believe it's done to remove bias, I believe it's done so that the answers make the clients happy so that contracts get renewed. As for "reputable", I believe the most important reputation is the reputation for making the clients happy. As for "responsible", I believe they see themselves as responsible for making clients happy by telling them what they want to hear.

No one wants to hear bad news. Give a client bad news, and the client will find a new research company. I suspect one of the reasons why I have such a bad reputation in here is that I say things that aren't what you suits want to hear. It makes you either sad or angry to be told that the audiences don't love you like you think that they do, so you reject any voice telling you what you don't want to hear.

There's also a hell of a lot more to liking a song than just how you ask the question. Have people listen to enough similar sounding songs to get them into a particular mood, then throw a change-up at them, and chances are they won't like it. But set them up in a different mood, and you'll get a totally different answer. But you already know that.

 

[TheBigA]

I suspect one of the reasons why I have such a bad reputation in here is that I say things that aren't what you suits want to hear.

No that's not it. You have a bad reputation because you continually attempt to talk about things you know nothing about. That's not unusual on message boards, but some at least attempt to provide documented proof for their opinions. You don't even try to do that. To me you're another aging boomer crying to be heard.

 

[davideduardo]

I believe the carefully crafted part. I don't believe it's done to remove bias, I believe it's done so that the answers make the clients happy so that contracts get renewed.

First, most radio research is done on what would be called an "ongoing relationship" basis. The station or the group does regular research, and the research company respects exclusivity in the market or market segment. The relationship continues as long as the station(s) has a research budget and the research company brings real, actionable data that retains or grows the station(s) position in the marketplace.

Delivering intentionally inaccurate data will destroy the research company and the broadcaster.

Clients are "happy" when research achieves its objectives which is to maximize station potential. That means, more than anything else, finding if a station is doing anything that is negative... bad songs, bad mix, bad jocks, being perceived as being not as good as the competition, vulnerability to a new competitor, etc.

As for "reputable", I believe the most important reputation is the reputation for making the clients happy. As for "responsible", I believe they see themselves as responsible for making clients happy by telling them what they want to hear.

As BigA said, you know nothing about this. The reputation a radio research company has is based on the success of its clients, word of mouth among industry leaders, understanding of new trends and actionable data once hired.

No one wants to hear bad news. Give a client bad news, and the client will find a new research company.

Bad news is generally accompanied by ways of finding solutions, which make a station get more audience. The first thing that we look for in research is indication of anything going wrong and we want to address that first.

Keep in mind that probably 90% of music station research today is music research. So the "bad news" is "there are songs you should no longer play". We get MScores from MediaMonitors to tell us that so that we can yank the stiffs. We do "callout" so we can find the klunkers. We do music tests to weed out the duds. And then we use the good news... the data about the positive testing songs... to make a powerful playlist.

So we desperately want to hear the bad news. Because by fixing it , we may get better ratings.

I suspect one of the reasons why I have such a bad reputation in here is that I say things that aren't what you suits want to hear. It makes you either sad or angry to be told that the audiences don't love you like you think that they do, so you reject any voice telling you what you don't want to hear.

No, you are annoying because you show no respect and you state things that demonstrate a high degree of ignorance of the way things are actually done.

There's also a hell of a lot more to liking a song than just how you ask the question. Have people listen to enough similar sounding songs to get them into a particular mood, then throw a change-up at them, and chances are they won't like it. But set them up in a different mood, and you'll get a totally different answer. But you already know that.

What radio wants to know is whether to play a song or not. And if we play it, how much.

I've told you before that we commonly repeat a number of songs in a test, placed in different places between different songs. The scores, using 1-100 dials, generally differ by less than 1% which is amazing considering that this means that people placed their dial at almost exactly the same place each time the song came up.

There is no position bias. And I've conducted or participated in waaaaay over 1000 tests... as moderator for hundreds of them, too.

When we test in individual settings, either on appointment-driven tests rather than larger groups, or where the test is done on the Internet, the song order can be scrambled each time so that the sequences are random, minimizing position bias.

I've also done replication studies where the same music, in different sequences, is tested in a different location on a different by close date in the same city. The results are essentially the same. I did at one time 3 separate test groups, each divided into 2 sessions, in three parts of a Top 5 marke... one set on the East Bay, one South Bay and one on the Peninsula... 6 sessions where the results were statistically identical each time.

That is what I know, not what you think I know.

 

[semoochie]

Let's see. Top five market, East Bay, South Bay and Peninsula: I'll say "San Francisco"! Am I close? I must be within four.

 

[JeffM]

.

No one wants to hear bad news. Give a client bad news, and the client will find a new research company. I suspect one of the reasons why I have such a bad reputation in here is that I say things that aren't what you suits want to hear.

There it is! The "you suits" rant again! It's like waiting for a duck to quack.

 

[K.M. Richards]

I don't know which is more shocking. The fact that you and I appreciate an outstanding Britcom, or the fact that after going to all the trouble of actually finding the clip and a link to it, you have absolutely no comment on the whole point of me mentioning it, which is that any sort of public opinion research can be terribly skewed by how the questions are asked.

I thought I was helping your argument by providing the actual clip so everyone could hear what you were referencing. Pardon me for trying to be useful.

 

[davideduardo]

Let's see. Top five market, East Bay, South Bay and Peninsula: I'll say "San Francisco"! Am I close? I must be within four.

Well, since there are no bays in Dallas and Chicago could not have an "East Bay" that does narrow it down. Odds are 1 in 3 you are correct.

 

[Avid Listener]

I thought I was helping your argument by providing the actual clip so everyone could hear what you were referencing. Pardon me for trying to be useful.

Since you made no comment, other than to point out that it was the latter seasons of the show when Jim Hacker was elected PM, I thought you were attempting to "one-up" me by pointing out an error and demonstrating that you were better at finding links.

 

[semoochie]

Well, since there are no bays in Dallas and Chicago could not have an "East Bay" that does narrow it down. Odds are 1 in 3 you are correct.

Probably the only east bay in Los Angeles is on one of the FM antennas on Mt. Wilson. This reminds me of a story I can't quote but involved(I believe)Ira Blue. An edict came down at KGO to refer to the various areas of the market as the East, South, North and West Bays. The exasperated host involved said he would(Fill in your own story here.)"...before I will refer to MY San Francisco, as the West Bay!".