Research: Friend or Foe

[deltas69]

too much air clutter i supoose ..used to hear

 

[deltas69]

and he continued..hear WLS AND WABC at night back in the sixties/early seventies..but can't anymore..just as well..no "boogie check" to hear anyway.. :'(

 

[fireworks]

As I mentioned earlier, I believe Research is a great tool, that has to be used with care and balance. If you subscribe to David E, Research will be the death of programmers. Keep in mind, that not all research is the same. As David said, if you use the right research incorrectly, you could be doomed. If you get the wrong research, you could be doomed. If you do well with the wrong research, does that mean the research is actually wrong or does it mean there is no real competition in your market? By the way, Mr. E, you contradicted yourself in a previous post. Makes me wonder if even you believe what you claim to know.

 

[davideduardo]

As I mentioned earlier, I believe Research is a great tool, that has to be used with care and balance. If you subscribe to David E, Research will be the death of programmers.

I definitely did not say that. I said, with "illustrations," that research which is badly done or badly implemented can be worse than no research at all. There are many variables, starting with whether the recruiter a station's research company uses locally actually screens the respndents well. And, if the station specifies the wrong recruit, equal harm is done. And, if the results are perfect (no such thing in an affordable random sample) failure to understand how to implement or failure to implement on the right sort criteria will also be disasterous.

This is no different than saying that a new antenna, if installed without getting moisture out of the line, will burn up. There are do's and dont's in every aspect of business.

Keep in mind, that not all research is the same. As David said, if you use the right research incorrectly, you could be doomed. If you get the wrong research, you could be doomed.

Correct. Those of in markets where nearly every station researches have seen evidence of both the good and the bad.

If you do well with the wrong research, does that mean the research is actually wrong or does it mean there is no real competition in your market?

How, then, would you know that the research is "wrong?"

By the way, Mr. E, you contradicted yourself in a previous post.

Try reading it again. Perhaps a second time will help.

"Eduardo" is a second given name, not a surname. Sort of like "Billy Bob" is not Mr. Bob.

Makes me wonder if even you believe what you claim to know.

Click the link.

I am simply answering another poster's excellent question as to whether cases have been seen of bad things hapening in the name of "research."

 

[fireworks]

I definitely did not say that.

I didn't say you said that.

How, then, would you know that the research is "wrong?"

Bingo. Interesting, you didn't try to answer the question. You only asked another question. I'd recommend a course in Grammer Etiquette.

Try reading [the previous post] again. Perhaps a second time will help.

...without even questioning what was possibly misinterpreted. This is the mark of real research.

 

[TheBigA]

...without even questioning what was possibly misinterpreted. This is the mark of real research.

I think that point was handled well by onetake earlier in this thread:

But even good research is lost to a stubborn programmer who can't believe what he sees and his mind was made up before the project started.

 

[davideduardo]

Bingo. Interesting, you didn't try to answer the question. You only asked another question. I'd recommend a course in Grammer Etiquette.

I am not going to even try to guess what "Grammer Etiquette(sic)" is. Or why it is capitalized.

I can not answer a question that is, as Churchill said, "... a riddle wrapped in a mystery inside an enigma. " Of course, he was talking about the Soviets, but it applies to your question. If a research project is implemented and the stations continues to be successful or increases its success or debuts successfully, there is no way to call the research "bad" because it achieved its objectives.

If, in the course of human events... er, analysis of the research, procedural faults are discovered, then the question would be, "did we serendipitously stumble upon another format or derivitave?" The issue is whether to adjust the specifications for the next project to purposely incorporate the accidental elements and, at the same time, try to understand why the project yielded good results.

This is one of the reasons that the station's job, which is defining the objectives, specifying the recruit specifications and analyzing the results requires a skillset that not all managers, program directors or station staffers in general posess. All this brings us back to the fact that research is only as good as the correctness of the individual parts or steps.

 

[loves radio 2]

Bingo. Interesting, you didn't try to answer the question. You only asked another question. I'd recommend a course in Grammer Etiquette.

I am not going to even try to guess what "Grammer Etiquette(sic)" is. Or why it is capitalized.

I can not answer a question that is, as Churchill said, "... a riddle wrapped in a mystery inside an enigma. " Of course, he was talking about the Soviets, but it applies to your question. If a research project is implemented and the stations continues to be successful or increases its success or debuts successfully, there is no way to call the research "bad" because it achieved its objectives.


If, in the course of human events... er, analysis of the research, procedural faults are discovered, then the question would be, "did we serendipitously stumble upon another format or derivitave?" The issue is whether to adjust the specifications for the next project to purposely incorporate the accidental elements and, at the same time, try to understand why the project yielded good results.

This is one of the reasons that the station's job, which is defining the objectives, specifying the recruit specifications and analyzing the results requires a skillset that not all managers, program directors or station staffers in general posess. All this brings us back to the fact that research is only as good as the correctness of the individual parts or steps.

David are you a radio professional? Not trying to be a jerk, but I never heard of you untill a few weeks ago.

 

[fireworks]

As in politics, if you can't impress them, baffle them. Mr. "E" has the basis for any business. But you can copy these lines off an Internet site and re-word it to fit your agenda. My guess is he's looking for a job.

 

[romer979fm]

My guess is he's looking for a job.

Google "David Gleason"...
(you may then want to guess again)

 

[davideduardo]

My guess is he's looking for a job.

;D

 

[davideduardo]

David are you a radio professional? Not trying to be a jerk, but I never heard of you untill a few weeks ago.

Coincidences are funny (funny=peculiar, not funny=haha). This week, I was talking about pendulum swings and mentioned Bill Drake, Ron Jacobs, Tom Rounds and some others who put Top 40 back on the tracks... and nobody I was speaking to knew who they were.

I also referred to the pioneers of call-out, like Todd Walace, Jack McCoy and Bill Tanner. Call-out in 1975? No way, I was told.

So the probability that you have not heard of me is immense.

 

[Tibbs4]

http://www.dave-gleason.com/


;D That one's for Buddy and Chris and Shirley.

 

[Tibbs4]

David are you a radio professional? Not trying to be a jerk, but I never heard of you untill a few weeks ago.

Coincidences are funny (funny=peculiar, not funny=haha). This week, I was talking about pendulum swings and mentioned Bill Drake, Ron Jacobs, Tom Rounds and some others who put Top 40 back on the tracks... and nobody I was speaking to knew who they were.

I also referred to the pioneers of call-out, like Todd Walace, Jack McCoy and Bill Tanner. Call-out in 1975? No way, I was told.

So the probability that you have not heard of me is immense.

None of your clients have studied the past have they? Okay that was a bad attempt at humor (haha) from David's line under his moniker. Seriously, I have never thought about the fact that "the art of radio" has been forgotten. Not saying all that is a good thing to have happen, but seriously, (Shirley) how can radio
improve and grow if so many decision makers don't grasp the process?

 

[davideduardo]

None of your clients have studied the past have they?

I would not expect any of my clients to know the chronology of US CHR / Top 40, but I would hope that PDs in the US might have an idea about it... particularly as we go through yet another transformation of what radio is.

 

[fireworks]

http://www.dave-gleason.com/

Well, it's nice to know he's got a song coming out. Maybe he can use it for an audience test one day.

 

[davideduardo]

Well, it's nice to know he's got a song coming out. Maybe he can use it for an audience test one day.

Very funny. I am not a singer, which is no doubt a great relief for the entire universe.

If you would re-read the last few posts you might realize that you have been spoofed. :-\

Try following the link below instead.

 

[fireworks]

If you would re-read the last few posts you might realize that you have been spoofed.

... and when will you realize you've been spoofed?

Try following the link below instead.

I saw it. It's an ode to you. Congratulations. Nothing boosts the ego like having your own website. Try Head Hunters, or Kelly Services. They can find you some temporary work. And send them to your website instead of giving them a resume. They might be a little intrigued at first, but that will subside quickly.

 

[Tibbs4]

I have a bad feeling this is headed south...and not to Birmingham.

Kiss up time ~ Hey David, gotta a legit question. When you play 8-10 seconds of music clips, have you ever tried rotating the order of the songs and gotten back to back reverse reactions?
To be more clear, like in a wine tasting, if you get "the order wrong" the next couple of wines will taste "wrong" or "worse". I have to think it's possible in music rotation, as well. One of my best decisions was to go over the music and keep a general flow going vs. train wrecks. Same has to be true in testing. I have also always thought it was interesting to see and compare the results of these tests among two groups. One was tested at night after work. The other was tested on Saturday or Sunday. People are more relaxed, etc. They'll prefer different music under different conditions. How do you feel about the fact that external conditions may alter the survey and adjust for them? If this is unclear, let me know. Hope it makes sense.

 

[davideduardo]

I have a bad feeling this is headed south...and not to Birmingham.

More like Hartselle... :-\

Hey David, gotta a legit question. When you play 8-10 seconds of music clips, have you ever tried rotating the order of the songs and gotten back to back reverse reactions?

Absolutely.

The answer is more complex, though.

Hooks that are 10 seconds are probably too long; if too long, the participant will fatigue sooner. 8 seconds is beyond the time when essentially everyone has scored the song, any longer and the test takes too long and has to be limited to fewer songs.

What is done first by a research company is to first prove that there is no poisiton bias. "Position bias" is the placement of each song in each set of songs... typically, songs are presented in batches of 50 to 70. Among, let's say, 500 songs, some songs are put at the beginning of a set and repeated in other positions. And songs that are in the test in the first 100 or so are repeated towards the end. The finding is that position is not a significant consideration.

Some companies do a 100 person test in two sessions of 50. The orders are reversed or scrambled in each test to avoid fatigue from some respondents getting tired towards the end.

If testing is individualized, each respondent can have a differently scrambled or reordered set of songs. Callout typically uses 5 to 6 different start points for 25 to 30 total songs.

While we see via internal testing that position is not a bias factor of significance, we use technology to shuffle or reorder as much as we can.

To be more clear, like in a wine tasting, if you get "the order wrong" the next couple of wines will taste "wrong" or "worse".

I've never seen any evidence of this, despite including a control group of duplicates in most tests and comparing the same song across multiple markets and multiple tests. Since songs on a test are slated (a number is spoken ahead of each song for paper based score sheets) or separated by a short silence for tests using scoring dials.

I have to think it's possible in music rotation, as well. One of my best decisions was to go over the music and keep a general flow going vs. train wrecks.

While a very useful programming technique, it's not particularly useful in testing. In fact, an objective of list preparation is to randomly shuffle and get uniform same artist separation.

Same has to be true in testing. I have also always thought it was interesting to see and compare the results of these tests among two groups. One was tested at night after work. The other was tested on Saturday or Sunday. People are more relaxed, etc. They'll prefer different music under different conditions. How do you feel about the fact that external conditions may alter the survey and adjust for them? If this is unclear, let me know. Hope it makes sense.

Similarly, two groups in the same night, one at 5:30 and another at 8 PM will have differences. This is because the two groups are different! The ideal situation would be to do 25 people at 2 PM, 25 at 6 PM, 25 at 8 PM, and 25 Saturday afternoon, but the cost would increase astronomically and just doing two sets is usually enough. In any case, what is important is the relative score vs. all other songs. The actual score is less important than the relative score.

Further, when factor / cluster analysis is done on the results using packages like SPSS, lifestyle or taste subsets can be identified, just as age variations can be via collection of the age of each participant.