Doc: I heard a word about Internet stuff called “Spim.” Do you know what that is? - Anonymous
Anon: SPIM is a relatively new term that refers to SPam sent via Instant Messaging. Here is a good explanation: SPIM
I also set up an Internet search for you for more information. Click here: More SPIM Information
Doc: I recently bought a socket
set and there is a small blue disc included with the set. The directions call
it a "spin disc." I never saw that piece before. Do you know what it's for? -
Anon: If you have ever used a socket wrench or any ratchet tool, you probably know that sometimes the nut or bolt you are trying to turn is too loose for the ratchet to work properly—the ratchet just spins around and does nothing to the nut or bolt.
The spin disc is designed to solve that problem. If you have a nut or bolt that is too loose, slip the spin disc on the handle (the head end where you attach sockets), and then slip on the socket you plan to use. You'll notice that you can turn the spin disc with your thumb and index finger to turn the loose nut or bolt until you get it tight enough for the ratchet to work properly. It's a neat little invention.
I am trying to find freelance work in play-by-play broadcasting. I have years of experience calling games in various formats (Internet webcasts, video dubbing, etc.), but not in radio proper. I am not in a situation where I can pack up and work at a small market station for $1000-$1500 per month (I have had those offers). Are there talent agents that could help find a match? Can you recommend a better course of action? Thanks. - Anonymous
Anon: I sent your letter to Paul Douglas, former PD and now with Cox Radio Syndication in Atlanta. This is what he said:
"Based on what this person said about his/her experience, I would have to say that it would be difficult, if not impossible, to be considered for a major league professional sports or big time college play-by-play sports job. There are talent agents who represent high priced talent, but I’m not aware of a "talent matching" service. In addition, most (if not all) play-by-play jobs are not freelance, but rather full time employees of the station under contract to carry the games, or employees of the team itself, which is the most likely scenario in college sports. I’m afraid the "usual" road one would take to the top is the one you are not able to consider, which is start small and work your way up."
I agree with Paul and will add that I checked with a few other people who also said that your lack of experience wouldn’t become a positive if you hired an agent. In other words, your agent—like you—would only be able to promote the fact that you have only had experience in webcasts and video dubbing. I know this is tough, but the best way to get into this field is to start in a small market somewhere. However, as an alternative, you might want to send a demo tape to a few people to get their reaction.
I know that many sports play-by-play announcers read this column, and maybe one of them will send a note with another suggestion.
Please tell me your thoughts on programming sports radio. - Anonymous
Anon: I have no idea what you mean, so I’ll take a guess. I’m sure you’ll love this response.
I think a lot of time can be saved with sports, including sports programming on radio, if these rules were followed (major sports only):
Baseball: Give each team one run and let them play one inning.
Basketball: Give each team 100 points and let them play for one minute.
Football: Give each team three points and let them play for two minutes.
Hockey: Give each team one goal and let them play for two minutes. Name every player Mario.
For each of the sports . . . all the teams could be brought to one location and each season would last one day.
Is that what you’re looking for?
We are running a spot load of 10 or more minutes in our Breakfast Hour plus a good amount of sponsored games that include sponsor messages. This is in addition to news, local information, weather, entertainment bits, and music. I feel the show is getting a bit congested. Is there any research on a the maximum amount of spots in breakfast before people tune out or away? - Anonymous
Anon: Your question addresses one of the most controversial topics in radio. It is a controversy that has two definite sides—sales and programming. This is what I know . . .
I have conducted (and seen) many studies that investigated the willingness of radio listeners to listen to increased spot/promo loads—particularly in morning shows. In every case, the listeners’ enjoyment of the station (TSL) decreases in direct relationship to the increase in the number of spots, contests, and promos that a radio station airs.
The listeners understand that commercial radio stations must have commercials to stay on the air—they don’t complain about that. What they complain about is the amount of commercials in a given hour. They say that they still listen to their favorite radio stations even as the spot loads increase, but they progressively reduce their TSL. When they get tired of (or upset over) the commercial content on their favorite radio station, which steadily increases over time, they switch more frequently to other radio stations, CDs, or tapes. (If another radio station choice has the same heavy spot/promo load, the process repeats itself.)
The data clearly show that many radio stations (by themselves, not outside influences) are systematically reducing listener TSL. However, the comments I have heard from radio station owners (or regional VPs, etc.), do not suggest a concern for the problem. The concern is to produce a good bottom line and "shareholder value." The emphasis is placed on everything but quality programming that radio listeners say they want.
The data show that radio stations are forcing their listeners to go to other entertainment sources, and I’m surprised at the lack of concern by the people who are in charge. Maybe there will be more concern when their radio stations have a "0" TSL. The question will then be, "What the hell happened? Get a new PD (or GM).
This scenario reminds me of TV shows I have seen about murders committed via arsenic poisoning. In those cases, the murderer gives arsenic to the victim in small doses over several weeks or months. The victim suffers, but attributes the problems to the flu or other sickness. Eventually the arsenic concentration builds up to a lethal level and the victim dies. Everyone but the murderer is surprised and shocked by the "sudden" death. The data show that increased spot/promo loads are radio’s arsenic. (The same thing applies to commercial TV.)
Spot Loads - More Questions
You do a great job pointing out speculation that is presented as knowledge. Have you seen any data to confirm the widespread belief that increasing spot loads are pushing down TSL? (I ask this, aware that a company on your client list has tried reduced spot loads as a competitive weapon.)
Also, any observations relative to the Arbitron/R & R/Edison Media Research finding that obnoxious spots are a bigger tune out among listeners 35+ than are stopset lengths, per se?" - Sponsored-Out in Spokane
Sponsored: Thanks for the comment about the column.
Question 1: I am not breaching any confidentiality by saying that there is a lot of evidence showing that increased spot loads affect TSL. Many radio listeners have simply given up because of the number of commercials they have to listen to during any given hour. From what I can remember, the spot load affect on TSL started showing up around 1994. But as is the case with many things in radio, the management looked the other way.
Question 2: I can’t comment specifically on the Arbitron/R & R/Edison Media Research study because I haven’t seen it (the methodology included). However, the hundreds of studies I have seen since about 1982 show that obnoxious commercials and long stopsets are virtually tied as tune outs. The Arbitron/R & R/Edison Media Research is only one study. I tend to go side with the results of many studies.
I’d like to get your opinion about something. Is it better to have more breaks with fewer spots, or fewer breaks with more spots? I’m getting pressure from my regional VP to go to fewer breaks with more spots. - Anonymous
Anon: As I have said many times before, my opinion doesn’t matter. What matters is what your listeners say. Ask them.
However, what I have seen in many research studies is that most radio listeners prefer more spot breaks with fewer spots. I know there are many Urban Legends floating around about this topic, but that’s all they are. I hear many consultants (and others) talk about the need to have fewer breaks because of programming flow and other nonsense. However, these are only opinions, and opinions about what listeners want mean nothing.
The research from listeners shows that radio stations that have long stop sets train their listeners to tune away from the station. Case closed.
A PD recently repositioned our stop sets from :15 and :45 to :23 and :53. Reasoning: people are on "appointment schedules," meaning they need to do things (i.e., clock in at work, hair appointments, etc.) at the top and bottom of the hour. So, they are turning off their radios just before their appointments (the top and bottom of the hour) so it's a good time to play spots (the channel changing menaces). This makes sense on the surface but I started wondering, then why doesn't every station do this?
Is there solid research that shows the best placement of stop sets (say, 14 :60 units per hour)? Or does it depend on what your competition is doing? (Countering them?) Thanks. - Lynn
Lynn: Let’s test the hypothesis that the people ("they") are on :30 or :60 appointment schedules. If you are near a highway right now, look out your window at a time other than :30 or :60. If there are any people driving around, then "they" are either early or late for their appointment. Oops. So much for the hypothesis that "they" all set appointments for :30 or :60. (Although they might have set a :30 or :60 appointment to drive around the city for a while. But would they have their radio in the car on or off? Hmmm…)
Does this mean that your PD’s justification to place stop sets at :23 and :53 is wrong? No, but it is just as "right" as the PD who places stop sets at :19 and :49 because that’s the year he/she was born. In other words, this is a "seems like" decision: "It seems like a lot of people make appointments at :30 or :60 and that’s how I will place the stop sets." Now if there is proof of that, then OK. However, my guess is that it’s just a "seems like" statement.
If you’re trying to figure out when to place your stop sets, the best thing to do is ask your listeners. So let’s say that you do and you find out that a majority make :30 and :60 appointments. Well, a majority could be as little as 51%, which means that 49% don’t make :30 and :60 appointments. If going with the majority is OK with you, then that’s fine, but just realize that there are many "theys" out there who won’t fit your model.
We humans are a curious lot when it comes to describing a sample situation to the population. For example, assume that a radio station puts on a new morning show. How many times have you heard something like this from a person who works at the station?
Person: Everyone is listening to the new morning show! It’s great!
You: What do you mean by "everyone?" Who did you talk to?
Person: A whole bunch of people. They’re all listening!
You: How much is a "whole bunch?" How many did you talk to?
Person: Oh, about 20 or 30.
You: That’s everyone? That’s only 20 or 30 people.
Person: I know, but everyone is listening!
You: (I give up.)
You asked what the research shows in reference to placement of stop sets. What it shows is that there is no single approach that works better than any other because there is such a broad range of listeners who behave in an equally broad range of ways.
Finally, I have never seen any information showing that placing your stop sets similar to or different from your competitor has an affect on the audience. For example, assume that you place your stop sets at different times than your competitor. How do you know that your competitor’s listeners tune to you only when they go into a stop set? (Sure, there is sharing data in Arbitron, but does the sharing happen only when there are stop sets? No.)
When your competitor goes to a stop set, maybe some of their listeners actually listen to the commercials. Maybe some go to your station, but maybe some go to a different station. Maybe some go to tapes or CDs, or maybe some go to "off." From all the research I have seen and all the things I have learned from the top PDs and consultants is that you must know what your competition is doing (don’t operate in a vacuum), but you must base your programming decisions on what your audience wants—not what your competitors are doing. This will always be the best choice.
Our GM saw a commercial in another market and wants to use the same spot for our radio station (with appropriate call letter changes, etc.). Most of us think that the spot isn’t good for us. Is it a good idea to use another radio station’s spot. - Anonymous
Anon: My experience in testing radio station TV spots during the past 20+ years shows that it’s not a good idea to assume that a syndicated spot (usually promoted as "tested") is good for all radio stations. You need to test the spot with your listeners. Just because a spot is used in another market does not mean that it’s a good spot (one that communicates the message you want to communicate) for your radio station. The odds are high that the radio station in the other market didn’t test the spot either…so it’s merely a perpetuation of an ineffective spot.
Most radio spots are selected either because someone at a radio station saw the spot in another market, or it came on a demo tape from a production company. In most cases, radio people don’t ask, "What message is this spot communicating to our listeners?" ("Who knows, or cares, they’re using it in the New York!") What a silly way to make a decision.
I have heard the word "squid" used to describe people who ride sport bikes (crotch rockets). Where did that term come from and what does it mean? - DT
DT: The origin of the term seems to be lost, although I did hear once that it was first used in a motorcycle magazine in the mid 1980s. The term is usually considered derogatory and there are several possible explanations for the term. Some say that the word squid:
Describes the multi-colored suits many wear while riding multi-colored bikes.
Is a combined and shortened form of the words "squished kid."
Describes how the riders squirt or slide through traffic.
Describes how unprotected skin looks after sliding on the pavement: calamari.
Describes the typical attire of the rider: shorts, t-shirt, and tennis shoes. Male squids often have a "hugger" on back—a female similarly dressed who hangs on for dear life trying to make it look like she enjoys the ride.
Describes how the riders look when they are sprawled out on the pavement.
Being a motorcycle rider, I don’t condone labels of any kind to describe "types" of riders. (I don’t like labels of any kind to describe anyone or anything.) Just like any other label, the term "squid" often only alienates people. I prefer the term "rider."
You mentioned standard deviation in one of your answers. What is that? - Anonymous
Anon: In technical terms, standard deviation is the square root of the variance of a data set. What is variance? Variance is a number that indicates how different the elements of a data set are from one another. The larger the variance, the more difference there is in the data set. If variance, and therefore the standard deviation, is small, it means that the elements (let’s say music test scores from 100 respondents) are very close to each other—the respondents agree in their rating of a song or songs. (Obviously, if the standard deviation is large, the respondents don’t agree in their ratings.)
Another way to look at standard deviation is to consider it as the "average difference" each element is from the mean (average) of the data set. So why don’t research people use average difference? Oh, I don’t know. Probably because it doesn’t sound as cool as "standard deviation."
Standard Deviation - 2
My question is: How did you get 1.1 as the standard deviation? Can you give us the formula? - Anonymous
Anon: I just made up the 1.1 standard deviation for the example.
Variance: The formal definition is: the mean of the squared deviations from the mean. In words, here is what you do: Subtract the mean of the group from each score in the data set. These are "deviation scores" (the difference each score is from the mean). Square each deviation score and add them together (the "sum of squares"). Divide the sum of squares by N-1 (the number of elements or subjects minus 1).
Standard deviation: The standard deviation is the square root of the variance.
Standard Deviation - 3
I know that I’m probably a pain in the butt with my standard deviation questions, but I just want to learn. What is the standard deviation of these numbers . . . 5, 5, 7, 8, 11, 14, 15, 16, 18, 21? I come up with 5.34. Is that right? - Anonymous
Anon: Nope, as it is said, that don’t be right. The answer is 5.63. You divided the sum of the squared deviations by 10 instead of 9. Remember, the denominator in the formula is N-1.
Standard Deviation - 4
This may be too simple for you, but what does “standard deviation” mean? - Anonymous
Anon: Hey, nothing is too simple if you don’t know the answer. Don’t worry about that.
Your question is an example of why some people don’t like statistics, and it doesn’t have to be that way. Once you know what the terms mean, statistics is much easier. And your question provides a great example. I hope the following discussion will help.
Let’s say you have listeners’ ratings on a 1 to 10 scale for a song from callout or an auditorium test. This data set (group of scores) will have an average score, or mean. As you probably know, the mean is computed by adding together all the respondents’ ratings and dividing by the number of respondents. Let’s say that the mean is 5.5, right in the middle of the scale. The 5.5 is the average score, but there is something missing.
What’s missing is that we don’t know if the mean score of 5.5 was produced by all the respondents giving mid-range scores or whether it was produced by one-half of the respondents giving the song a low score and the other half giving the song a high score (a “love-hate” rating). And that’s what we need to know—where did the mean score come from? That is where the standard deviation enters the picture.
Sometimes statistical terms themselves create problems. Instead of “standard deviation,” try this term—“average difference.” That’s what the standard deviation is: the average difference each respondent is from the mean of the set of scores.
If the standard deviation is close to zero, it means that the respondents agreed on their rating of the song—their scores are very close to each other. If the standard deviation is high (it can only be as high as the highest number in the rating scale—in this case, 10), it shows that the respondents did not agree on their rating of the song—the ratings are mixed from low to high. (By the way, in a music test, you always hope that the standard deviation for each song is low—each song has its own standard deviation—because it helps in your analysis. If all your songs have high standard deviations, it means that the respondents didn’t agree in their ratings. This doesn’t help with your decision to play a song.)
Standard Deviation - 5
Doc: Would you please explain “standard deviation?” Thanks. - Anonymous
Anon: Statistics and research have several terms that drive people crazy. Standard deviation is one of them. Think of it this way…standard = average and deviation = difference. The standard deviation of a set of numbers gives you the “average difference” each element (a person or thing) is from the mean (average) of the data set. A small standard deviation indicates that the elements or people in the data set are the similar. A large standard deviation indicates that the elements or people are different from each other.
For example if you conduct a music test using a 1-10 scale, and you have a song that has a standard deviation of 4.5, you know that the respondents who rated the song do not agree much in reference to their rating of the song. On the other hand, if a song receives a standard deviation of .1, you know that the respondents agree in their ratings of a song.
A standard deviation cannot be a negative number. The smallest standard deviation is 0.0, which means that there is 100% agreement among the respondents or elements in the data set. A standard deviation cannot be larger than the largest number in the scale use. That is, in a 1-10 scale, you cannot have a standard deviation above 10.0. However, in reality, with a 1-10 scale, the largest standard deviation you’ll get is a 4.6 if 50% rate the song a “1,” and 50% rate the song a “10.”
For a detailed discussion of standard deviation click here. You may also want to read about some of the other statistics shown on the page.
Standard Deviation Calculation for Grouped Data
Hi Doc. I’ve read all your articles on standard deviation and understand the concept and how useful it can be in interpreting song data. But I’m having trouble calculating a song’s standard deviation score in Excel.
Let’s say we’re using a scale of 1 to 5 to score a particular song—5 respondents score a “1,” 5 score a “2,” 24 score a “3,” 44 score a “4,” and 25 score a “5.” So the spreadsheet has the totals in each cell and would look like this: Cell A1=5; B1=5; C1=24; D1=44 and E1=25.
Now if I use the formulae =stdev(A1:E1) I get a standard deviation of 16.32 (which I know is totally wrong) because the correct answer should be 1.03 (I think). Incidentally, I worked out the second score by entering a score of 1 into five cells, 2 into five cells, 3 into twenty four cells and so forth.
Is there a formula that makes the job easier, or am I doomed to having to enter enormous amounts of data into separate cells? I hope this makes sense.
And thanks for the column, I learn something every day from both yourself and contributors. - Huw
Huw: I’m glad you enjoy the column and learn things. That’s my goal. Your question is very interesting. First, you are correct in saying that the standard deviation of 16.32 is incorrect. That’s the standard deviation of the sum of scores and that isn’t what you want. Second, you are also correct that the standard deviation is 1.03. So now what?
I tried to save time and searched the Internet for an Excel spreadsheet example for computing standard deviations for grouped data. Bummer. Couldn’t find one.
My only option was to create one for you and you can find it by clicking here: Standard Deviation for Grouped Data.
Click Here for Additional S Questions
All Content © 2018 - Wimmer Research All Rights Reserved