Broadcasting Telecasting (Jan-Mar 1958)

.•iMRTISERS S AGENCIES continued What happened in the Syracuse rating test QEORGE ABRAMS, vice president and advertising director of Rcvlon Inc. and chairman of the Assn. of National Adversers Radio-Tv Committee, explained his proposal for a single rating service in an exclusive recorded interview with Broadi \mi\(, editors last week. With Mr. Abrams was Miles A. Wallach, head of M. A. Wallach Research Inc., a company which conducted a test survey, of the kind which Mr. Abrams proposes, in Syracuse, N. Y. Here, slightly condensed, is the transcript of the interview. Most of the answers were given In Mr. Abrams. Some of those dealing with technicalities of the Syracuse survey came from Mr. Wallach. Mr. Abrams, what is the basic conclusion you draw from your Syracuse study? I think the answer to that is that we have found that personal interviewing while the program is on the air can be done and can be done inexpensively enough so that you can operate a national rating service by that method. Will you review what the Syracuse study was. when you did it, how you did it? It was done the week of March 9, 1958. A sample of 10,200 homes was set up in the city of Syracuse or broadly Onondaga County. This is 10% of all television homes in S\racuse. There were three checking periods used, Sunday, March 9, between 4 and I 1 p.m., purposely to determine whether we could do late-night interviewing. The next check was done on Tuesday, March I I, between 5 and 8:30 p.m. to determine whether we could interview during the dinner hour. The third check was done on Thursday, March 13, 10 a.m. to 6 p.m.; in essence this was a daytime check and also a luncheon-hour check to determine whether we could get into the homes while mothers were preparing lunch for their family or doing their household chores. In all three cases we found we could. Why did you pick Syracuse? Syracuse for a long time has had the national reputation of being Test City No. 1 . It is used by most advertisers as a typical American city. Unfortunately from a television standpoint it has only two major stations. But this was not what we were checking. We were checking methodology. Could we do personal interviews and get the information we wanted, and how did our information compare with other methods of rating services? How many people did you put in the market to do this job? We had 58 — 29 of them doing telephone interviews and 29 doing personal interviews. The next step was that we divided the sample equally, 50% by personal coincidental interview and 50% by telephone coincidental interview. A matched probability sample using U. S. Census data projected to 1957 was also employed so that we had a good sample of homes. Based on what the Syracuse survey cost y ou, what do you estimate it would cost to do this same job for a week, once a month, on a national basis? The answer to that is you could do 80,000 personal interviews per week and 10,000 telephone coincidental interviews in the farm areas in probably 53 major television markets for an annual cost of $5 million. Could you give us some of the Syracuse findings? One of the important elements checked was the sets-in-use figure and the rating comparisons by telephone vs. personal interview, [n trying to measure this we also took our figures and compared them with the ARB figures for the period of Feb. 9, the preceding month, which was then available. Our experience as both advertising men and research men has shown that during the winter season the sets-in-use figure for established programs — programs that have been on the air for several years — does not vary very much. ARB for the city of Syracuse shows a 67.1% sets-in-use on Sunday night from 10 to 10:30. Our T-PI— Telephone-Personal Interviews — showed 47.2%. When this is subdivided by telephone and personal interview, this breaks down into a sets-in-use figure by telephone of 49.7% and 44.3% by personal interview, so that if for a moment you discard the telephone interview aspect of this, and just make a direct comparison between the diary method and personal interview, you have a 50% differential in sets-in-use for Syracuse. Now in terms of ratings, the ratings on the Lor etta Young Show (NBC-TV) and $64,000 Challenge (CBS-TV) have followed a pretty uniform pattern over the past several months. The comparison by telephone coincidental interview, personal coincidental interview and diary shows this: The diary method for Feb. 9 showed 46.6% of sets tuned to Loretta Young, and 19.7% tuned to $64,000 Challenge. The personal interview method showed 25.7% tuned to Loretta Young and 15.7% tuned to $64,000 Challenge. The telephone method, 31.5% to Loretta Young and 16.4% to $64,000 Challenge. You therefore can see if you want to apply simple statistics to it, that you get a greater variation between the two shows on the diary method as compared to personal interview. What aboui the refusal rate you mentioned— the number of people who wouldn't admit you to the house? This was presented to us at the outset as one of the problems we could expect to face. In other words, could we get into homes? I would like to add that one of the reasons I was encouraged to go further with this with Miles Wallach [head of M. A. Wallach Research Inc.] was that he had done a study for a food company in which he had actually gotten into, roughly, 7,000 homes during their eating hours. If an interviewer could stand there and watch what people were eating I then believed he could watch what they were viewing. From 10:30 to 11 p.m., which is the stopping point of our interviewing, we had a refusal rate — on telephone and personal interviewing exactly the same — of 7%, which shows that there's absolutely no difference in refusal rate between the two methodologies. In the personal concidental method, the person actually was granted access to the home to ask the questions? In over 90% of the cases, regardless of time period, the interviewer not only went into the home but went into the room in which the television set was located. There was less than a 10% refusal rate as far as entry into the house. In those, 8 or 9% of the cases we had to conduct the interview at the door or right inside the house. How long is the average interview in the home at 10:30 p.m.? The average interview took between three and four minutes. I would like to get it down to below that, but it is not feasible to do so because time is taken in going to the room where the tv set is. How many interviews actually were completed in your sample? The base was 10,200 homes. Of this, 9,100 interviews were considered acceptable. That is they were complete, determined to be completely honest interviews. The rate of interviewing by telephone was 160-170 per half hour. By personal interview 140-150 per half hour. The average interviewer did 7-8 telephone interviews per half hour. The average interviewer did 5-6 personal interviews per half hour. How did the interviews break down during the week? The breakdown of the 9,100 completed and accepted interviews was as follows: Sunday, approximately 4,000 interviews; Tuesday, 1,700 interviews, and Thursday, 3,400. What did you find with the number of viewers per set? We found on the Loretta Young Show on diary technique 2.4 viewers per set, based on their February period. On personal interviewing we found 2.0 viewers per set. On telephone, it was 1.8. Telephone gives us the lowest figure of all three techniques. Nielsen doesn't measure viewers so they are out of the picture completely. Do you know why telephone shows the lowest? I think one of the reasons is that you call up and ask a few questions and you are in the middle of a questionnaire and you say to the person, "How many people are watching the Loretta Young Show?" and they say, "Oh, I'm watching it" and just throw it off that way. I don't think they want to be bothered trying to count. We also feel low sponsor identification is due to the same desire to throw off the inter Page 32 • March 31, 1958 Broadcasting