Sponsor (1956)

THREE-HOUR RADIO RATINGS Pro: Hooper likes larger samples that would result. Con: Pulse feels long-span rating hides important data. Agencies doubt long span is answer : . ■ : . . . ■ I . ' ! i : . PROPOSAL iuthor of plan lor threehour radio ratings is If. Ward Dorrell, Blair research chief. He said lhe\ would provide stable figures, bigger samples, quicker reports. Dorrell also argued that pre sen t-day saturation Inning techniques make the current 15-minute radio rating, nun in common use. unnecessary. llllll!llllllll!lllll!lll!l!lll!illlll!lllllilllllllllll!lll!llllll!IIIIIIIIIUIIIHIIIIIIII!llllll!l!llllim REACTION Pulse's Roslow holds three-hour ratings would hide much data on radio listening habits, feels stations need 15-minute data. Hooper's Knipe says monthly pattern of day-part indices supplemented by seasonal 15-minute figures will meet the needs of the sponsor l)i. >\<lney Roslow James L. Knipe niiniliiiiii mum ■ minimi 11 iiii:iiiiii!ii;iiiniiiiiiiimiiiiiiiiiiiiii!iiiiiiiiiiiiiiiiiiiiiiiiiiiiiii!iiii:i!iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiir I he proposal that radio rating services measure listening in large blocks of time, rather than in 15-minute segments as at present, is not regarded as a basic solution to the radio ratings problem b\ admen; two ratings firms differed In their appraisal. Made in the 8 December issue of sponmik b) W. Ward Dorrell, research chief at John Blair and Blair Tv. the proposal embraced a suggestion thai three hours be chosen as the standard period foi radio ratings! Dorrell pointed out, however, that an) lengthening in the time span would provide greatei u i huh \ and stability to radio ratings, ll< also argued that present-da) saturation Inning techniques make the 15minute measurement span unnecessary. Vmong those questioning the value of longer time spans was Dr. Sydney Roslow. head of The Pulse, Inc. While agreeing that the idea was a good one for buyers of saturation radio. Roslow cautioned that the station itself needs the detailed programing profile that is provided b\ 15-minute measurements for pinpointing the station's audience strengths and weaknesses. As for the buyer, Roslow said, "longer time spans make his work easier but. otherwise, there is no big advantage. The saving in COSt which could be put into larger samples is not big enough to increase the sample much. Don't forget. you need to quadruple vour -ample to cut your error in half." Roslow. agreed, however, that the longer time span would increase Pulse?!S rating accuracy to a degree large enough to be measured. Roslow maintained that three-hour periods would hide much of importance going on in radio listening habits. He pointed to the important morning slots where, in a brief period, the home is suddenl) denuded of its males. Thus, he said, much valuable audience composition data is lost. James L. Knipe. president of C. E. Hooper, Inc., whose telephone coincidental method inherently provides that the sample size is increased in direct proportion to the increase in time span measured, stated: "Every honest researcher everywhere, always, wants a larger sample. But larger samples create costs which are impossible to all concerned. So. the answer is the measurement of longer time periods. Here is a fine example of where the economics of the radio business fits the realities of present-day timebuying. Knipe said that Hooper's present pattern in main markets of providing monthly day-part indices supplemented by seasonal reports breaking down the audience by half or quarter-hour periods would meet the needs of radio buyers today. Agency researchers, all of whom are well aware that the statistical error for many radio ratings is almost as big as the rating itself, were quick to point out that timebuyers can provide themselves with more stable ratings from current measurement figures bv merely averaging 15-minute figures over as long a period as thev want. The media director of an agencv with substantial radio-tv billings said three-hour spans would he useful if it can be proved without doubt bv the rating services themselves that radio listening is usually on a level keel. "But if a special study with a lariie sample shows peaks and valleys of listening, then we need the short time periods." A number of agencv people pointed out that, while run-of-schedule huv ing is increasing, there is still much fixedposition buying — which requires the 15-minute rating. The radio-tv research boss at one of the to]> agencies felt the Dorrell proposal implied a compromise was necessar) in the radio ratings business. He said. "We need both bigger samples and we need the 15-minute rating. \ number of agencies are willing to pa) for this. I think stations should ante up. too. ^r 38 SI'ONSOK 15 DECEMBER 1950