Tobacco Documents Online

SUBJECT: FTS's proposed cigarette warning size study 1. FTC [-Loesn't know what it's trying to do (or at least won't admit it in the suppt-irting statement). They say that they are trying to find "... the best siz=! for a health warning ...". But they don't know what best means. They seem to want to make the warning more noticeable, and yet not detract from the advertising message (page 2, top). Now you can only fit a given amount o7' information onto a page, and dn increase in the warning can only come fro,,L a decrease in the advertising. (There is no free lunch.) So I don't knt)w what they want. Perhaps ;they have some golden mean in which the warning i.s more apparent, but not too much, while the advertising is decreased, but only a bit. I'd like to see a definition or two. 2. Givern that FTC is trying to make the warning more apparent to the reader, this stu~y only looks at a piece of the problem. They have already studied shape, tlf.ey say; now they look at size. How about design? Content? Any advertis=r or graphics person knows that print is a lousy.way to get a message across qi:.ickly. How about something like the following, if they're really serious ?bout all of this? , CTG-I'QETTCS 3. Given that FTC is constr&ined to a simple print measage, with size and shape the only variables left, the study may make some sense. The universe is not defined (i.e. the population that one is studying). Since we are worried about print messages, it should be all those who read magazines-or newspapers. The populgtion actually investigated is those who go to shopping malls and who consent to be interviewed for a half hour. It's ha.rd to make a case that the Latter are representative of the former. Besides, ang1 statistician will tell you that quota samples are not representative. So the study may give ideas about the effects of different size warnings, but will not be able to say anything definite (no statistical accuracy or anything like that). With thin understanding, the categories chosen for sampling (young-old, male- female, snoker-nonsmoker) are there for convenience, mothing more. The education_.31l level of the household head is apparent-j:y a way to sort on the general cil.tural level of the family, useful for minors. 4. I susaect that most market surveys are done by second parties, precisely so that t'ze company sponsoring the survey will not influence the results. You want cespondents to think they are talking to somebody neutral. I see no problem in doing the same thing here. You have volunteers, after all; there's no compulsion. ram ft "M of FTC: ~'1,4Tc6oa P.O.BOZZe. t, SPI{tNGFIF.Li., VA t .1 t 7C3/494-.;f3 i . r. C t- i 61 TIMN 0061096