498 



-17- 



ilgnificaad/ b«cv««n 0, C *od B, but 41rtctioaall7 all but on* prcf«rr*d 

 cither C or •) Mokcri of lov-prie*4 el(arttc«s och«r tlua Playtrt lo. i 

 w«rc ta apparcob kftb«pcloa aa Ch« Cabla tbova chac chay placad 1 top, 

 though la fac^^ aod.B vara ao eloaa togetbar aa to b« virtually aqual. 

 At tha bottoa of oSk tabla aaoVTera wttoac regular br«n4 fall ia the 



"low nicotine" categopj, (up Co 0.9 mg delivery) showed a tendency to 



>^ 



prefer B wich a delivery ik(.0.9 ■(, but directionally A wai again tha 



least preferred cigarette even ^or these svokera. The coopiratively 



snail difference between A and B iooicotioe delivery coupled with the 



a^arent rejection of A by nearly all ^tegori«s of smoker suggesta 



some v^duct deficiency in addition to ibylow olcotine delivery, though 



Ic is noi ^lear from the "spontaneous dislikel^ (Tabl* 3) exactly what 



this daficie«^ could have been. "^ 



^6 o 



(b) Straigftiorward Preference (^ \ 



'^, -^ 



It is possible'^ show the influence of variables wtGch war* aot 



included in Che Round R^^i" analysis by considering the overall preference 



<^ 

 distribution of the foul teil) cigarettes (the total number of tiaes 



that each cigarette wes prafer('«d when coopared wich the ocher three 



cigaretces in turn, expressed •■ a 'percentage of che Cotal nuaber of 



tests). The preference discribucioo ia^pfyivn in Table S, which for 



comparison includes two of the variablea fr<>4 che Round Robin analysis. 



Coopared wich the tocal sample, staokers wt(> a high degree of 



concern for healch (score 25-35) have a slight cendervcy to prefer A 



and B (lower nicotine) ac Che expense of C and D (hithe/ nicocine). * 



This cendency is also seen among chose aged 60 or over, chos^ belongiag ^ 



Co social class AB and those smoking fewer chan 10 cigarettes pi<4e day; 



9> 



