509 



'^^ 



-29- 



Proa t^ti tpoatuieoua coancnCt ic would app«ar th«c eltarttt* t 

 vtt |ec«railT (Ma as •troctt«r aod aort •acitfyias, ei|artcc« t vat ««c 



a* mildar, •llg^i|^aoocb«r and rachar fate buraios. Coamtatt that t 

 vat liked became of it« ttreagcb and tacisfaccioo vcr« Co b« found at 



all Itvelt of Inn^-^eed, %ac inrcicularljr aaoos Clutters 4 and 10. 



Coaments that E was I'ixfd bacause it was aild caae particularly froa 



the Low Need clusters. /. 



/. Test Cigarettes - Pref*reil<;^e 



y ^ 

 -■'. Consumers were asked two queations ralating to preference: "Which 



"^ the two cigarettes did you preferrr (referred to below as "free 



choTo^" - i.e. they were allowed to say ctjat cHey had no preference) 



Ay w> 



and "If "Qau had to choose one which would you ^efeirt" (referred to as 

 "forcad chofca^'). Table 12 shows the answers to"^:fe>th qnettions, by 

 cluster. la ehi^ree choice" situation the pcrceDCn»-^^ra£arance hat 

 been calculated on ru^basii of those who expressed a prej^renca one 

 way or the other, ignoriRj chose who had no preference. 



PERCESTAa 





TABLE 12 



RRIVC E OR F BY CLUSTER 



^^ 



a 

 c 



c 



N 

 I. 



b 



O 



