473 



-31- 



tetpondeat oa « scale raaging fron 7 (aialnja concern) to 35 (maxiauv 

 concern). Th« sooret for the toc«l taaple oo S of Che 7 quesdont were 

 ooc slgnificuitly different fros 3.0, the aid-polat on the 1-S scale, 

 indicating that Che ntober of respondents expressing an opinion in one 

 direction was roughly balanced by the number expressing an opinion in 

 the opposite direction.'- however in the remaining two questions the 

 answers were very unevenly split and the score differed significantly 

 from 3.0: in Question 18a only 17Z of the sample had bought a brand 

 because of its favourable position in the Government list, and in • 

 Question 20c 83Z of the sample agreed either strongly or mildly that 

 tipped cigarettes are healthier than non-tipped_ cigarettes. After 

 sttidying tfeft resulting distribution of scores for Che 7 questions, 

 respondeotswei^^classified in one of three ranges: -'hl9 (low concern) 

 20-24 (mediua conotni), 25-35 (high concern). Table 10 Sbows the percentage 

 of respondents in eacn'M these three categories broken down by cluster, 

 age, social class, cigarette- consumption and nicotine delivery of 

 regular brand. 'y 



Based on the percentages in t>v^2S-35 bracket, those clusters which 

 tend to be the most concerned about tfae'^ealth effects of smoking are 

 Numbers 9, 6 and 12, in other words on* Low Need, one Medium Need and 

 one High Need cluster. Those which tend to b^ ^east concerned are 

 Numbers 7 and 1, both of which are Low Need clusteT». On this evidence 

 it would appear that no relationship between health concern and Inner 



08 



Need has been demonstrated. This is contrary to what was expected, ^ 

 though views differed as to whether those most concerned about h»«ltl 

 risks would logically be expected to be low in Inner Need or high in 



