Tobacco Documents Online

-6- QUZSTIO~ 5 "How do you think this blocking would be brought about?" DEPENDING ON THE WAY THE CIGARETTE ZS HELD, AS A CONSEQUENCE OF THE NORMAL HOLD ON THE CZGARETTE DON'T KNOW 712 292 (N - 41) qUESTION 6 "About what percentage of the t/me you spend puffing on the cigarette would you think so~e of those holes could be blocked - O, 25, 50, 75 or 100 ?" DON'T KNOW " 12.5% 0 - 0% 25 - 15% 5O - 42.52 75 = 20% I00 = 10% ) ) 72.5% ) (. = 40) It will be noticed t.hat this post interview data comes from a smaller sample size. This represents a constraint of the research desIEn since it was necessary to avoid direct reference to clgarettes in the filmed pre-lntervlew (set up ostensibly as a llfe stress interview) it became impracticable in some cases to consider turning to a very specific focus on cigarettes in the post interview as this may have incurred a loss of faith with the subject. CONCLUSIONS AND RECOMMENDATIONS It is clear that the systematic interview data dealing with subjects* own perceptions/est/mations of partial blocking behaviour is quite in accord with the gozlowskl findings which insinuate a h~h level of partial blocking. It is, however, equally clear that actual observed behaviour collected from the same subjects shows no such high incidence of uart/al blocking. CY~ C~ O~ t.m ClibPDF - wvvw.tastio.com

-7- ~hac subjects do and vhac they chink they do are almost diametrically opposite, and we muse conclude chac che use of Kozlc~sk£ data co make inferences about accuai puffing behaviour is demonstrably Invalid. Recommended supplementary analyses are as follows: Io o e 4, Comparisons by sex, age, socio-economic class and smoker status (degree of commitment Co low car smoking). Examinacic~ of the hFpochesis that increased incidence of cermlnal puff parclal blocking is related Co (shorter) bucc length. Examination of the burrs collected from SINOS subjects (for which we know the extent of partial blocking) for (a) the possibility of predicting parciai blocking indicence from end stain appearance and, (b) chemical analysis. Smoking machine experiments designed co stmulace the CTpe and levels of blocking observed, and therefore indicate the order of delivery increases implied. This would be valuable since Kozlowski has asserted possible increases from Img to 23mE on ~he basis of his (invalid) interview data. R.P. FERRIS O CC Cr~ Cr~ ClibPDF - www.fastio.com

PROJECT S'rNOS : ADDENDUM It has been hypothesised that partial blocking Incidence on the terminal puff may be related to butt length, such that a shorter butt length may require more support in the month during puffing and would therefore be more likely to be flnssr blocked. This hypothesis could be examined within SINOS since each interviewee's butts were collected along with the observational data. Results show that: For those smokers who partially blocked the terminal puff, average butt length = 2.Smm. For chose smokers who did not partially block the termlnal puff, average butt length - 4.7mm. R.P. FEREIS 0 C", CO "-M F',,.,) CT', ..,..j C li b P D F - vvvv v,y~ f~a:s~t~ireq.Tc 6tlf~. 7. ~3

PROJECT SINOS : ADDENDUM IC has been hypochesised thac partial blocking incidence on che terminal puff may be related Co buCC length, such ChaC a shorter bucc Iength may require more support in the mouth during puffing and would therefore be more likely to be finger blocked. This hypothesis could be examined within SINOS since each Interviewee's butts were collected along with the obaervatlonal data. Results show that: For those smokers ~.-ho partially blocked the temlnal puff, average butt length - 2.Smm. For chose smokers who did aoc partially block the terminal puff, average butt length - 4.7mm. R.P. FERRIS eJPF/W~L/! 9.7. s3 o co o-,, co ClibPDF - www.fastio.com

PROJECT SINOS : ADDENDUM IC has been h~pochesised chat partial blockins incidence on the terminal puff may be related co buCC Iengch, such chac a shorter buCc lengCh may require more support in the mouth during puffing and would therefore be more likely Co be finger blocked. This hypoches£s could be examined within SINOS since each Interviewee's butts ~ere collected alonE with the obeervaCional data. Results aho~ Chat: For Chose smokers who partially blocked che terminal puff, averase butt length - 2.hmm. For those smokers who did noc partially block the terminal puff, average butt length - 4.Tmm. LP. FERRIS IPF/~,'3L/19.7.83 CO ClibPDF - www.fastio.com

R.P. FERRIS/~L 15oh July 1983 PROJECT SINOS: USE OF SYSTEMATIC OBSERVATIONAL AND INTERVII~DATA TO EVALUATE INCXDENCZ OF PARTIAL BLOCKING OF VENTILATED LOW DELIVERY CIGARETTES OBJECTIVE To examine critically the KozlowsEt assertion that: N ... we have found in systematic interviews chac 32X co 69Z of low car smokers have blocked the holes rich fingers, lips or cape." KACKGROUND GR&DC maincaln Chac the Xozlowskl asserclon is mlsleadlng since Ic is based upon smokers' opinlons as to whether they par~lally block or uoc, rather than syscematlc observaclon of smokers' behavlour during actual puffing. A pilot study using filmed ineervlews of low tar smokers during acClve smoking ~ndicaCed ChaC for 20 subjects 89Z of puffs were unblocked, 8Z were parcially blocked and 3Z were undefinable. IC was decided Co conduct a large scale observational study using academic referees and accounting for possible regional differences in partial blocking. CUItEENT REPORT The large scale study (SXNOS) was successfully completed with 113 subjects distributed evenly across a Notch (Sheffield) and South (Bristol) location. ClibP'BF - www.fas.t-i.o.com m P~ C~

-2- The structure of eke final sample achieved :Ls showu below: SEX AGE S.E.C. TOTAL 16-34 35-54 55 + ABC1 C2DE BRISTOL 21 24 45 22 15 8 22 23 15 53 68 29 31 8 31 37 TOTAL 36 77 113 51 46 16 53 60 SMom "EXCLUSIVE" LOW TAE SMOKERS (smoked low car only in last year) "LESS COMMITTED" LOW TAR SMOKERS (low Car most often, buc middle tar also used) 78Z 22Z The video films have been analysed and prelL~atnary flndLngs reflect pLloc scudy expeecacions: Of coCal puffs (ex. lighctn8 puff): URBLOCKED - 80Z PAB.TT.ALLY BLOCKED - 13% INDETERI~ItA~ - 7Z V,,D O", CO .,-,,.3 ---..,I ClibPDF - vvvvvv.fastio.com

-3- NGET~T~ LOCATION UNBLOCKZD - 8 IX PARTIALLY BLOCKED - 10Z ZNDETERMII~TE - 9Z SOUTHERN LOCATION U~LOCKED - 78Z PAETIALLY BLOCKED - I~'~ INDETERMINATE - 5Z The dom£nanC proportion of unblocked pufflng observed in the pilot study has been substanClated. There is some indication chat partial blockin8 incidence =ay be higher for the Southern sample. This iuterpretatioa should be viewed with caution since it may reflect an artefact of video quality (see percentages of indeterminate puffs). This is some evidence to support the view that the terminal puff is =ore likely to be partially blocked (th£s is part£cularly so for the Southern sample): UNBLOCKED - 64Z PAETIALLY BLOCKED - 26Z INDETERMINATE - IOZ NOITHERN LOCATZON UNBLOCKED - 71Z PARTIALLY BLOCKED - 22~ INDETEE~ATE - 7Z 0 co ClibPDF - wwwifastio.com

-4- $O~ 50CEION UnLOC~D - 5 i% PAATT~LLY BLOCKED - 36% INDETE~INATE - 13% / Distribution of partial block/ng within the puffins patterns of the sample population supports the view chac £C represents a scattered inadvertent £ucidence: Excluding lighc£ng puff (average puff number per clsarecce equalling 6) :- for only 9 subjects do partially blocked puffs exceed 2 per cigarette (8% of sample) for only 6 subjects do partially blocked puffs exceed 3 per cigarette (SZ of sample) for only 4 subjects do partially blocked puffs exceed 4 per cigarette (3.5Z of sample) Excluding l~hC~nE puffs and indecerm£naCe puffs, 72 subjects (64Z of sample) show no evidence of partial blocking whatever, whilst only 7-subjects (6Z of the sample) parcially block all puffs. Further comparisons Co be made ~chln the observational data include by sex, age, social class, mnoker status (coan~tted or less committed low tar smoker). The hypothesis thac inc£dence of partial block£ng on the terminal puff may be related To butt length (shorter butts being associated with more partial blocking) is also to be tested since butts were collected for each subject. SYSTEMATIC INT~VIDg DATA To clarify the relationship between Chts observational data and systematic interview data of the type used by Kozlovskt, a post observation session inter- view was admtn£eteced To establish subjects' own perceptions mud esctmacious of their puffing behaviour. The inCerviev was designed To collect the Type of Information vhtch forms the basis for ~ozlowski's asserC£ous: m ClibPDF - wwvv.fastio.com

-5- qUESTZON I "Many low Car cisarectes, including the one you have Just smoked, have several small holes in the tip (demonstrate) ... have you ever noticed Chase before?" YES ', 62I H0 " 38Z (N ', 97) QUESTTON 2 "~mt do you think the function of Chase small hoZas miEhC be?" TO STOP GETTI~ TOO MUCB NICOTTNE TO STOP GETTZNC TOO MUCH TAR VENTILATTON (TAR AND NICOTINE) RZDU~ ~40KE ~E AL~ INCREASE FILTRATION DON'T TAKE RACX SO MUCH/EEDOCE STRENGTR LET IMPUETTIES OUT THE CTGARETTE BUR~ STOPS YOU SMOKING MORE LESS DANGEROUS kLLOWS THE NICOTINE TO COMZ THOUGH HELPS TO TZLL YOU ETBAT TEE CIGAI~TTE ZS YOU MUST NOT S~01~Z PAST IT DOH'T KNO~ Z 28 14 11 li 9 7 3 2 1 1 t 1 1 1 9 (N - 100) qUESTZON 3 "Do you Chink durin8 normal smoking you would be i£kely Co block Chess hole8 aC all?" 'iCES - 52Z NO - 27Z DON'T KNOW - 21Z (N - 97) qUESTION 4 "Would this blocking be conscious or tnadverl:ant?" INADV~RTANT 8 7! CONSCIOUS 2Z DON'T KNOW 11Z O C~ .-.,.j. ClibPDF - wwwifastio.com