Tobacco Documents Online

i c I ab ~ wi C'b N-10 A Psychophysical study on the effectiveness of cigarette filters E.P. lCoster Psychoioqical Laboratory Utrecht University 351'1' SZ Utrecht The Netherlands

1' 1. Introduction In the cigarette indiustry, increasingi efforts have been made over the past decade to reduce the amounts of tar delivered by cigarettess to the mouth. In order to reach this goal, many different types of filters have be.en, developed. When mere: retention of the tar in the filter was. not judged effective enough to further reduce the~ tar content of the delivered smoke, ventilated filters were introducedd which allowed the smoke to be dilutedi by pure air drawn in through perforations in the filter. in an effort to compensate for the loss of taste introduced, by the higher ventilation a new type of filter has been introduced. This filter, the Actron Filter, is especially designed to spread the smoke effectively through the mouth and thus to ensure that as, many, taste receptors are hit simultaneously as possible. This leads to spatial summation, the phenomenon that the same, amount of stimulus energy, when spread simaltaneously over more receptors, wi'll cause a stronger sensation than when it is concentrated on a small number of receptors. Earlier test& by Caiin (1) and Solms (2) have already demonstrated that this filter is indeed very effective. In this study the, fllter' was again tested. In order to do so,_a number of specially composed filter andi tobacco rod, combinations, so-called hybrids were made for comparison with commercially available cigarettes. Thus, it is possible to single out the contribution of the filter and' of the tobacco rodi to the strength of the taste. Also, special attention was given to the ease with which the tobacco could be drawn from ci~garettes equipped' with different types of filters. This latter' point is of particular importance since it is claime& by some that the''greaR:er effectiveness of the Barclay cigarette is diue to the fact that through human manipulation by the smoker one or more of' the four separate ventilation channels which are characteristic of the Actron filter, are blocked. If such blocking took place it would not on1y, have an effect on the strength of'the cigarette.but it would'al''so discriminate the ease of draw, since it reduces the ventilation and .a thus forces the smoker to draw against more resistance (the filter + the tobacco rod). Therefore, if the greater strength of a cigarette equipped with the. Actron filter is due to blocking of the ventilation channels that cigarette s!houli7 be difficult to dra!w upon.

2' . Method 2.1~ Cigarette& Fi've brands of commercially available cigarettes and eight tiybrids composed of' different cigarette filters and, tobacco rodis were used, in the experiment. All cigarettes were presented wiith their trade names covered by a special white cover carrying, a three digit code. Table I gives a survey of the cigarettes used. Table 1. Cigarettes used in the experiment A. Co®erciali Brands Nam 1 Barclay Type of' Filter Actron Tar 1 mg Nic 0.2 2 St llorits Special Filter 2 mg 0.2 3 Peter Stuyvesant Extra Mild Conwentional S' mg 064 4 Peter Stuyvesant Ultra Mild Conventional 1 mg 0.1 S Philip liorris Ultra Lights Conwentional 1'mq 0~.1 B. Hybrids Filter Tobacco-Rod 6 St Morits Barclay 7 Peter Stuyvesant Extra Mild Barclay 8 Peter Stuyvesant Ultra Mild Barclay 91 Philip Morris Ultra Lights Barclay 101Actron St Moritz 11 Actron; Peter Stuyvesant Extra Mild 12' Actrom Peter Stuyvesant Ultra Mi!1'd 13 Actron Philip hlorris Ultra Lights

2.2. Subgects Forty two subjects (20, Barclay smokers and 24 smokers of thee other cigarette brands used) took part in the experiment. The subjectss were reeruted by' an independent marketing company (Interview, Amsterdam)i and were paid for their partic'ipation. The subjects took part during three two-hour sessions. The first two days they smoked nine cigarettes in each session and! the last day eight cigarettes. Each of the thirteen different cigprettes was thus smoked twice by each subject. The order of the cigarettes was randomized. Different randomisations were used: for groups of'the subjects. 2.3. Procedure At the beginning of'the first session the subjects were placed at tables containing an instruction sheet, a response s!heet, an ashtray and! a lighter. They were asked to read, the instruction and to ask questions if they, felt that further explanation was*'needed. The instruction told them that during the experiment of that day nine cigarettes would be saooked'.and that they were asked to judge thee stFenqth- of the cigarette first an& there after to judge the ease of draw of the cigarette. The scale used! in the.se experiments was originally developed by Dr. W.S. Cain of Yale University. It consists of a 2'3' cp long line with the dutch word for weaker on the left end! of the scale and'. the dl4tch word for stronger on the right end (see Appendix I1). The subjects were asked to compare the cigarette with their own normal brand and to indicate whether they considered the experimental cigarette to be stronger or weaker than their own brand.. in the scale for ease of draw these words were replaced by the words for much more difficult and much easier respectively. 2.4. Scoring and data treatment The marks on the scales were measured for their' distance in mm to the 'own brand' mark. Distances to the left were given a negativ sign and' distances to the right a positive one. Averages and standard'

4 deviations per ciyarette were calculated for the total group of .ubjects and for both the Barclay surokers and the smokers of other low tar cigarettes separately. The standard deviations were high because each subject used' a scale based on comparison with his or her own brand. Therefore non parametric statistics ('wilcoxon mtched pair s'igned' rank test) was used to determine the significance of the di'fferences between the different ciqarettes.

3. Results 3. 1. ¢ompari~son of the commerci'ally available brands The asin results for all smokers are given in Tab1e 2 and in Table 3 and 4' for Barcl'ay smokers respectivefiy,, and smokers of other brands Table 2. Means and standard deviations of' all smokers on two repetitions (ni88), Number Filter 01i barclay 02 st moritz 03 p s extra 04 p s ultra 05 p a ultra 06 at moritz 07 p s extra. 08 p s ultra 09 p sr ultra 10 barclay 11 barclay 12 barcl'ay 13 barclay Tobacco rod Mean StD. barclay +26.4' 45.5 sti moritz -1i8.1! 50.2' p s extra -1'S'.6 45.3' p s ultra -56.0 49.8 p a ultra -57.7 47.9 barclay +24.3 4I5.9 barclay + 1'8. 9 53.2 barclay -40.4 6'1I.6' barci'ay -31.9 57.7 at moritz + 4.1 4'5'.3 p s extra 0.0 47.9 p s ultra + 7.8 45.9 p a ultra +26. 0 41.0 ltean StD, -18.4 45.6 + 9.3 47.3 +11.0 38.1 +38.9 54.0' +29.4 54.4 - 0.3 43'.81 - 3.0 S0. 6 +31.3 59.3 +32.9 51!.5 + 4.1 43.7' - 6.2 4'3.9 - 6.2 410.7 -1!5» 7 43.0 In the first five rows of Table 2 the results are given for the five different commercial'~ly available cigarettes. Two conclusions can be dlrarrn immediately from the comparison of' these resus'ts: Barclay, is the only cigarette that is )ludged to be stronger in taste than the brand the smokers are used to and it is also the only brand that is easier to draw on than the own brand of the smokers. The differences between Barcl!ay and the other brands in both respects are all statistically sigrificant at the p <.001 1'eveli. .P Strenqth DraN.

Table 3. Means and standard deviations of'Sarciay smokers (n-4i0-) Number Filter Tobaeco rod 01 02 03 04 05' 06' 07 08 ~~ 09 10 11 12 13 barclay barclay st .oritz at s+oritz p s extra p s extra p a ultra p a ultra p a ultra p a ultra at' aoritz barclay p ai extra barclay p s uLtra barclay p a ultra barclay barclay at mwritz barclay p s extra barclay p s ultra barclay p a ultra Strength Draw Mean StD. Mean. StD. +16'.8 34'1. 0 - 2• 3' 37.5 -42. 0 49.1 + 17. 6 45'. 1 -33.2 36.9 +19.5 32.7' -67.0 53.0 +44.5 59.2 -76.0 34.6 +28'.4 57.4 + 1'5. 1 46.5 ' + 0.9 39.2 - 7.2 56.8 + 4.2 41.4 -53'.8 59.6 +34.9 63.7 -51.0 58.8 +42.2' 55.7' - 6.6 37.6 + 9.6 34.1 -22.2 47.3 + 2.6 34.8 - 6'.6' 4'T.8 + 7'.1 30.2 + 7.5 35» 3 - 2.2 22.6 Table 4. Means andi standard deviations of all other smokers (n-48) '. Number Filter Tobacco rod 01 barclay barclay 02 at aoritz~ at aori'tz 03 p a extra p a extra 04 p a ultra ps ultra, 05 p m ultra p a ultra 06 st moritz barclay 07 p a extra: barclay 08 p s:ultra barclay 09 p a ultra barclay 10 barcla~ at moritz 1 1 barclay p r extra 12 barclay p s ultra 13 barclay p a ultra Strength Draw Miean StD. Mean StD, +34..4 52.2 -31.9 47.7 + 1.8 42.2 + 2.6 48.5 - 065 4i6.7 + 3.7 4'1.1 -46.7 415.3 +3!4,. 1 49.3 -40.0 51.8 +30.3 48.3 +31.7 44.5 - 1.3 47.5 +410.6 38.6 -10.0 49.6 -29.3 61.6 •28.J 55.8 -16.0 52.2 +2'S.1 4'7'.0 .12-8 49.4 - 0.6: 50.3 + 18.6 40. 2: - 131. 6 49.5 *19.7 401.9 -17.2 45.0 •t1.3i 39. 3 -27.0 52.11

7 If one studies the first f'ive rows of Tables 31 and 4 one sees that the same conclusions can be drawn from the~results of the Barclay smokers and of the smokers ofl other cigarettes respectively. Furthermore, it cam be seen that the Barclay smokers are used' to smoke a stronger tasting cigarette than the smokers of other cigarettes. Their judgments on the strength of all the cigarettes tested are l~ower than the judgements of the smokers of other, ccigarettes. With regard to ease of draw a similar conclusion can be drawn: The Barclay smokers are used to a cigarette with.a relatively easy draw. As a result of t'his, they tend' to rate the draw of the experimental cigarettes (v3'th the exception of one) as being more difficult than do the smokers of other brandis. These results are illustrated in Figure 1'1 and Figure 2' for strength and!ease of draw respectively. Fig. 1. A comparison of' the~ average STREtCTIW FIVE~~ BRAM0'S' .1I01 judgements on strengthi of taste for the five < a M ~. t o:' Stm a Q mng commercially available < «- Cr x ..' J 0 44J ~ - cigarettes. ~ F w N = m %A a a 0. -t0i DA'!t Smokers 0 ®arcdar Smokers m Other &nnd Smokers Weak ~

8 EASE OF ORW FIVE eRhNCS .aon Figr 2. A comparison of' the average judgements on ease of' draw for the five commercially available cigarettes ~ a v a 0 Lfl a -C DJCItuIK. ~ ~ ?I I.- , vn~. V1 Zr %A m C7A(1 Smokers Morclo'y Smokers E30ther BRo'rb Smokers 3.2. Comparison of the hybrids cmntaining, the Barclay tobacco rod.. The contribution of the Barclay tobacco rod to the better performance of Barclay in, comparison to other commerciall!y available cigarettes with respect to strength of taste and ease of draw can be estimated by comparing hybrids with different filters but with the same Barclay tobacco rod, with cigarettes in which these fil'ters are combined with the original rods (the commercially available cigarettes). The results which make this comparison possible are taken from rows t to 9 of Table 1'. These results are illustrated in Figures 3 and 4!. Easy

9 I STiRENGTH' HY8'RIIiQS Fig. 3. A coa-parison between the judgements' on s!trength for the hybrift containing, the Sarclay rod and for the original' cigarettes .4 ~ x %Wi Strong O'8crclcy Rod + ElOcigirnol Rod We©k As can be seen fros this figure,, putting a Barclay tobacco rod on the filters of the other coam-ercially available cigarettes improves the strength of the taste in all cases. All of the differences obtained are slgnificant at at least the p<.01 level.

10 Fig. 4. A comparison between, the: judgements on ease of diraN for the hybrids containing the Barclay rod and' for the original cig¢rettes. EASE OF ORtw HYBRIDS ObarcbnY Rod tai0ngino1 Rod 0 OIVicUl" Ecsr Where ease of draw is concerned, putting a Barclay tobacco rod on the other filters also seems to improve the draw in all cases except one (Philip Morris Ultra Lights). llbne of the effects tound are statistically signifi'cant, however. The only significant differences found are between Barclay and all, the other cigarettes and, between the two ultra lights and the other two non Barclay brands. The better performance: of Barclay on ease of draw seems to be almost completely due to the filters. 3.3. Comparison of the hybrids containing the Actron Filters The contribution of the Actron Filter to the performaiace of the Barclay cigarette with regard to strength and'ease of draw can be estimated by comparing the results of' hybrids made with the Actron .P

ll1 Filter but with different tobacco rodic with the rods equipped with their original filter. I These comparisons have been made in Figure 5 for strength Figure 6 for ease of draw. STRENGTH HyBRID'S •10 Fig. S A comparison between the j ~ _ ~ average j!udgements on ~ v L strength for the hybrids I m N containing the Actron filter and for the original cigarettes re'~s!u:lts of~ the salOe~ and' in Strong a a' Cr ~ M ~ ~ J x W D %n „n = i a m. d a -110J [a OR'1GlNAL FItT'EiR'. Inspection of this figure shows immediately that replacement of the original filters by the Actron Fillter lea'ds to improvement of the strength of the cigarette taste. In all cases the differences between the cigarettes equipped with their ori!ginall filters and those equipped with the Actron Filter were significant at the p c.001 level. One hybridl, the Philip Morris Ultra Lights rod equipped with the Actron Filter, equalled Barclay in strength. The other hybrids, although greatly improved! in strength by the adidition of' the Actron Filter remained weaker than Barclay (i'n all cases at least p<.01),.

12 Fig. 6. X compa'rison between the ave'raqs jludgements on ease of drawn for the hybr'idr containing the Actron filter and for the original cigarettes EASE ' Of DRAW wYB'RiD'S .10r -10 AE T P 0N F'ILT E'R E2 ORIGiUMAL FILTER DO flcult Emr Fig, 6 shows that replacement of' the original filters by Actron Filters results in an improvement ofl the ease of' draw ia all four cigarettes. In the case of St Moritz this change is statistically non-significant. In the other three cases the level of significance is at least p < .01 and for the highly ventilated cigarettes even p <.0001. In one of the hybrids (Philip Morris Ultra Light rod' with Actron Filt,er)' the ease of draw is almost equal to that of Barclay. In all other cases Barclay is significantly easier to draw on than the

13 hybrids equipped vith Actzon (at least p<' .0'S'). This gives some support to the id+ea, that the Barclay, Tobacco rod is in part . responsible for the ease of draw of Barrlay. However, the role of the filter is much moze substantiaS, as was already hinted at in the pravious section (discussion to: Figure 4).

!. Conclusions and Discussion From the experiment. the following main conclusions cam be drawnis 1. Barclay is judged to be stronger in taste than any of' the other. commerci'al'1y available cigarettes in this study, including Peter Stuyvesant extra mild, a 5 mg tar cigarette. This statement is further corroborated by the differences in judgements between the Barclay smokers and those who smoke other brands. 2. Barclay is also.jOdlged to be easier to draw on than the other commercially available cigarettes in this!study. 3. The Barclay tobacco rod has a strongar taste than the tobacco rods of any of the other commercially avai~lable brands used:. 4. The: Barclay tobacco rod is only a little easier to draw upon than the tobacco rods of the other commercially available brands use& 5. The Actron Filter is far more effective in increasing the strength of the taste than any of the filters vith which the other commercially availab,l'e,brands used in this study are eqaipped» 6. Z'h llctron Filter provides a much better ease of draw than the other highly ventilate& filters with which the t mg tar cigarettes used'in this study vere equipped. The.l'ast of'these conclusions makes it very clear that the very large increases in strength obtained by combining the tobacco rods of' the other 1 mg cigarettes with the Actron filter are not due to blocking of the channels by the smoker. The fact that the standard deviations obtained for the strength of the cigarettes equipped with the Actron filter are in the normal, to loww range corroborates this concl'usion. IY' huemn manipulation with the channels was involved, it would, be highly unlikely that all human subjiects mnlpulate6 to the same extent. Ifl this wasi indeed not the case, one would' expect moree variability (higher standard~ deviation) in the results of the Actron, filter equipped cigarettes. This i's certainly not the case. .a

xpPFNOrx I Response foru cigarette nu,enber:........ subject nuniber:.......... date:........... .1 find that th,i's cigarette tastes: your own weaker brand stronoez I I ( than my owm brand. I find that this cigarette draws: your own . easier brand more difficult I II 11 than my own brandl