56 



2. Analyses of the Trend of Tar Level on Year: 



The tar data provided by the FTC was parallel to that for 

 the nicotine, consisting of two additional vectors: 1) Tar(i,j) = 

 mean tar level of cigarettes tested in year i, tar category j; 

 and 2) SDT(i,j) = std. dev. of tar in year i, tar category j. In 

 this data, the mean and the std. dev. were computed, according to 

 Joe Mulholla'nd by taking sales weighted means and sd's of all of 

 the varieties tested within a given year and tar category. 



We did four analyses of the trend of Tar Level on Year, from 

 1982 through 1991. These were as follows: 



1) weighted least squares regression of tar, Tar(i,j) on 

 year i, stratified by tar category j = (high, low, and 

 ultra-low) . The weights were proportional to reciprocals of 

 the variances of the sample means * l/VarT(i,j) > 

 N(i,j)/SDT(i,j)-2. 



2) weighted least squares regression of average tar on year, 

 ignoring tar category. In this regression, the dependent 

 variable was the sales weighted average of tar of all 

 varieties tested in year i. This was computed from the FTC 

 data as Tar(i) ■ S{S(i, j) *Tar(i, j) ) /£S(i,j) where the sums 

 were over j > ultra-low to high. The weights in this 

 regression were again proportional to reciprocals of 

 variances of the Tar(i) . Specifically, the weights were 

 w(i) « 2:{S(i,j)-2) / j:{VarT(i,j)*s(i,j)-2) » 

 S{S(i,j)-2) / Z{S(i,j)-2 * SDT(i,j)-2 / N(i,j)). In all 

 cases, the sums are over j^ultra to high. 



3) Spearman rank correlation of tar and year, stratified by 

 tar category. 



4) Spearman rank correlation of average tar and year, 

 ignoring tar category. (Average tar is defined exactly as 

 in analysis 2.) 



Findings were that tar was statistically indistinguishable 

 from being flat within tar categories. This may be because 

 veurieties within a tar category retain a deliberately maintained 

 chemical identity, with lowering of tar in the marketplace 

 occurring by proliferation and increased sales of lower tar 

 brands. It may also be an artifact, since the association found 



