Smoking and Tobacco Control Monograph No. 7 
I think many smokers, while they may not be as accurate in estimating 
FTC tar yields, still can rank cigarettes by tar. Now, how could they possibly 
do that if compensation were extensive? 
DR. KOZLOWSKl: I think you do not want to think of compensation as 
something that influences everybody's smoking behavior. What we found 
in the Pharmacology, Biochemistry and Behavior study, half of the people block 
vents quite a lot, and the other half did not block them at all. 
And if you looked further at those who did not block, you found that 
they did not smoke as many cigarettes per day. If you did taste ratings and 
how they liked the taste of the cigarette, they seemed to be consumers who 
were after a really low-yield smoke. They weren't blocking the holes. Not 
everybody smoking a low-yield cigarette blocks the vents. But this gets back 
to the issue of subject self-selection biases. We have to expect that there are 
individual differences in how much nicotine a person might want and also 
to the extent that a person is smoking for nicotine. 
So, half of those subjects who were smoking ultralow-yield cigarettes 
in the long term were not blocking vent holes; they did not smoke many 
cigarettes per day; and they had low CO levels. The other half smoked 
a lot more cigarettes a day, smoked earlier in the morning, and got higher 
nicotine levels. You average them, and you get the kind of figures that are 
commonly described as "intermediate." Some people were showing a lot 
of compensation; some were showing very little; and that figure of mean 
compensation can be misleading. 
DR. DEBETHIZY: 1 think you have pointed out an important fact: No 
machine-smoking method can predict individual behavior. This method 
was never intended to predict individual behavior, and it does not. I think 
people use different strategies when they smoke cigarettes, and it is rather 
obvious in the data you presented today. 
DR. COHEN: Is it your intuition that a great many people who compensate 
are just following classic learning theory and do not even know they are 
doing it? 
DR. KOZLOWSKl: Some people are not aware they are doing it; that is clear. 
They are not aware they are blocking the holes. I think that some people 
find the cigarettes relatively difficult to light. You push it a bit further in 
your mouth and it is a lot easier to light. Blocking could get started in a 
number of ways. 
DR. SHIFFMAN: We have all been struggling with the issue of variability 
within a given product or products of equal FTC yield, I think, in talking 
about compensation and in the difference between the machine yield and 
the human biological exposure. 
Now, with this issue of color matching, you are introducing something 
that I think has to do with true exposure rather than FFC yield. 1 wonder 
what you could tell us about the prospects of using a system like this to 
170 
