Smoking and Tobacco Control Monograph No. 7 
DR. FREEMAN: Dr. Hughes? 
DR. HUGHES: I would like to ask your opinion about ordinal vs. cardinal 
scales because I think that was not made clear yesterday. You compared 
the FTC method to the ERA gas mileage. If I buy a car that has 38 miles per 
gallon and my sister buys a car with 19, 1 get twice what she gets. Now, 
even your own data show that is not true with tar and nicotine yields. 
When you have a tar yield that is twice another cigarette, you do not get 
twice the tar. So, I find those numbers misleading. I think the normal 
consumer when they see a cigarette that says, "1 milligram tar," and they 
see another cigarette that says, "8," they think they are getting one-eighth 
the tar, and that is not true. It seems to me if that is the case, and all you 
want is rankings, that we should do away with the numbers because they 
are misleading, and I would like to hear your thoughts about that. 
DR. TOWNSEND: A relative ranking of cigarette yields is what is essential 
in the marketplace. To date we believe that the FTC method provides useful 
information for the consumer. Do you really believe that your car gets 
19 miles per gallon when you drive it? 
DR. HUGHES: I believe that my car that gets 19 gets half the mileage of 
somebody else's car that gets 38, and I think most consumers would believe 
that if they saw the numbers 19 and 38. 
DR. DEBETHIZY: But what else is on that label? The other part that is on 
there is, "Your actual mileage may vary," and that is important because again 
this particular method was not set up to predict what an individual will get. 
It was set up for relative ranking, and I think it is really important to stick 
with that. 
DR. HUGHES: I agree, and with relative ranking, when you have rankings 
and ordinal categories, you do not have numerals attached. 
DR. DEBETHIZY: You just gave a good example of that when you gave ERA 
gas mileage. 
DR. HUGHES: ERA gas mileage is a cardinal system. It is not a relative 
ranking. The ERA gas mileage, 38, cars that have 38 miles per gallon do, 
in fact, get twice the mileage as cars of 19. 
DR. DEBE3 HIZY: Only if driven under standard conditions. 
DR. HUGHES: No. You are confusing variability around the mean with 
ordinal vs. ranking. 
DR. DEBE'FHIZY: I think I understand the difference, and I think that we 
could argue about this all day, but I think that the I'FC method was intended 
as a machine-based standardized method to provide relative ranking, no 
more than that. 
DR. FREEMAN: Dr. Henningfield? 
