123 
Now, your other question, I am sorry, I got a little carried away with 
my answer, and I don't remember your other question. 
DR. AHMED: Those were the two questions I asked. 
DR. FREDRICKSON: I think you have answered both of them. 
DR. AHMED: Yes, I think you have answered my questions. 
DR. FREDRICKSON: Professor Rosenblith. 
PROFESSOR ROSENBLITH: I have a fairly technical question. I have 
heard you refer repeatedly, and the Green Book does too, to a one-step reduc 
tion. That leads me to ask whether you consider levels of containment an 
ordinal scale or an interval scale. In an ordinal scale, all you know 
about is the direction of increase, but you don't know what the intervals 
are, and specifically the Green Book talks about the reduction from P4 to 
P3 as greater than one step. That makes even assumptions beyond what an 
interval scale is. 
I was impressed with the fact that you talked once about a reduction 
in P level, and once a reduction in EK level, and sort of set them equal as 
one step. I think that there is a serious problem here, that we are dealing 
indeed with an ordinal scale as I had assumed when I saw low, moderate, high 
and maximum, or whether we are dealing with something that is on an interval 
scale that claims a lot more properties than an ordinal scale does. 
DR. HELINSKI: I guess I used the term "step" because I think it is — 
we generally use it in committee — and I think it is a term that is also 
used in the Guidelines. I only used it to indicate to you where we are 
on the containment levels as far as physical and biological containment. 
That is, it is P3 to P2, or EK2 to EK1. I didn't attempt to quantify that 
in any way. Perhaps at some time the Committee may want to go into detail 
on that. 
PROFESSOR ROSENBLITH: Then I have serious questions whether under 
those circumstances one ought to use that kind of language in the Guide- 
lines, because other people will not be as wary of that as you and your 
Committee are, and might look for equivalences where there are none. 
DR. FREDRICKSON: Professor Rosenblith, you might give some thought 
to advising us better, or the Committee better on the language that might 
then be used . 
DR. NEEL: Just to follow up Professor Rosenblith, what I think I hear 
you saying is that we are working, perhaps, with nonparametric statistics. 
PROFESSOR ROSENBLITH: Well, that is another way of saying something, 
but there are certain statistics that can only be applied to interval scales 
and other statistics have to be applied to ordinal scales. 
[ 327 ] 
