Day 256 - 04 Jun 96 - Page 56


     
     1        A.  Right.
     2
     3   Q.   This top mean is if the highest quartile is 35 per cent.
     4        Obviously there is no significant difference at the lower
     5        end, is there?
     6        A.  Well, ours was 6 to 24 and this one is 15 to ----
     7
     8   MR. JUSTICE BELL:  You move from 6 to 24 and the average was
     9        about ----
    10
    11   MR. RAMPTON:  I am sorry, my Lord, was your 6 a mean?
    12        A.  Yes, it was.
    13
    14   MR. JUSTICE BELL:  Was it a mean?
    15        A.  Yes it was a mean, surprisingly.
    16
    17   Q.   So that was a mean in the lowest community, as it were?
    18        A.  Right.
    19
    20   Q.   Your question, Mr. Rampton, before I interrupted was
    21        whether the differences were significant and what were you
    22        going to say about that, between these means and your
    23        means?
    24        A.  I thought the point of the question was to know whether
    25        or not these differences were really real or not and I am
    26        saying, if that is the question then my answer is yes, they
    27        are.
    28
    29   MR. RAMPTON:  His Lordship is right, perhaps it was not a very
    30        good question.  I meant to be specific.  These findings as
    31        to the strength of the association are similar to yours?
    32        A.  Yes, except for the fact that the ranges are rather
    33        different.
    34
    35   Q.   That was going to be my next question.  What difference
    36        does that make?
    37        A.  Well, that tends to suggest, and in fact it is good
    38        evidence to suggest, that over the whole range all the way
    39        from the 6 to the 35, if you will, just taking 2 sets of
    40        data, we get this sort of rather general slope and
    41        consistent relationship between dietary fat intake and
    42        breast cancer, although the relationship, when is examined
    43        just for fat, we get a significant relationship just for
    44        fat, but it also says that we are taking away from that
    45        relationship some of these adjustment factors that, in my
    46        view, ought not to be subtracted.
    47
    48   Q.   I see.  You would not disagree with these orders, that for
    49        their study group, or their study population, their figures
    50        are not statistically significance because that is what 
    51        they are saying, is it not? 
    52        A.  Yes.  They are saying though after adjustment.  They 
    53        are saying it is not statistically significant after
    54        adjustment for these factors, and my point is:  Why are
    55        they making these kind of adjustments? It comes back to
    56        this question concerning how is the reductionist nature of
    57        the hypothesis for investigations?  They are trying to just
    58        focus just on fat and turning around and taking away some
    59        of the things that that actually causes.
    60

Prev Next Index