Day 257 - 06 Jun 96 - Page 21

     
     1   Q.   Somewhere between 385,000 -- no, sorry, it would be
     2        somewhere between 425,000 and whatever double that would
     3        be.  A bit more than double than that, something like
     4        900,000?
     5        A.  Yes, the only caveat would be that on the first sheet
     6        you get down into numbers so small, because it is a from a
     7        small sample, that you get out of the bounds of statistical
     8        reliability.
     9
    10   MR. JUSTICE BELL:  I can see that because the figures are so
    11        small that if you took the next 2,100 of the 15,000 you
    12        might find you got very different sort of figures?
    13        A.  Yes, that is correct.
    14
    15   Q.   You might have, it is statistically quite possible to have,
    16        8 nearly every day in the first 2,100 and 18 nearly every
    17        day in the second and none in the third, and so on?
    18        A.  Yes, anything below 100 is normally taken to be
    19        unreliable so far as the sample goes.
    20
    MR. MORRIS:  From McDonald's own experience, when they looked at
    22        these figures they were satisfied that this was a fairly
    23        accurate picture?
    24        A.  Yes.
    25
    26   Q.   Of what they knew?
    27        A.  That is correct.
    28
    29   MS. STEEL:   Just going to your statement, if you get page 3.
    30        In paragraph 10 you originally said that the research is
    31        compiled by Taylor Nelson who conduct random telephone
    32        interviews to 60,000 people per year.  So are you saying
    33        that is now incorrect, it is done on a house to house
    34        basis?
    35        A.  That is correct.  Right.  I was mistaken in that point,
    36        yes.
    37
    38   Q.   Do you know which regions the interviews were done in?
    39        A.  Yes, it is on a representative basis across the UK to
    40        give us a representative sample of the UK population.
    41
    42   Q.   There might be some in -- I do not know -- the Outer
    43        Hebrides and Cornwall, and then there might be some in
    44        London?
    45        A.  Not quite as far flung as that, mainly where the main
    46        populations centres are.  But, yes, across all towns
    47        broadly above 50,000 population within the UK, that is
    48        broadly around the size of town they would be looking for.
    49        Quite simply because those are towns in which McDonald's
    50        and our competitors exist below 50,000 population, that is 
    51        the case. 
    52 
    53   Q.   All towns broadly above 50,000 population?
    54        A.  Yes.
    55
    56   Q.   I wanted to ask you about the questions that are in tab 5
    57        which relate to this.  On the second page, question one, it
    58        has 'person interviewed is...' and the third one down has
    59         'new occupier whole household year new'?
    60        A.  Can you clarify which document you are looking at?