186 The Marketing Book
makes different assumptions regarding the way
in which the numbers reflect the situation
under measurement.
Nominal scalesassign numbers to objects,
variables or people to show that they belong to
some stipulated category, categories which are
mutually exhaustive and mutually exclusive.
In this scale, numbers have no mathemat-
ical value, they merely show that the people,
objects etc. belong to a nominated group. Thus,
people who read The Guardian might be
assigned to the value 7 and those who read The
Timesto the value 456. The only mathematical
function which can be undertaken is to count
the number of objects inside each category. Bus
numbers, bank accounts and football team shirt
numbers are all nominal scales.
Ordinal scales rank order objects/people
etc. according to the amount of a property
which it/they possess. But respondents in a
research programme must be able to discrim-
inate between items of interest with respect to
an attribute, i.e. they must have the ability to
say that this tea, for example, tastes better
than that tea. They are saying that this tea, the
preferred one, has more of the attribute ‘good
taste’ than the other does – their second
choice. Ordinal scales do not enable research-
ers to know/infer by how much one item is
preferred over the others in the same category.
Thus, it is not possible to say if the difference
between the first and second and between the
second and third is the same, more, or less.
Students’ examination results, first place, sec-
ond place etc., is an example of an ordinal
scale.
Interval scalespossess order and distance,
but not a unique origin, i.e. their zero point is
arbitrary. Thus, meaningful statements about
the distance between two objects on a scale may
be made. It is permissible to say that the
difference between scale points 7 and 8 is the
same as the difference between scale points 57
and 58. However, interval scales do not allow
researchers to make meaningful statements
about the value of a scale point being a multiple
of another value on the same scale.
Ratio scalespossess order, distance and a
unique origin indicated by zero. All mathemat-
ical operations are allowed here, so it can be
said that a reading of 80 on a scale is four times
a reading of 20 on the same scale. Three feet is
three times larger than 1 foot and 10 pounds is
twice as large as 5 pounds. Measures such as
height, weight and volume are examples of
ratio scales.Attitudes and their measurement
Marketing research constantly seeks to measure
respondents’ attitudes towards, for example, a
change in packaging, price, a new product,
politicians etc. But attitude measurement can
sometimes be a rather difficult concept for such
a practical subject as marketing research to
come to terms with. How have attitudes been
defined?
Two of the most useful and illuminating
definitions are as follows:... an individual’s enduring perceptual, knowl-
edge-based, evaluative and action-oriented
processes with respect to an object or
phenomena.
(Kinnear and Taylor, 1996)
... mental states used by individuals to struc-
ture the way they perceive their environment
and guide the way they respond to it.
(Aakeret al., 2001)
While there are many definitions other than the
above, there is broad agreement that an attitude
is a learned mental state of readiness, a way in
which individuals construct their own worlds
such that when confronted with a certain
stimulus they act in a certain manner.
Attitudes are not held to be the only cause
of human behaviour, there are many other
factors having an impact upon the individual at
the moment at which the behaviour under