Marketing research 187
investigation becomes manifest. Attitudes,
though internal to the subject, are conditioned
through external experience, but experience is
not a random or arbitrary process, but one
which is organized through a process known as
learning.
Components of attitude
Attitudes have three components:
1 Cognitive– represents an individual’s awareness
and knowledge about an object, person etc.
They say, ‘I have heard about Brand X’ or ‘I
believe that Brand X will carry out this
function’.
2 Affective– represents an individual’s feelings
(good/bad etc.) towards an object etc. and is
usually expressed as a preference. They say, ‘I
do not like Brand C’ or ‘I like Brand D better
than Brand F’.
3 Behavioural– represents an individual’s
predisposition to action prior to the actual
decision being made, or their expectations of
possible future actions towards an object etc.
When researching the link between attitude
and behaviour, consultants may try to use the
information in one or two ways:
1 By measuring the cognitive and affective
components to predictfuture possible
behaviour.
2 By altering the cognitive and affective
components in order to influencefuture
behaviour.
The measurement of attitudes
Cook and Sellitz (1964) put forward, among
others methods, the following way in which
measured responses may give an indication as
to an individual’s attitude. They used tech-
niques which rely on a relatively direct style of
question which respondents answer in a way
which enables an inference to be made as to the
strength and direction of the attitude towards
the research’s object.
Measurement scales may be divided into
two groups: rating scales and attitude scales.
Rating scales measure a single component of an
attitude, a respondent typically indicating their
attitude to an object by means of a placement
along a continuum of numerical values or of
ordered categories. Scales can be labelled with
verbal or numerical descriptors, but in using
the former, the researcher should be aware that
some respondents may not think that there is
the same psychological difference between a
‘very’ and an ‘extremely’ as does the con-
structor of the scale; a pre-test should check on
this. By allocating a numerical value to the
object, depending on the strength with which
they hold a given attribute, measurement scales
may be used to measure:1 A respondent’s overall attitude towards an
object, product, person etc.
2 The degree to which something possesses a
certain attribute.
3 A respondent’s feeling towards a certain
attribute.
4 The importance with which a respondent
invests a certain attitude.Non-comparative rating scales
Respondents are asked to rate, assign a number,
to the object of interest in isolation, there being
no standard against which measurements are
made. Respondents mark their attitude posi-
tion on a continuum (a graphic scale), or they
may chose a response from a limited number of
ordered categories (an itemized scale).
For example, the question might be: ‘How
do you like Brand X of chocolate?’Comparative rating scales
Respondents make an assessment of the object
of interest against a stated standard.
For example, ‘How does Brand X of choc-
olate compare with Brand Y?’