Understanding and predicting human behaviour

Abstract

Understanding and predicting human behaviour has been of particular interest to researchers for many years. Moreover, the assumption that knowledge of attitudes will help in the task of predicting human behaviour has formed the basis for much consumer and social research. Attitudes are assumed to play an important role in human behaviour theory as the crucial link between what people think and what they do. Ajzen and Fishbein's (1980) attitude-based questionnaire framework has been widely used for the purpose of predicting behaviour. However, despite much study and refinement, limitations still exist with both the application and the predictive ability of their approach. Labaw (1980) offers an alternative approach to predicting behaviour in which behavioural aspects of people's lives form the basis of questionnaire design. Although less widely operationalised and tested than Ajzen and Fishbein's approach, recent investigation found that Labaw's approach to predicting behaviour was equivalent in terms of predictive ability, and was superior from a survey research perspective. Thus, Labaw's behavioural approach presents a feasible alternative to attitudinal-based approaches to predicting behaviour.

Full text

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
1
Understanding and predicting human behaviour
Judith Holdershaw and Philip Gendall
Massey University
Judith Holdershaw is a lecturer in the Department of Communication, Journalism and
Marketing at Massey University. Her research interests include social and health
related issues in Marketing. Email: j.holdershaw@massey.ac.nz
Philip Gendall is Professor of Marketing in the Department of Communication,
Journalism and Marketing at Massey University. His research interests include
questionnaire wording and questionnaire design, and survey research methodology.
Email: p.gendall@massey.ac.nz.
Abstract: Understanding and predicting human behaviour has been of particular
interest to researchers for many years. Moreover, the assumption that knowledge of
attitudes will help in the task of predicting human behaviour has formed the basis for
much consumer and social research. Attitudes are assumed to play an important role
in human behaviour theory as the crucial link between what people think and what
they do. Ajzen and Fishbein’s (1980) attitude-based questionnaire framework has
been widely used for the purpose of predicting behaviour. However, despite much
study and refinement, limitations still exist with both the application and the
predictive ability of their approach. Labaw (1980) offers an alternative approach to
predicting behaviour in which behavioural aspects of people’s lives form the basis of
questionnaire design. Although less widely operationalised and tested than Ajzen and
Fishbein’s approach, recent investigation found that Labaw’s approach to predicting
behaviour was equivalent in terms of predictive ability, and was superior from a
survey research perspective. Thus, Labaw’s behavioural approach presents a
feasible alternative to attitudinal-based approaches to predicting behaviour.
Introduction
It has generally been assumed that prediction of behaviour is best achieved by the
understanding and measurement of cognitive variables. Occupying a central position
in the study of behaviour research is the concept of attitude (Krosnick, Judd &
Wittenbrink, 2005). Kraus (1995) observed that the computerised database PsychLit
indexed more than 34,000 studies published since 1974 that address attitudes in some
way. Similarly, a review of empirical and conceptual developments on attitudes
between 1992 and 1995 by Petty, Wegener and Fabrigar (1997) reports that “a
voluminous amount of material was produced concerning attitude structure, attitude

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
2
change, and the consequences of holding attitudes” (p. 609). Recent observation of
computerised literature databases suggests many more attitude-related studies, across
a range of research disciplines, have been published since Kraus and Petty et al.’s
earlier observations.
Perhaps the most fundamental assumption underlying the attitude concept is the
notion that attitudes in some way, guide, influence, direct, shape, or predict actual
behaviour (Ajzen & Fishbein, 1974; Gross & Niman, 1975; Kraus, 1995). Thus, it is
not surprising that researchers interested in human behaviour theory ascribe great
importance to the role of attitudes in predicting and explaining human action. With
few exceptions, the assumption that attitude is useful for predicting behaviour went
unchallenged until the 1960s (Ajzen & Fishbein, 1980). In fact, Kraus describes the
first few decades of the twentieth century as an era of indifference to the attitude-
behaviour relationship. Many researchers simply assumed implicitly that attitudes
would be closely related to behaviour. The need to demonstrate that attitudes
predicted behaviour was not seen.
However, between the 1960s and the late 1970s, attitude research received much
criticism (Tuck, 1976; Eagly & Chaiken, 1993; Kraus, 1995). Years of early research
failed to provide strong support for the behavioural consistency or predictive validity
of attitudes. It was found that people neither behaved consistently in different
situations, nor acted in accordance with their measured attitudes. Thus, in time the
feeling grew that stated attitudes are not always consistent with overt behaviour. In
particular, a review by Wicker (1969) of 47 empirical studies of attitudes and
behaviours concluded, “it is considerably more likely that attitudes will be unrelated
or only slightly related to overt behaviours than that attitudes will be closely related to
actions” (p. 65). This review resulted in considerable controversy and caused many
researchers to question seriously whether attitude was still useful as a scientific
construct to predict behaviour (Fishbein & Ajzen, 1975; Tuck, 1976; Ajzen, 1987;
Kraus, 1995).
Ajzen and Fishbein’s approach to predicting behaviour
One explanation offered for the inconsistency in attitude-behaviour findings is that
historically researchers have not universally agreed on the components or elements of

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
3
the attitude construct, and as a consequence, nor have they agreed on an explicit
definition of attitude. With no clear definition of attitude available there was no clear
approach as to how attitudes should be measured, leading to a variety of measures of
‘attitude’ reported in the early literature (Gross & Niman, 1975; Tuck, 1976;
McGuire, 1985). For example, in a review of research published between 1968 and
1970, Fishbein and Ajzen (1972; cited in Fishbein & Ajzen, 1975) found more than
500 different procedures that had been used to measure attitude.
In recognition of early difficulties with the attitude-behaviour relationship, Ajzen and
Fishbein advanced a theory in which the attitude concept is examined in separate parts
(Fishbein, 1963, 1967; Ajzen & Fishbein, 1973). Specifically, the foundation for
Ajzen and Fishbein’s conceptual framework is provided by their distinction between
four components: beliefs, attitudes, intentions and behaviours.
Beliefs
A person’s attitudes are believed to form in response to the acquisition of certain
beliefs. Beliefs, therefore, are the fundamental building blocks upon which Ajzen and
Fishbein’s conceptual framework is based. Fishbein and Ajzen (1975) posit that
people may acquire beliefs on the basis of direct observation, or information received
from outside sources, or by way of various inference processes. East (1990) explains
that most people hold both positive and negative beliefs about an object (e.g. person,
action), and attitude is viewed as corresponding to the total affect associated with their
beliefs. For example, the belief that ‘the prime minister is an effective leader’ links
the object ‘the prime minister’ with the positive attribute ‘effective leader’. On the
other hand, a person may also believe that the object, ‘the prime minister’, is linked
with the negative attribute ‘is out of touch with ordinary people’.
Therefore, the attitude concept can be viewed as a set of beliefs, each belief can be
thought of as a separate attribute, and a person’s overall attitude toward the object is a
function of his or her evaluations of those attributes. Different people may have
similar beliefs about various objects but may give them quite different evaluative
weights. Thus, similar beliefs may result in different attitudes, depending on the
different evaluative weights given. Hence, individuals will vary in their attitudes
about, say, political voting preferences, depending on the strength and mix of beliefs

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
4
they have about this concept. Moreover, in the course of people’s lives, their
experiences lead to the formation of many different beliefs about various objects,
actions, and events. Some beliefs may persist over time, others may be forgotten, and
new beliefs may be formed. Some beliefs can be relatively stable, whereas others can
vary considerably. Therefore it follows, as beliefs are not static, neither are attitudes;
some attitudes may be relatively stable over time, and others may frequently alter.
Attitudes
Fishbein and Ajzen (1975) introduced the idea of corresponding measures of attitude
and behaviour, which has become central to their conceptual framework. They argue
that the attitude to an object is not necessarily related to the attitude to behaviour
towards that object, and that researchers’ failure to recognise this attitudinal
distinction has led to inaccuracies in behavioural predictions. For example, someone
may have a very favourable attitude towards a political party (the object), but not be
inclined to vote at the next election (the behaviour towards the object). Hence,
correlations between attitude to the object and action toward that object may not be
high. Therefore, Ajzen (1988) suggests if it is the action toward the behaviour a
researcher wishes to predict, it is the attitude towards performing this action that
needs to be measured.
Intentions
For many years researchers assumed that the relationship between attitude and
behaviour was direct. That is, the more favourable the attitude, the more likely
someone is to behave in accordance with that attitude, with no other variables
intervening the relationship. However, Ajzen and Fishbein (1980) disputed this
assumption, and argued that attempts to predict behaviour simply by measuring
attitudes will not succeed.
Within Ajzen and Fishbein’s conceptual framework, attitude is viewed as one major
determinant of the person’s intention to perform the behaviour in question. However,
other beliefs are also considered to be relevant for the formation of behavioural
intentions (Ajzen & Fishbein, 1970). Normative beliefs are those that occur due to
other people’s influence on whether an individual should or should not perform the

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
5
behaviour in question. For example, the influence of friends, family or work
colleagues may impact upon a person’s intention to vote for a political party. This
explains why two people may have the same attitude toward, say, The Green Party,
but may behave differently in terms of political voting, depending on the degree to
which other people influence their actions. In addition to attitude and normative
beliefs, Ajzen (1985) acknowledged that the formation of intentions to act may also
be influenced by aspects that are not under a person’s volitional control, such as the
requirement of certain abilities, or necessary resources. For example, in addition to
attitudinal beliefs or normative beliefs, a person’s intention to vote may also be
influenced by his or her perceived ability to travel to a polling booth on Election Day.
For this reason, the concept of a person’s perceived ability to act, should he or she
want to, was later included in Ajzen and Fishbein’s conceptual framework, to account
for situations where behaviour is not considered to be under a person’s voluntary
control.
Behaviours
Ajzen and Fishbein view behavioural intentions as the immediate antecedents of
corresponding overt behaviours; hence, the best prediction of behaviour is a person’s
intention to perform the behaviour. The apparent simplicity of this approach is
somewhat deceptive, however. Fishbein and Ajzen (1975) assert there are two factors
that can disrupt the intention-behaviour relationship. The first is the intervening time
between the stated intentions and the actual time of the act. Since it is often
impractical to measure a person’s intention immediately prior to performance of the
behaviour, the measure of intention obtained at one time may not be representative of
the person’s intention at the time of the behavioural observation (East, 1990). This is
due to the fact that behavioural intentions are affected by many situational factors,
which may intervene and disrupt the attitude-behaviour relationship. In turn this leads
to a situation where behavioural intentions do not correspond well with actual
behaviour. For example, Fishbein and Ajzen (1975) explain that if a person states an
intention to buy a car in three months time, any change in his or her financial position,
the price of the car, or the availability or price of petrol may influence that stated
intention.

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
6
A second factor that Fishbein and Ajzen (1975) suggest causes problems in attitude-
intention-behaviour measurement is described as the degree of compatibility in levels
of specificity. That is, an intention can only provide an accurate measure of a
predicted behaviour if there is compatibility in what exactly is being measured.
Therefore, Fishbein and Ajzen state it is important that the measures of attitude and
intention that are obtained are at the same level of specificity as the behaviour they are
trying to predict, in order to match cause and effect. That is, the more precise the
behavioural intention which is obtained, the more likely it is to be accurately related
to the subsequent behaviour.
Fishbein and Ajzen (1975) refer to an early, and frequently cited empirical study by
LaPiere (1934) to illustrate this aspect of their theory. LaPiere undertook an
investigation into racial prejudice in which he travelled across the United States with a
young Chinese couple. They visited 251 hotels and restaurants, and were refused
service only once. Studies at the time indicated that there was much anti-Chinese
sentiment in the USA. Six months after the trip, LaPiere wrote to each of the
establishments asking if they would offer service to Chinese guests. Of the 128 who
responded, 118 (90%) claimed that they would not serve them, in spite of the fact that
all had previously done so. This early study was frequently cited as evidence that
little correlation exists between attitudes and behaviour. However, in Fishbein and
Ajzen’s view, the measure used, which was whether the hotel and restaurant owners
“would accept members of the Chinese race as a guest in their establishments” may
have received a different response if they had worded the question “would you accept
a young, well-dressed, well-spoken, pleasant, self-confident, well-to-do Chinese
couple accompanied by a mature, well-dressed, well-spoken….educated European
gentleman as guests in your establishment?” (p. 375). That is, in Ajzen and
Fishbein’s view there was no compatibility between the attitude toward the behaviour
measure and the behavioural intention measure used in this study.
The predictive ability of Ajzen and Fishbein’s approach
Ajzen and Fishbein’s conceptual framework has led to improvements in the prediction
of behaviour since Wicker (1969) earlier questioned the accuracy of predictions based
on attitudinal data, and forms the basis of most attitude research today (Foxall, 2005).
Yet meta-analyses of research using their framework show it only explains, on

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
7
average, between 40% and 50% of the variance in intention, and between 19% and
38% of the variance in behaviour (Sutton, 1997). Thus, questions are still being
raised about the performance of attitude models in predicting and explaining
intentions and behaviour (Chandon, Morwitz & Reinartz, 2005; Foxall, 2005). Ajzen
(2002) recently conceded that, despite numerous attempts to improve the limitations
and predictive ability of their models “vexing problems remain” (p. 666).
Foxall (2002) argues that the main problem with investigations of human behaviour to
date is the prevailing view that prediction can occur from measures of beliefs,
attitudes and intentions, regardless of situational factors. Foxall does not single out
the Ajzen and Fishbein approach per se for criticism but suggests that as their
approach represents the most sophisticated methodology available to researchers,
whatever limitations apply to it will also affect the less sophisticated methods of
investigating or predicting behaviour. Clearly, there is a need to consider alternative
approaches to predicting behaviour than those based on attitudes.
Labaw’s approach to predicting behaviour
One avenue of reappraisal of the use of cognitive variables to predict behaviour is to
consider an alternative approach based on behavioural variables, such as the one
proposed by Labaw (1980). Labaw’s background was as a researcher in the area of
public opinion surveying. She proposed a foundation for a systematic theory of
questionnaire design, which places much emphasis on the underlying framework of
questionnaire development (Gendall, 1998).
Labaw’s (1980) approach to questionnaire design developed as a result of her lack of
success using the accepted attitudinal approach to predicting behaviour. Labaw does
not believe that accurate prediction of human behaviour using attitude concepts is
possible using questionnaire techniques. She states:
Frustrated with the lack of predictability of purely attitudinal
questions, and rather stunned by the huge gap between what
people say and what they then do, I felt it necessary to re-evaluate
the role of these types of questions within surveys and to find
alternatives to them which could be used in predicting behaviour.
Consequently, attitude questions have become a minute part of
surveys I design (p. 32).

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
8
To Labaw, attitudes are often what she refers to as “mere surface manifestations of
larger, structural movements beyond the control or even the consciousness of
individuals” (p. 82). The inference of this statement is that rather than ask
respondents about their attitudes, which they may not understand themselves, it is
more helpful simply to ask direct questions about the aspects of their lives upon which
their attitudes are based. Labaw’s rationale for this approach is that people are much
better able to say what they do now and have done previously, compared with what
they might do in the future.
An underlying theme of Labaw’s approach to predicting behaviour is to determine
levels of respondent consciousness. She argues that, as a rule, people have not
thought about, and do not know their feelings about most issues that have not directly
affected them. Thus it is important to determine how close to the issue or action of
interest the respondent is. Closeness for survey purposes is equivalent to firsthand (or
fairly direct) personal experience. Accordingly, Labaw’s approach to predicting
behaviour is to determine respondents’ levels of consciousness about the behaviour of
interest using experience-related questions. Using an analogy based on blood
donation, Labaw’s approach suggests that someone who has donated blood knows his
or her feelings about blood donation better than someone who hasn’t, and similarly,
someone who knows someone else who has donated blood has a greater level of
consciousness about the topic than someone who doesn’t know someone who has
donated blood. Hence, she argues, there are various levels of consciousness,
depending on a person’s level of experience with an action. As a consequence,
Labaw’s approach to predicting behaviour assumes that someone with greater direct
or even indirect experience of blood donation would be more capable of accurately
indicating their willingness to donate blood themselves than someone with no
experience.
Labaw argues that by replacing attitudinal questions with behavioural questions, it is
the researcher not the respondent who analyses and projects meaning from the
responses. In her view, cognitive questions, based on hypothetical situations, often
shift the responsibility for analysis from the researcher to the respondent. Instead, she
argues the researcher should use questions in which the respondent can accurately
describe what he or she does, and in this way, the answers do not require imaginative

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
9
skill, or projectable consciousness on the part of the respondent. In essence, Labaw
concluded that to predict behaviour, attitudinal questions, the answers to which can
never be externally validated, should be replaced by questions that respondents can
‘truthfully’ answer: that is, questions with a verifiable answer on which researchers
can base predictions of future behaviour.
Specifically, Labaw argues that respondent consciousness in relation to the topic or
action of interest relates to three components, which she believes provide a means of
predicting actual behaviour. Labaw argues that by adopting an approach in which we
establish the respondents’ environment, what respondents know, and their past
behaviour, researchers could make better predictions of their behaviour than is
possible by measuring their attitudes. Each component of Labaw’s approach to
predicting behaviour is discussed in the following sections.
Environmental Influences
The first component of Labaw’s approach to predicting behaviour is the environment,
which Labaw describes as the physical aspects of people’s lives over which they have
little control but which impinge on their ability to act or respond in specific ways,
regardless of their attitudes. These aspects include age, gender, health status, location,
mobility level, and education level. Labaw argued these aspects are important
because they provide greater depth to understanding human behaviour than attitudes,
which may be much shorter lived.
Knowledge
Labaw refers to the second component in her approach, knowledge, as a respondent’s
level of knowledge about the topic or related action. Whilst definitions of the exact
components of knowledge and techniques to measure knowledge may vary, it is
generally agreed that an individual’s level of knowledge about an action relates to his
or her subsequent behaviour to that act (Brucks, 1986; Allen & Butler, 1993;
Andreasen, 1995). Thus, Labaw argues that the way people behave often results from
what they know about an action, or equally, their behaviour may relate to what they
do not know. Hence, based on her approach it could be reasoned that a respondent’s
level of knowledge about, say, a product’s country of origin may have a direct

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
10
influence on his or her purchase behaviour. For example, purchase behaviour of fresh
produce from a supermarket may be influenced by whether or not people know if it
was produced locally or imported from overseas.
Past Behaviour
The third component of Labaw’s approach to predicting behaviour is people’s actual
behaviour; in particular, her approach emphasises the importance of past behaviour as
a predictor of future behaviour. Debate frequently arises among behaviour
researchers as to the role of past behaviour in determining future behaviour. For
example, within Ajzen and Fishbein’s conceptual framework, future behaviour is
shaped by beliefs, and beliefs are thought to incorporate past experience. However,
even proponents of attitude-based models have found that people’s future behaviour is
more accurately determined by measures of past behaviour, compared with those
provided by cognitive measures (Foxall, 1997; Sutton, 1998).
Labaw argues that behaviour questions are particularly important in areas where
potential future behaviour is under study. In particular, the frequency with which
behaviours have been performed in the past tends to correlate well with later actions.
Thus, she posits, when designing a questionnaire aimed at determining potential
future behaviour, the most important design technique is to include a battery of
behaviour questions detailing past and present behaviour that is similar to or related in
some way to the potential behaviour under study.
Comparison of two approaches to predicting behaviour
Labaw’s approach to predicting behaviour, using questions that can be ‘truthfully’
answered, has not been as widely operationalised or tested as Ajzen and Fishbein’s
conceptual framework. Numerous studies have been undertaken in which various
aspects of Ajzen and Fishbein’s approach have been examined and tested, and the
results are widely disseminated in the academic literature. By contrast, Labaw’s
approach is not cited in any of the well-known survey research or questionnaire
design texts. Nevertheless, given the limitations currently identified with cognitive
approaches to predicting behaviour, a need clearly exists to consider alternative
approaches such as Labaw’s.

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
11
The main distinction between Ajzen and Fishbein’s approach to predicting behaviour,
and Labaw’s approach lies in the nature of the questions used to formulate a
questionnaire. Ajzen and Fishbein’s questionnaire design incorporates questions in
which respondents are asked their feelings and beliefs about the behaviour under
investigation. This typically involves the use of numerous questions using standard
attitude scaling procedures, most commonly the semantic differential. By contrast,
Labaw’s approach uses questions that, at least theoretically, have verifiable answers.
Two recent studies (Holdershaw, Gendall & Wright, 2003; Holdershaw, 2006), in
which direct comparisons were made of the ability of each approach to predict blood
donation behaviour, found that Labaw’s approach was equivalent to Ajzen and
Fishbein’s in terms of the variance explained. In absolute terms, neither approach was
good at predicting blood donation behaviour. Lack of variation in the sample may
have been a factor in this finding; only 12% of respondents reported that they had
donated blood. Thus, additional research is needed to further test the predictive
ability of Labaw’s approach using various behaviours of interest.
However, an important finding of the two studies was that greater ease of
questionnaire application in the field was achieved using Labaw’s survey design,
compared with Ajzen and Fishbein’s attitudinal approach to questionnaire design. No
discernible difficulties occurred with the application of Labaw type questions in the
field, but limitations were identified with application of Ajzen and Fishbein’s belief-
based questions. Liska (1984) also noted that Fishbein and Ajzen models are more
strongly supported in laboratory than field studies. Given that Ajzen and Fishbein’s
conceptual framework is so pervasive in survey methodology, this finding was
important.
One aspect of respondent dissatisfaction with cognitive questions occurred due to an
apparent difficulty respondents had in differentiating between the two components of
the belief-based measures, which consist of paired items. Ajzen (1985) provides
sound methodological reasons for the wording and question order of paired items, yet
application in the field does not appear to complement the rationale for the
methodology. For example, one respondent in Holdershaw’s (2006, p. 252) study
commented, “there are too many scale questions which seem completely irrelevant.
They all seem to be asking the same thing.” In a similar study, Knight (1983) too

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
12
reported difficulties with respondent fatigue when presenting an Ajzen and Fishbein
type questionnaire in several parts. Initially he had intended to use three parts to a
question, but proceeded with two parts after pilot testing found respondent fatigue to
be a problem, even with only two sections to each question. Knight states:
It was impossible to avoid repetition which led to irritation and
boredom on the part of some respondents. In fact, two questions
were dropped mid-way through the survey and two others were
deliberately omitted for certain respondents to ensure that the most
important questions were answered (p. 68).
Oppenheim (1992) explains a difficulty with attitudinal based research is that
attitudinal questions are more sensitive than factual questions to changes in wording,
context, emphasis, and so on; therefore it becomes almost impossible to assess
reliability by asking the same question in another form. Hence, typically numerous
questions are asked and then it is necessary to ask the same question in similar forms.
Of course, the rationale for this approach is not obvious to respondents.
When comparing Ajzen and Fishbein’s and Labaw’s survey methodology,
Holdershaw (2006) also found that some respondents experienced difficulty
understanding what certain attitude-based questions asked. No such difficulty with
question understanding occurred with the Labaw-type questions, which are factual in
nature, rather than based on internal thoughts and feelings. Labaw defines one
example of a bad question as one that is incomprehensible to the respondent because
the wording, the concepts, or both, cannot be understood. In her view, bad questions
are any questions that obscure, prohibit, or distort the fundamental communication
from respondent to researcher. Instead, Labaw posits that a questionnaire should be
designed to prevent it becoming simply an instrument of the writer’s perceptions,
values, and language, which is then inflicted upon the respondent. Arguably, many
attitude-based questions routinely used in survey research would be labelled ‘bad’
questions according to Labaw’s definition.
Conclusion
Examination of current research practice suggests that, by and large, researchers opt
for a cognitive-based questionnaire framework, designed to attempt to understand
‘what is going on inside people’s heads’, as a basis for predicting future behaviour. In

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
13
particular, Ajzen and Fishbein’s attitude-based approach is considered the most
sophisticated survey methodology available to researchers for behavioural prediction.
However, recent research supports the use of a viable alternative to the continued, and
often unquestioning, reliance on attitudinal questions as a basis for understanding and
predicting behaviour. Instead, it is suggested that greater use of questions that, at least
theoretically, have verifiable answers is incorporated into research design.
Comparison of Ajzen and Fishbein’s attitude-based survey methodology with
Labaw’s behavioural approach found that the predictive ability of the two approaches
was equivalent; however, Labaw’s behavioural approach was superior from a survey
research perspective.
References
Ajzen, I. (1985). From intentions to actions: A theory of planned behaviour. In J. Kuhl
& J. Bechmann (Eds.), Action control: from cognition to behaviour. Berlin:
Springer-Verlag, 11-39.
Ajzen, I. (1987). Attitudes, traits, and actions: Dispositional prediction of behavior in
personality and social psychology. Advances in Experimental Social
Psychology, 20, 1–63.
Ajzen, I. (1988). Attitudes, personality, and behavior. USA: The Dorsey Press.
Ajzen, I. (2002). Perceived behavioural control, self-efficacy, locus of control, and the
theory of planned behaviour. Journal of Applied Social Psychology, 32(4),
665–683.
Ajzen, I., & Fishbein, M. (1970). The prediction of behaviour from attitudinal and
normative variables. Journal of Experimental Social Psychology, 6, 466–487.
Ajzen, I., & Fishbein, M. (1973). Attitudinal and normative variables as predictors of
specific behaivours. Journal of Personality and Social Psychology, 27(1), 41–
57.
Ajzen, I., & Fishbein, M. (1974). Factors influencing intentions and the intention-
behavior relation. Human Relations, 27(1), 1–15.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social
behaviour. USA: Prentice-Hall, Inc.
Allen, J., & Butler, D. D. (1993). Assessing the effects of donor knowledge and
perceived risk on intentions to donate blood. Journal of Health Care
Marketing, 13(3), 26–33.
Andreasen, A. R. (1995). Marketing social change. San Francisco: Jossey-Bass
Publishers.

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
14
Brucks, M. (1986). A typology of consumer knowledge content. Advances in
Consumer Research, 13(1), 58–63.
Chandon, P., Morwitz, V. G., & Reinartz, W. J. (2005). Do intentions really predict
behaviour? Self-generated validity effects in survey research. Journal of
Marketing, 69(April), 1–14.
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. USA: Harcourt
Brace Jovanovich.
East, R. (1990). Changing Consumer Behaviour. London: Cassell Educational
Limited.
Fishbein, M. (1963). An investigation of the relationships between beliefs about an
object and the attitude toward that object. Human Relations, 16(3), 233–239.
Fishbein, M. (1967). Readings in attitude theory and measurement. USA: John Wiley
& Sons.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An
introduction to theory and research. Reading, Massachusetts: Addison-Wesley
Publishing Company.
Foxall, G. R. (1997). Marketing psychology: The paradigm in the wings. London:
Macmillan Press.
Foxall, G. R. (2002). Marketing’s attitude problem – and how to solve it. Journal of
Customer Behaviour, 1, 19–48.
Foxall, G. R. (2005). Understanding consumer choice. UK: Palgrave MacMillan.
Gendall, P. (1998). A framework for questionnaire design: Labaw revisited.
Marketing Bulletin, 9, 28–39.
Gross, S. J., & Niman, C. M. (1975). Attitude-behavior consistency: A review. Public
Opinion Quarterly, 39(3), 358–368.
Holdershaw, J. (2006). Comparison of two approaches to predicting behaviour.
Unpublished Doctoral, Massey University, Palmerston North.
Holdershaw, J., Gendall, P., & Wright, M. (2003). Predicting willingness to donate
blood. Australasian Marketing Journal, 11(1), 87-96.
Knight, R. J. (1983). The stimulation and planning of blood donation: A marketing
problem. European Journal of Marketing, 17(6), 65–73.
Kraus, S. J. (1995). Attitudes and the prediction of behaviour: A meta-analysis of the
empirical literature. Personality and Social Psychology Bulletin, 21(1), 58–75.
Krosnick, J. A., Judd, C. M., & Wittenbrink, B. (2005). The measurement of attitudes.
In D. Albarracin, B. T. Johnson & M. P. Zanna (Eds.), The handbook of
attitudes (pp. 21–76). New Jersey: Lawrence Erlbaum Associates, Inc.

ANZCA08 Conference, Power and Place. Wellington, July 2008
ANZCA08: Power and Place: Refereed Proceedings: http://anzca08.massey.ac.nz
15
Labaw, P. J. (1980). Advanced questionnaire design. USA: Abt Books.
LaPiere, R. T. (1934). Attitudes vs. actions. In E. Aronson & A. R. Pratkanis (Eds.),
Social Psychology (Vol. 1, pp. 576–583). Great Britain: Edward Elgar
Publishing Ltd.
Liska, A. E. (1984). A critical examination of the casual structure of the
Fishbein/Ajzen attitude-behaviour model. Social Psychology Quarterly, 47(1),
61–74.
McGuire, W. J. (1985). Attitudes and attitude change. In G. Lindzey & E. Aronson
(Eds.), The handbook of social psychology: Vol. 2. (3rd ed., pp. 233–346).
New York: Random House.
Oppenheim, A. N. (1992). Questionnaire design, interviewing and attitude
measurement. London: Pinter Publishers Ltd.
Petty, R. E., Wegener, D. T., & Fabrigar, L. R. (1997). Attitudes and attitude change.
Annual Review of Psychology, 48(1), 609–647.
Sutton, S. (1997). Theory of planned behaviour. In A. Baum, S. Newman, J.
Weinman, R. West & C. McManus (Eds.), Cambridge handbook of
psychology, health and medicine. (pp. 177–180). Cambridge, UK: Cambridge
University Press.
Sutton, S. (1998). Predicting and explaining intentions and behaviour: How well are
we doing? Journal of Applied Social Psychology, 28(15), 1317–1338.
Tuck, M. (1976). How do we choose: A study in consumer behaviour. London:
Methuen.
Wicker, A. W. (1969). Attitudes versus actions: The relationship of verbal and overt
behavioural responses to attitude objects. Journal of Social Issues, 25(4), 41–
78.
Copyright Statement: Articles submitted for ANZCA08 remain the copyright of the author, but authors
by virtue of submission agree to grant the Department of Communication, Journalism & Marketing at
Massey University a copyright license to permanently display the article online for public viewing as
part of this conference proceedings, and to grant the National Library of Australia a copyright licence
to include the ANZCA08 Proceedings in the PANDORA Archive for permanent public access and
online viewing. Articles first published in this ANZCA08 proceedings may subsequently be published
elsewhere by authors, provided the next version acknowledges this original publication in the
ANZCA08 refereed proceedings.