CIGARETTE SMOKING, HEALTH, AND DISSONANCE
(PROJECT LIBRA)
1. INTRODUCTION AND ~IETHOD
REPORT NO. RD.1670 RESTRICTED
23.4.1979
ISSUED BY: D.J. Wood
DISTRIBUTION:
Dr. S.J. Green
Dr. I.W. Hughes
Dr. R.A. Sanford
R.M. Gibb, Esq.
R.S. Wade, Esq.
R.G. Nicholls, Esq.
Herr E. Rittershaus
Dr. F. Seehofer
Mr. A.J. Kruszynski
Dr. C.J.P. de Siqueira
Dr. D.G. Felton
Library
PROG. REF. 13.02.05
COPY NO.
CZD
(-T1
(J'l
<3 N
r1 j
r1 j
AUTHOR: H. Oldman
Copy No. 1, 2, 3
4
5, 6
7
8, 9, 10
11, 12
13
14
15
16
17
is, 19
BATCo document for Province of British Columbia 29
October 1999


Group Research & Development Centre,
British-American Tobacco Co. Ltd.,
SOUTHAMPTON.
MO/CAL/46J 23rd April 1979.
CIGARETTE SMOKING, HEALTH, AND DISSONANCE
(PROJECT LIBRA)
1. INTRODUCTION AND METHOD
(Report No. RD.1670 Restricted)
SUMMARY
Project LIBRA is a comprehensive quantitative
survey undertaken
among current cigarette smokers, ex-smokers and
never-smokers in the
United Kingdom. The principal objectives of the
research are to
explore
(i) the ways in which smokers, ex-smokers and
never-smokers
can be distinguished in terms of their attitudes
to health
in general, and smoking and health in particular,
and (ii) the indices of psychological difference
which best predict
the mode(s) of conflict resolution an individual
will employ
in relation to smoking and concern for health.
This report describes the background to the study,
the selection
of items for inclusion in the questionnaire, and
details of the sampling
and administration procedures adopted. The results
of the data analyses
will be presented in future reportc.
BATCo document for Province of BritiSh Columbia 29
October 1999


INTRODUCTION
It is now nearly thirty years since Wynder and
Graham reported that
tobacco smoking "seems to be an important factor
in the induction of
bronchiogenic carcinoma" (1). Since then
anti-smoking campaigns have
expressed themselves in the media, schools, and
workplaces. In the
U.K. and U.S.A., and other countries, health
warnings have appeared on
cigarette packets. Smoking in certain public
places has been restricted
and advertising on television has been banned in
some countries and
prohibited altogether in others. As a consequence,
the social climate
regarding smoking has changed and it is unlikely
that anyone in the
U.K., at least, is unaware of reasons why they
should not smoke. Hence,
most smokers wish to quit smoking.
Results from the recent Operation Aquarius (2)
have indicated that
more than sixty percent of the U.K. smoker
population can be described
as dissonant. That is, they are experiencing some
conflict between what
they do and what they believe they should do.
Specifically, they wish
to stop smoking. For more than half of those
smokers who had attempted
to quit, general or specific health reasons were
cited. Of the forty
percent of non-smokers who used to smoke, three
quarters were motivated
to give up smoking by health concerns.
It is possible to define four groups within a
smoker population,
which describes the market according to the amount
of consonance /dissonance
expressed by their behaviour and attitudes to
smoking:
Highly Consonant - those who have not tried nor
wish to quit;
Consonant - those who have tried but do not now
wish to quit;
Dissonant - those who wish to quit but have not
tried;
Highly Dissonant - those who wish to quit and have
attempted to do so.
ON
41-
BATCo document for Province of BritiSh Columbia 29
October 1999


-3-
According to data derived from Operation Aquarius,
the relative
sizds of these segments are as follows:.
Highly Consonant 20%
Consonant 16%
Dissonant 13%
Highly Dissonant 50X
The highly dissonant smoker is most likely to be
in the age range
25 to 34 years and to smoke 15 to 29 cigarettes
per day. In contrast,
sixty percent of all highly consonant smokers are
aged at least 45 years
and one quarter of them smoke five or less
cigarettes per day.
When asked "what were the reasons that made you
give up?" (ex-smokers),
or "what were the reasons that made you try to
give up?" (s mokers), the
following distri bution of responses was obtained:
Ex-smokers Smokers
(7.) M
Cost 33 43
Health (specific) 35 30
Health (general) 30 28
Other 9 13
Of particular interest in these data is the
apparent difference in
the relative importance of cost and health. For
smokers who had failed
to quit, cost was cited-most often as the
motivation to attempt, whereas
less ex-smokers were prompted by this reason. The
obvious interpretation
of this finding is that those for whom quitting is
the only satisfactory
means to resolve the conflict between smoking and
health concern have
already done so. Therefore, the difference between
smokers and ex-smokers
C=)
Un
L-n
Uri
BATCO document for Province of BritiSh COIUMbia 29
October 1999


-4-
in this regard, is historical. An alternative
hypothesis can, however,
be ddvanced: that those who have attempted to quit
but failed, deny the
importance of concern for health as a motive in
order to assuage the
dissonance which remains.
Whilst such an hypothesis is only tentative, it
highlights an
important gap in our knowledge of the consumer:
how does he remain a
smoker despite the apparently good reasons for him
to stop smoking? It
is our contention that all dissonant smokers, and
probably some apparently
consonant smokers, maintain their behaviour only
by making some psychological
adjustments to reduce the conflict between smoking
and concern for health.
It has already been suggested that one way of
reducing conflict is to deny
or devalue the health argument. We would suggest
that there are at least
five modes through which such a defense may be
obtained.
(a) By rationalising the health issue. For
example, a smoker
may maintain that there is too much fuss being
made about
the risks attaching to smoking, or he may assert
that "smoking
does not hurt anyone so long as they show no signs
of health
problems".
(b) By "statistical" rationalisation of the health
issue. This
mode would be reflected in agreement with such
statements as:
"Cigarette smoking only makes it more likely that
you would
get an illness you would get anyway" and "Many
non-smokers die
of illnesses that are often said to be caused by
smoking
cigarettes".
Both modes (a) and (b) share a fatalistic
disposition and offer the
rationalization that, through accidents or ageing,
people die any~4ay .....
CD
ON
r,-)
ON
BATCo document for Province of BritiSh Columbia 29
October 1999


-5-
(c) By modifying smoking behaviour. This is a more
truly rational
approach and would elicit concordance with such
statements as:
"There is nothing wrong with smoking so long as it
is in
moderation!', and "Low tar cigarettes are safer
than high
tar cigarettes". However, belief and action may
not match.
Faith in "safer" smoking is probably not
sufficient to resolve
conflict and would need to be supported by
appropriate buying
and consumption behaviour to be effective.
(d) By denying the authority of anti-smoking
information. Evidence
from Operation Aquarius (op.cit.) suggests that
smokers are less
likely than non-smokers to believe "the things
that the government
says about cigarette smoking and its effects on
health" and less
likely to accept "the things doctors say about
cigarette smoking
and its effect on health" - the latter because
doctors themselves
smoke and "doctors don't always agree with one
another".
It is not suggested that these modes of conflict
resolution or reduction
are exclusive or independent. Recent Canadian data
(3) supports the
existence of several factors in this area. The
fifth suggested made
of dealing with conflict is perhaps the most
rational and is likely to
be found amongst apparently consonant smokers
also.
(e) By acknowledging the risks attached to
smoking. Such a mode
would be expressed in agreement with a statement
such as
11 smoking might be harmful but I cannot stop" or
"the enjoyment
I get from smoking is worth any small risks it
might involve".
In the latter case, agreement presupposes that the
risks are
genuinely perceived as "small" or that some
rationalisation has
allowed that conclusion to be reached.
BATCo document for Province of BritiSh Columbia 29
October 1999


-6-
For the smokerg these and perhaps other ways of
reconciling smoking
with concern for health are available and
operating singly or in
combination. They are, however, unsuccessful.
Otherwise, half of the
smoker population would not express the wish to
quit because of their
health concern. Because denial, or reduction of
its importance, is
itself a means of partially resolving conflict,
then concern for health,
determined solely by direct questioning, is likely
to yield an underestimate
of its importance to some smokers.
In terms of sustained continuation of smoking
behaviour, only
twenty percent of the current U.K. market can be
considered "safe",
that is they are highly consonant, neither wishing
to nor ever having
attempted to quit. The immediate "loss potential"
of dissonant smokers,
who wish to but have not yet attempted to give up
smoking, is estimated
at 13%, with a further 50% highly dissonant, who
have tried to quit and
failed - so far.
It is our view that the true loss potential is
much greater than
those waiting to try but that we do not have
sufficient means of identifying
the long term behavioural fate of current smokers.
The present research,
Project LIBRA, will begin a programme to better
characterise the smoker
population in relation to concern for health,
which dimension is clearly
one of the most influential in the segmentation of
a contemporary market.
PROJECT LIBRA
Project LIBRA addresses itself to the following
questions:
1. In what ways can smokers, ex-smokers and
never-smokers be distinguished
in terms of their attitudes to health in general,
and smoking and
health in particular? (Attitudes to health.)
C=~
(-n
NJ
co
BATCo document for Province of British Columbia 29
October 1999


-7-
2. Which indices of psychological difference best
distinguish between
individuals who reduce conflict between smoking
and health concern
by various means (such as modifying their smoking
behaviour, adopting
various modes of dissonance reduction, etc.) and
between these and
successful quitters and never-smokers? (Individual
differences.)
Attitudes to Health
Data obtained from Operation Aquarius indicate
that smokers, ex-smokers
and never-smokers express different attitudes
towards smoking and health.
For example, 87% of ex-smokers and 89% of
never-smokers agreed with the
statement that "cigarette smoking is harmful to
health", whereas only
70% of smokers agreed with this assertion.
Although the majority of
smokers agree with the proposition in a general
way, they are less
concerned when asked about the possible effects.
of smoking oa their own
health. This apparent contradiction is a further
reflection of cognitive
dissonance (4).
In a recent market segmentation study carried out
in Canada (3),
four major segments of the smoker population were
identified. Concern
for health, or lack of it, was a major
differentiating feature for each
of the segments. Members of the largest segment
appear to have resolved
their conflict by modifying their smoking
behaviour. Such consumers
were found to more often want a low tar and
nicotine product, perceive
mildness to be of greater importance, to smoke
fewer cigarettes, and to
generally present a more health conscious profile.
In terms of a socio-
economic index they are classified as
predominantly medium-high to high.
An almost diametrically opposed segment was also
revealed, comprising
about 25% of all Canadian smokers. High inner need
motivation, a decided
lack of concern about tar and nicotine, a high
smoking volume, a
LJ1
C711
r1,)
110
BATCo document for Province of BritiSh Columbia 29
October 1999


-8-
requirement for a strong product. and a low level
of health concern
cliaracterised this group. They were,predominantly
young and male and
proportionately more represented in the medium-low
to low socio-
economic group.
From the foregoing, two important points arise.
Firstly, overt
questioning about smoking and health concern
differentiates smokers,
ex-smokers and never-smokers and, within the
smoker population, has
an important role to play in segmenting a
contemporary market. Secondly,
questions about health have hitherto always been
linked with smoking
behaviour with no attempt being made to view
expressions of concern
about specific health issues (i.e. related to
smoking) in a wider context
of general attitudes to health. Could it be, for
example, that non-smokers
are, as a group, more generally conscious of their
health and physical
well-being than smokers?
Because we doubt that for some smokers, real
concern for health
is adequately reflected in their answers to
specific smoking and health
questions, it is an important aspect of this study
to compare general
attitudes to health between smokers, ex-smokers
and never-smokers. This
will provide a context in which to judge specific
expressions regarding
smoking and health. Furthermore, it will be
important to determine
whether, for example, consonant and dissonant
smokers are differentiable
in terms of general attitudes to health. We might
expect that the
rational, consonant smoker is likely to have a
fatalistic attitude to
health generally, whereas the dissonant smoker,
depending upon his
particular mode of conflict reduction, might have
a more "anxious"
perception of his well-being. Comparisons between
smckers and ex-smokers
C:D
UN
N-)
CD
BATCo document for Province of BritiSh Columbia 29
October 1999


-9-
and never-smokers can similarly be expected to
provide further insights
i!ato broad health perception and its. effect in
determining behaviour
Individual Differences-
It is to be expected that an individual's attitude
and behaviour
regarding his own health will be reflected in
various indices of
individual difference. For example, the dimension
of locus of control
recognises that some people have an internal locus
of control O.e. they
perceive themselves as having considerable
influence in determining the
course of their lives), while others are more
externally (i.e. fatalistically)
controlled. Persons with an internal locus of
control are less likely
to be cigarette smokers; people with an external
locus of control who in
fact smoke are less likely to quit than those with
an internal locus of
control.
The finding that smokers are, on average, more
extraverted than
non-smokers is well established (5) and there is
some good evidence
to suggest that scores on the dimensions of
extraversion and teuraticism
have some power in predicting changes in smoking
behaviour, with stable
extraverts among men being most likely to stop
smoking (6). More
generally, it is to be expected that neuroticism
is highly correlated
with introspective. fears about one's well-being
and that this dimension
is likely to have some predictive power in
differentiating consonant
and dissonant smokers and ex-smokers within the
general population.
Similarly, standard measures of anxiety can be
expected to be of relevance
in this regard.
Project LIBRA incorporates indices of locus of
control, neuroticism
and anxiety because these have been shown to be
related to smoking
CD
U1
(-n
C7 \
r1,)
BATCO document for Province of BritiSh Columbia 29
October 1999


_10-
behaviour and concern for health, and it is
expected that they will be
of further use in the sub-classification of
attitudes which is intended
here. of less immediately apparent utility are
three further indices which
have been included: of ego-strength, of social
desirability, and of
authoritarianism.
The Thomas-Zander ego-strength scale (7) was
developed originally
to measure ego-strength as it related to
susceptibility to group pressure.
Ego-strength is conceptualized in two parts. First
is a person's
ability to be self-directing and to translate
intentions consistently
into behaviour (i.e. executive ability). Second,
is the ability to
control and discharge tension without disrupting
other psychological
processes (i.e. tension control). The scale has
been found to have
reasonably good test-retest reliability and
satisfactory convergent
and predictive validity (8). The scale has not
been widely distributed
because there is some concern about social
desirability confounding.
However, the incorporation in this research of
further items to test
this possible feature should provide sufficient
check on the veracity
of the ego-strength responses given. The potential
relationships
between ego-strength and both quitting behaviour
and the management
of dissonance are readily apparent.
The tendency for some questionnaire respondents to
"fake good"
is well known. It seems particularly important in
the present study
to attempt to measure and control any such
tendency, for both intrinsic
and extrinsic reasons. Intrinsically, some check
on the truthfulness
of responses is required, particularly in relation
to the other
N-)
BATCo document for Province of BritiSh Columbia 29
October 1999


psychographic data. Extrinsically, the actual
prevalence of "faking good"
'Within and between the various smoking/
non-smoking sectors is itself of
interest and of direct relevance to the modes of
conflict resolution under
investigation.
The final dimension of psychological difference to
be studied is
authoritarianism. The authors of the original
F-scale (9) conceived of
the authoritarian personality as comprising nine
variables: conventionalism,
authoritarian submission, superstition,
stereotypy, power and "toughness",
destructiveness, cynicism, projectivity, and sex.
Of the several forms
of the F-scale now available only the four-item
questionnaire (10) is
appropriate for the present purposes.
Authoritarianism is included in the
present study because it is believed to have some
potential value in
differentiating the population sub-groups of
interest and can be expected
to co-vary with some of the other dimensions being
measured. Of particular
interest are the relationships between denial of
authority and authoritarianism
and between general attitudes to health and
authoritarianism.
Covert Indications of Concern for Health
It is an assumption implicit in our general thesis
regarding smoking
and dissonance that actual concern for health and
expressed concern for
health are not, for some smokers, perfectly
correlated. At high actual
levels, expressions of concern may be suppressed
to give a distorted
picture of the individual's real attitude. This
hypothesized relationship
is presented schematically below.
Cr\
N-)
BATCO document for Province of BritiSh Columbia 29
October 1999


-12-
hi 00,
Expressed concern
for health
10
10 actual concern hi
for health
It would seem important to establish a "true"
measure of concern
for health for two reasons. Firstly, and of
implicit value to the
present project, the amount of mismatch between
expressed and actual
health concern is a measure of attempted conflict
resolution. Secondly,
and of wider eventual application, any measure
which was found to correlate
highly with true health concern but which did not
rely on explicit
reference to health would be of potential value in
determining the
level and distribution of health concern in
markets where direct methods
are considered inappropriate or inopportune.
With any questionnaire, the validity of the data
obtained depends
upon the ability and readiness of the respondents
to answer the questions
as accurately and honestly as possible. The
results of a recent study
to assess the present level of public knowledge
concerning the relationship
between the harmful effects of smoking and health
(11) support the
suggestion that smokers may deliberately deny to
others, and possibly
to themselves, a connection between smoking and
associated diseases. It
could be argued that smokers are at least as well
informed as non-smokers
about the possible health dangers of smoking.
Indeed, smokers are as
likely as anyone else to know of the association
between smoking and
Qn
(-n
CN
N-)
BATCo document for Province of BritiSh Columbia 29
October 1999


-13-
some diseases, and any apparent discrepancy
between the level of knowledge
found in smokers and non-smokers may not simply be
a reflection of how
well informed either group is. The probability has
to be considered
that smokers are more ready to deny the validity
of the evidence, or to
consciously suppress their awareness of overt
propaganda.
Conversely, it has to be allowed that non-smokers,
especially
ex-smokers are probably more prepared to attribute
harmful effects to
cigarette smoking, and this may serve to increase
any observed difference
in knowledge between the various groups.
Insofar as the information to be collected in
order to answer the
other questions posed for this research is of
relevance, the opportunity
will be taken to investigate the feasibility of
developing a covert index
of health concL-rn. It is to be expected, however,
that additional research
beyond the scope of the present study will be
required in this connection.
Additional Information to be Collected
In addition to the questions relating to general
health and
psychographics described above, which all
respondents answer, smokers
and ex-smokers have been presented with questions
concerning their smoking
(or previous smoking) behaviour and motivation. In
this way information
on consumption, brand choice, and switching will
be obtained together
with measurements of fundamental dimensions of
smoking motivation.
It is acknowledged that responses about previous
behaviour and
motivation elicited from ex-smokers of long
standing are at the mercy of
potentially inaccurate recall. It is nevertheless
important to attempt
to collect such retrospective information both for
adequate description
of the ex-smoker and to compare recent and stable
quitters.
CZD
ON
N-)
BATCo document for Province of British Columbia 29
October 1999


-14-
RESEARCH METHOD
The Sample
The target sample size for the study is 2,000
respondents. Based
upon data from Operation Aquarius (op.cit.) this
can be expected to
yield three major sub-groups, as follows:
740 (37%) current cigarette smokers;
520 (26%) ex-cigarette smokers;
740 (37%) who have never smoked manufactured
cigarettes.
The quota controls have been set within each of
these sub-samples
proportionate to the demographic profile of each
of these groups, again
based on Operation Aquarius data. The quotas set
for smokers were:
Sex
Age: 16-24 years
25-44 years
45-75 years
Class: A, B, CI
C2
D, E
The quotas for never-smokers and ex-smokers were
set within men and women
by age (three groups) with overall class controls
(three groups). The
detailed quota information is given in Table 1.
For reasons which will become clear, the length of
interview varies
by sub-group but respondents who qualify for the
longer interview (smokers
and ex-smokers) received a small incentive payment
for their co-operation.
The 2,000 interviews were conducted in 100
sampling points spread
throughout the country and selected on a
representative basis proportionate
C:>
Ul
(-n
GrN
NJ
BATCO document for Province of BritiSh COIUMbia 29
October 1999


-15-
to the population. Each interviewer conducted one
pilot interview beford
starting fieldwork.
England, Grosse and Associates Ltd. were
commissioned by CR&DC
to undertake the fieldwork and top-line analyses.
Subsequent analysis
will be undertaken by GR&DC using an appropriate
data management system.
TABLE I
QUOTA INFOMIATION
Total
Sex
Male
Female
ABe
16-24
25-44
45-75
Class
A,B,Cl
C2
D,E
Current
Manufactured
Cigarette Smokers
740
(7.)
48
52
Ex-Manufactured Never Smoked
Cigarette Manufactured
Smokers Cigarettes
520 740
Male Female Male Female
M M M (Z)
65 35 36 64
14 7 7 27 18
38 26 34 38 31
47 67 59 35 51
29 41 42
38 34 32
33 25 26
BATCo document for Province of BritiSh Columbia 29
October 1999


-16-
The Questionnaire
Detailed design of the LIBRA questionnaire began
in October, 1978.
Through discussions with England, Grosse and
Associates Ltd., evolution
of the questionnaire was directed towards
achieving the most comprehensive
data base from which to answer the major
questions. In its final form,
it was expected that smokers and ex-smokers would
be involved with the
interviewer for up to 90 minutes and never-smokers
for 45 minutes.
The questionnaire is presented in such a way that
respondents are
unaware at the outset that the major interest is
in smoking and health.
Thus the psychographic data and responses to
general health items are
recorded first and before questions relating to
smoking behaviour and
motivation. Depending upon whether the respondent
is a current cigarette
smoker, an ex-smoker, or a never-smoker, the route
through the remainder
of the questionnaire is determined. Thus, the
sequence of data blocks
after the preliminary classificatory information
is as shc,.-n in Table 2.
C7,,
CXD
BATCo document for Province of British Columbia 29
October 1999


-17-
TABLE 2
STRUCTURE OF THE QUESTIONNAIRE
PSYCHOGRAPHICS
GENERAL HEALTH
SMOKERS
. I
SMOKING
BEHAVIOUR
SMOKING
MOTIVATION
SMOKING &
HEALTH
ATTITUDES
NEVER-SMOKERS
SMOKING AND HEALTH ATTITUDES
EX-SMOKERS
I
PREVIOUS
SMOKING
BEHAVIOUR
I
PREVIOUS
SMOKING
MOTIVATION
EX-SMOKING
& HEALTH
ATTITUDES
Qn
rNI)
BATCo document for Province of British Columbia 29
October 1999


-18-
Psychographics
. The reasons for selecting the particular indices
used here have been
discussed earlier. It will be recalled that
measures of the following
dimensions were included:
Locus of control (4 items)
Anxiety (4 items)
Authoritarianism (4 items)
Ego-strength (7 items)
Social desirability (6 items)
Neuroticism (6 items)
Respondents self-completed all these scales which
for locus of control,
anxiety, authoritarianism and ego-strength
required an indication of
agreement/disagreement using the following-scale:
Agree Agree Neither Agree Disagree Disagree
Strongly nor Disagree Strongly
For social desirability and neuroticism a simple
"yes" or "no"
response is indicated.
The items selected for the locus of control
measure were taken from
Rotter's Internal-External Locus of Control Scale
(12); those for the
anxiety measure came from the Taylor-Spence
Manifest Anxietv Scale (13);
for the authoritarianism measure Lane's Four-item
F-scale (10) was adopted,
and ego strength is indicated by responses to the
7 Guttman scale items
contained in the Thomas-Zander Ego Strength Scale
(7).
The six items measuring social desirability derive
from the Eysenck
Personality Inventory (14) and those relating to
neuroticism were taken
from Eysenck (15).
Q71
U1
C7,,
BATCo document for Province of British Columbia 29
October 1999


_19-
General Health
The intention of this section is to gauge
respondents' attitudes to
their health and well-being. Items include
measures of..the number of
occasions respondents have visited their doctor
and/or been in hospital.
A 5-point scale of agreement/disagreement is used
to determine the
respondent's general perception of his health
through fourteen items
(including, for example, "I eat more than I
should". "I exercise regularly",
"there is too much fuss made about what one should
and should not eat").
Further questions seek to measure the respondent's
readiness to use
medication when suffering four common acute
ailments (headache, cold,
sore throat, and indigestion). They are also asked
if they are currently
taking, or have ever taken, any
tranquillizers/sedatives. Level of
consumption 'of tea, coffee, and alcohol are also
recorded.
At this point respondents are routed through the
remainder of the
questionnaire according to their smoking status.
Smokers - Smoking Behaviour
Information regarding length of time as a smoker,
level of consumption,
current and previous brands, and reasons for
switching is recorded here.
Smokers - Smoking Motivation
The early items in this section were derived from
a recent qualitative
study, Project VIRGO, and have been described in
detail elsewhere (16).
Further questions enquire into the respondents'
attempts to quit smoking,
their reasons for doing so and their reasons for
starting again. The
possible consequences of quitting for the
individual and the pressures
to quit which he/she feels to be under are also
gauged.
CD
QI
(-n
Q^J
BATCO document for Province of BritiSh ColUrnbia
29 October 1999


-20-
Smokers - Smoking and Health Attitudes
The final section of the smoker-specific items
comprises a further
set of agree/disagree statements which reflect
contemporary attitudes
to smoking and health and embody some of the
current rationalizations
smokers use to reconcile their smoking and health
concern. Two items
for each of the following modes are included:
Rationalize health issue
Compulsive usage despite recognition of health
issue
I'statistical" rationalization of health issue
Modifying smoking behaviour
Denial of authority
Fatalistic attitude to death.
Also, items' on the social acceptability are
included:
"Cigarette smoking is harmful to the health of
non-smokers"
"Cigarette smoking is a dirty habit"
"Nowadays it is much less socially acceptable to
smoke"
"Just because other people don't like smoking I
don't see why
I should not smoke in their presence"
Ex-Smokers
The items for ex-smokers are in the most part
similar to those for
smokers, except that they are phrased
retrospectively (i.e. "when you
were a smoker...."). Thus information on previous
smoking behaviour and
motivation is obtained. Statements regarding
strength of intention
to remain a non-smoker and reasons for those
statements are also recorded.
Ex-smokers are asked what, if anything, they miss
most as a result of not
smoking and what consequences for them were
noticed as a result of not
smoking. c~
C7'\
BATCO document for Province of BritiSh Columbia 29
October 1999


-21-
All - Smoking and Health Attitudes
The final section of the questionnaire is
addressed.to all respondents
and enquires into awareness of substances in
cigarettes, potentially
harmful substances, the harmfulness of smoking,
and physical effects of
smoking, whether or not the respondent has tried
to persuade someone
to give up smoking and who, and whether friends
and colleagues are smokers.
Current cigarette smokers are also asked about
their awareness of
the tar and nicotine level of their brand, ways of
reducing the risks
of smoking, and their overall view regarding their
attitude to smoking
and health.
Ex-smokers are also asked about their awareness of
the tar and
nicotine level of their brand when they were
smokers.
DISCUSSION
Project LIBRA is an investigation of the
relationships between
cigarette smoking. concern for health, and
dissonance. It is being
undertaken because insufficient knowledge exists
regarding the characteristics
of individuals who successfully quit smoking,
those who are unsuccessful in
their attempts, and those who do not express the
wish to try. In addition
to the major objectives already noted in respect
of this study, answers
to a number of supplementary questions will be
sought. For example, is
the current (simple) definition of cons onance/di
s sonance adequate to
predict behaviour or can the definition be
refined? Are attitudes to
smoking and health predictable from attitudes to
health generally? Are
attitudes to smoking and health, and attitudes to
health generally,
related to psychological traits? Can various modes
of conflict resolution
be determined from an individual's behaviour,
attitudes, and trait
structure?
C:D
(-n
BATCo document for Province of BritiSh Columbia 29
October 1999


-22-
Analysis of the data from Project LIBRA can be
expected to take
St least twelve months. It is intended that a
series of reports
be issued as particular aspects of the information
are digested and
interpreted. The second report in the series will
be concerned with
a highlight analysis of the top-line data.
(-n
Un
ON
BATCO document for Province of BritiSh Columbia 29
October 1999


-23-
REFERENCES
I.- Wynder, E.L. and Graham, E.A. Tobacco smoking
as a possible
etiological factor in bronchiogenic carcinoma: A
study of 684
proved cases. J. Amer. Med. Assoc., (1950), 143,
329-336.
2. NOP Market Research Ltd., Operation Aquarius: A
presentation of
the main findings, July 1978.
3. Imperial Tobacco Ltd., 1977 segmentation of the
French- and English-
speaking Canadian cigarette markets, June 1977.
4. NOP Market Research Ltd. A commentary on: The
main findings of
Operation Aquarius, August 1978.
5. Smith, G.M. Personality and smoking: A review
of the empirical
literature. In: Learning Mechanisms and Smoking.
Ed. W.G. Hunt,
(1970), Chicago: Aldine.
6. Cherry, N. and Kiernan, K. Personality scares
and smoking behaviour:
A longitudinal study. Brit. J. prev. Soc. Med.
(1976), 30, 123-131.
7. Zander, A. and Thomas, E. The validity of a
measure of ego-strength.
Unpublished paper, The University of Michigan,
Institute for Social
Research (1960).
8. Robinson, J.P. and Shaver, P.R. Measures of
Social Psychological
Attitudes, University of Michigan, Institute for
Social Research (1973).
9. Adorno, T. et al., The Authoritarian
Personality. New York:
Harper (1950).
10. Lane, R. Four item F-scale. In: Political
Personality and electoral
choice, American Political Science Review (1955),
49, 173-190.
11. Ashton, W.D. and Lloyd, G. Cigarette smoking
and associated disease,
Update, Ist May 1978, 1095-1101.
BATCo document for Province of BritiSh Columbia 29
October 1999


-24-
12. Rotter, J.B. Generalized expentancies for
internal versus external
control of reinforcement. Psychol. Monogr. 1966,
80, (1 Whole No. 609).
13. Taylor, J. A personality scale of manifest
anxiety. J. Abnorm.
Soc. Psychol., 1953, 48, 285-290.
14. Eysenck, H.J. and Eysenck, S.B.G. The Eysenck
Personality inventory.
London: University of London Press Ltd., 1963.
15. Eysenck, H.J. A short questionnaire for the
measurement of two
dimensions of personality. J. Appl. Psychol.
(1958), 14, 14-17.
16. BAT Report No. RD.1668, Restricted, 18.4.79.
011
BATCo document for Province of British Columbia 29
October 1999


BATCo document for Province of British Columbia 29
October 1999


Check List
DEJWLINE DATE: Cover:
I 0-~ I L ' I L. I
REPORT NO.: Rb VV'70 CATEGORY: R,-r, Bindings:
,Aj.Cr-j& CH0WNC2,j H E,4L-7114 # ANt>
b1-5501-4Ar4f&s/Graphs No.:
R.EpORT TITLE: CIC'
C-T- L-jrsZ-4~
photographs No.:
ko C-
LT=DCT JOB NO l!6 - v' ;L - ato FILE NO 4 6-~-
Reductions:
AUTHOR: V~-O Total Pages:
GROUP TXAMER:
ISSUER: Aem -2~vv
DISTRIBUTION: SjG ( ) : Imi (1) RAS (2) : VIC (1)
RSW (3) : Angr., P. % D.D. Aust. (2) IIS (1)
FS (1) AJK (1) CJP de S (1) : DGF (1) : Library
(2) : File
TAPE REF:
Typing Checking
Signs
-ture Date Date Date Author Date Issuer Date
Sent Recd Returned
Draft for typing (1)
(2)
(3)
(4)
Prepare master: (5) ~%t< L 11A /-1 .110.. ;. 0-1
OL4
Co Z".. C4L .3. 71
0
00.
Prepare I ~,,d Copy 41-
Issue: 7s7LJ -AFrcaoP, 1iv.56WTirjc- JOTZF
~'7` IL
This pro forma is intended to facilitate and
detail the progress of G.R. & D.C. Reports
the typescript stage until the report is finally
approved and issued. The pro forma
should be commenced at the time of the initial
typing and accomparly the typescript (and
master after preparation) at all subsequent
stages. It must be sent to central filing with
the master when the report is issued.
Note: 1. Wherever possible, drafts for typing
should be written in ink.
.2. After typing, the typescript should normally
be returned to the Author, who
will subsequently obtain the Group Leader's and
Issuer's approval.
3. Circulation should be determined by the Issuer.
4. The typescript must not be presented for
printing until the Issuer has initialled
the pro forma against "Prepare Master".
5. After printing, one bound copy must be returned
to the Issuer for final approval
and the Issuer will ensure that one copy is seen
by the Manager.
c0
BATCO document for Province of BritiSh Columbia 29
October 1999