OVERDUE FINES ARE 25¢ PER DAY
PER ITEM

Return to book drop to remove
this checkout from your record.

 

 

 

 

SOCIAL COMMITMENT AND GOAL SETTING IN SMOKING CONTROL

By

Richard A. Strand

A THESIS
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

ABSTRACT
SOCIAL COMMITMENT AND GOAL SETTING IN SMOKING CONTROL
By

Richard A. Strand

The current study tested the effect of a behavioral commitment
to quit smoking on actual cigarette consumption. Behavioral commitments
were made in two ways. First, some Subjects were directed to make a
social commitment to their close acquaintances of.their intention to
quit smoking. Criticisms of previous attempts at inducing social var-
iables consisted of their lack of generalization to the smoker's total
environment and their consequent limited value to the smoker. The
second behavioral commitment consisted Of having subjects make smaller
and smaller daily goals with respect to cigarette consumption. It was
expected that both forms of behavioral commitment would produce dis-
sonance. A marginally statistically significant three way interaction
indicated that social commitment resulted in a slower recidivism rate.
Also, a statistically significant correlation was obtained between per-
cent of total friends positive about the effort to quit and three month
follow-up percent of baseline rate. Explanations of the effects of the
commitment treatments are based on social support. Recommendations are
made for further field testing of social commitment to quit cigarette

smoking and the social support variable.

To Linda

ACKNOWLEDGMENTS

I would like to thank the members of my committee, Terry Allen,
Larry Messe’, and especially the Chairperson of my committee, Neal Schmitt,
for their assisting with this work. I am also appreciative to the
Department of Psychology for the financial, material, and typing
assistance that was provided to me. Special thanks to my assistants,

Steven Armstrong and Angela Gondolfo, for their contributions.

ifi.

TABLE OF CONTENTS

 

LIST OF TABLES..... . . ..... . ........ . ..... . . ..............
LIST OF FIGURES..... ..................... .. ...... .. . ...... . ..
LIST OF APPENDICES ..................... . ...... . . ..............
Chapter
I. Introduction.. ...... .................................
Literature Review ................................. ...
Commitment Treatments ....... . ....... . ................
II. Method.. ............ ........... ...... . .......... .....
Subjects ............. . ................... . ...........

Procedure.........OOOOOOOOOOOOOO......OOOOOOOOOOOOOOO

Design and Treatments. ..................... ..........

III. Results..... ........ ... ...... . .............. .........
Dependent Variable Reliability............... ..... ...
Manipulation Checks ............. .... ..... .... ........

IV. Conclusions....................... ..... ..............
APPENDICES..... ......... . ...... ......... ....... . ...... ... ......
BIBLIOGRAPHY ..... . ..... . ............................. .. ........

iv

10

l9

19

19

22

25

25

25

42

LIST OF TABLES

Table Page
1 Analysis of Covariance Table .............................. . 27
2 Number of Persons Abstinent ................................ 30

LIST OF FIGURES

Figure

1 Group Mean Cigarettes Smoked Per Day...

vi

LIST OF APPENDICES

Appendix
A. Directions for Transaction Monitoring.......... ........
B. Directions for Prompts Control.... ............ . ........

vii

INTRODUCTION

Research in cigarette smoking control has shown that no one treat-
ment procedure is superior to the rest. Also, the research shows that
current treatments are only temporarily and marginally effective at best--
results are plagued with high recidivism rates and control groups are
often just as effective if not more so. Recent research in the area has
focused on the process taking place during purposeful quitting. The pur-
pose of this study is to use two of the process variables found to be
related to successful abstinence by operationalizing them into treatment
variables in a smoking control experimental study.

Through the use of post hoc questionnaire analysis and from the use
of certain analysis and from the use of certain control groups it has been
found from past research that amount of effort to quit and expectancy of
success in quitting is strongly related to the reaching and maintenance
of abstinence in cigarette smoking. In this study, two commitment type
treatment variables-~social commitment and goal setting-~are hypothesized
to induce high levels of these effective process variables. The two
treatment variables are crossed with one another and manipulation checks
are used to verify that they were employed by the subjects.

The theory, which the effectiveness of the treatment variables in
this study are based upon, is different than the theories which most
treatment variables are founded upon. A very large percentage of smoking
control treatments come from an operant conditioning framework. The

past treatments, in effect, are manipulations of contingencies to the

smoking behavior. The new process research has shown the relatedness of
cognitive and attitudinal behavior in abstinence. In this study the use
of the treatments and their hypothesized effectiveness is based on a cog-
nitive and attitudinal theory.

Literature Review

 

Cigarette smoking is predominantly considered in the smoking control
literature as a behavioral problem and as such, therapeutic attempts to
promote abstinence have been based heavily on behavioral modification
techniques. Reviews of the literature on behavioral modification tech-
niques in smoking control (e.g., Lichtenstein & Danaher, 1976; Bernstein,
1969) have shown that no one treatment technique has been clearly suc-
cessful. A number of patterns, however, have consistently occurred. One
common occurrence running through the literature is that there is a highly
significant reduction rate during treatment irrespective of the treat-
ment technique. A second pattern is that the reduction rates during
treatment, for all treatment techniques and across all studies, are re-
markably similar (McFall & Hammen, 1971). Probably the most discouraging
aspect of the results pertains to the relapse rate of those persons who
reach abstinence. In a review of 89 smoking control studies, Hunt and
Bespalec (1974) found a negatively accelerating rate of relapse with
most relapse occurring during the first month after treatment and level-
ing out at the three to sixth month period. The percentage of those
persons abstinent at the end of treatment who remain abstinent asymptotes
in the 20-30% range. Amongst other reasons, these results justify Yates'
(1975) title to his chapter on smoking: "When Behavior Therapy Fails."

The treatment effects and relapse pattern of cigarette smoking in-

terventions described above have been observed since the beginning of

decade and remarked upon in a number of review in the smoking control
literature (e.g., Hunt & Matorazzo, 1970; Bernstein, 1970). From the time
of these reviews in the early 1970's to now many of the ideas tested in
the literature on the treatment effects on cigarette smoking have orig-
inated from these earlier reviews. As a result, research in smoking con-
trol studies in the 1970's changed from the quest for the best or most
powerful techniques to a more humble search for the underlying processes
in a person's change of smoking behavior (Lichtenstein, 1971). The new
line of research in the 1970's has taken on four directions. In brief,
these directions are:

l. A search for the non—specific effective variables in the treat-

ment manipulations.

2. The testing and use of more rigorous methodologies.

3. The testing and use of abstinence maintenance procedures.

4. The testing of the relationship between psychological traits

and treatment effects.

The first new line of research, the search for non-specific vari-
ables, was initiated by Bernstein (1970). Non-specific effective vari-
ables are those variables which affect the outcome of the treatment but
are in themselves not psychological treatments per se. In a series of
three experiments, Bernstein (1970) compared the efficacy of a social
pressure group to a number of control groups. The control groups in—
cluded a group placebo, an individual placebo, an effort control, an
expectation control, and a contact control. In the conclusion of his
three experiments, Bernstein (1970) stated that "the sufficient condi-
tions for initial smoking behavior change in volunteer groups are: par-
ticipation in an experiment, expectation of change, suggestion, and con-

tact with an experimenter." The results of one of the groups in

4
Bernstein's (1970) study is noteworthy. The group, expectancy control,
contained subjects that were told that the most important variable in
quitting smoking is their determination to quit. These subjects were
neither promised nor offerred treatment. The results for the expectancy
control group were statistically no different from the social pressure
group. In Bernstein's (1970) third experiment the expectancy control
group tended to perform better than any of the other groups. Bernstein
(1970) concludes that for any treatment to be proven effective it must
not only be more effective than placebos but also the 8's effort itself.

In another study aimed at locating the role of non-specific factors,
McFall and Hammen (1971) compared the effects of a number of different
self monitoring procedures on smoking behavior. No significant differ-
ences between groups were observed. The groups, though, did reduce their
smoking levels commensurate with previous studies. The authors conclude
that the effective non-specific variables in smoking control studies are
motivated subjects, structured participation, and self monitoring.

Related to the search for the effective non-specific variables in-
volved in smoking control studies is the new line of research which
tests and uses different methodologies. Included in this category of
research are techniques which increase the internal validity of the de-
sign. One empirical issue is the drop-out rate of subjects. The use
of monetary deposits to remedy this problem was originally employed by
Mees (1966) and has since become almost a standard practice in most
smoking control studies (e.g., Best & Steffy, 1975; Marston & McFall,
1971; McFall & Hammen, 1971; & Winett, 1973). The deposits range from
$10 to $60. The deposit procedure has been very successful in reducing

the number of drop-outs in any one study. The use of monetary deposits

in itself, though, must be considered a modifier of a person's smoking
behavior. One of the most successful smoking control studies, reported
by Elliot and Tighe (1968), employed a $50 to $60 deposit that was re-
turnable in parts over the 16 weeks of the program contingent on no
smoking. 0f the 25 8'5 in the study 84% reached abstinence and 37.5%
were still abstinent at a three month folloWbup. Due to lack of controls
little could be concluded from the study. Winett (1973) tested the
effectiveness of deposit returns contingent or non-contingent to meeting
pre-assigned smoking level criteria. At the end of treatment the contin-
gent repayment group outperformed the non-contingent group. At the three
and six month follow-up there were no differences between the groups.
Another empirical issue in this second line of new research dealing
with increasing the internal validity of the experimental designs is the
self-reporting of the smoking rate of the subjects. Any study which
employs subjects' self-reports on the dependent variable must be suspect
to data falsification. Some attempts to increase the reliability of the
self disclosure of smoking rates during treatment include the use of
subjects' acquaintances to validate the report (e.g., Best & Steffy,
1975; & Winett, 1973) and the limitation of the subject's access to
cigarettes to the experimenter (e.g., Marston & McFall, 1971). Like the
use of security deposits to improve internal validity, monitoring pro-
cedures in themselves effect the smoking behavior of the subjects. In
a series of experiments by Frederikson, Epstein and Kosevsky (1975),
three self-monitoring procedures were tested for reliability and con-
trolling effects. The continuous recording procedure proved to be both
more reliable and more effective in reducing smoking rate than daily or
weekly recording. There were also more drop-outs in the continuous re-

cording procedure. Both monetary deposits and continuous recording

increase the internal validity of the design but potentially may become
obtrusive and effective enough to dilute any differences between treat-
ment groups .

Another methodological issue is the testing of therapist and
therapist-treatment effects. These effects have been substantiated
(Ober, 1968; Marston & McFall, 1971). Another issue is the question of
whether or not the subject is performing his outside assignments (Lich-
tenstein, 1971; Marston & McFall, 1971). >0ne possible resolution to the
problem is the use of paid informants as illustrated by Azrin and Powell
(1968).

The use of control groups (Lichtenstein & Danaher, 1976) is another
concern. There are two opposing views to the use of types of control
groups. Bernstein (1969, 1970) argues for the use of several varieties
of no-treatment controls in order to analyze such non-specifics as ex-
pectation and self-report. Marston and McFall (1971) argue for the use
of only a minimal treatment condition as a control. Due to the homogenous
results of most studies and the small N in most treatment groups it is
recommended that smoking control experimenters employ only one minimal
treatment control (Lichtenstein & Danaher, 1971). Another methodological
issue is the differential time and type of experimenter or therapist
contact. Increasingly, more studies are using standardized treatments
and assistants to collect the data (e.g., Winett, 1973). All these
methodological concerns are increasingly becoming more recognized and
controlled in the smoking control research of the 1970's.

The third new line of research developing in the 1970's is the
testing and use of abstinence maintenance techniques. As the trend

‘mentioned earlier showed, the abstinence rate at the end of treatment is

impressively high regardless of type of treatment. What is very dis-
couraging is the relapse rate across studies where the percentage of
abstinent subjects who remain abstinent reduces continually downward

and asympotes at the 20-30% level at the three to six month follow-up
point. From this realization alone, the need for abstinence maintenance
procedures became evident. One approach to abstinence maintenance would
be the use of treatment beyond the reaching of abstinence. Winett (1973)
tested the use of an additional two week abstinence maintenance period.
The maintenance procedure was an extention of the contractual agreement
of either stopping or attending meetings. At the end of treatment there
were no differences between the two groups, nor were there any dif-
ferences in either reduction or quit rate at a three and six month
follow-up. Another type of abstinence maintenance procedure tested in
the literature is the use of frequent contact between the experimenter
or assistant and the subject for purposes of self report on smoking rate.
Bernstein (1970) tested the effect of high frequency and low frequency
contact across all the treatments he tested. High frequency was experi-
menter contact two to three times per week for 16 weeks and low frequency
was experimenter contact once a month for the four months. There were
no effects at the follow-up periods of frequency of contact reported by
Bernstein (1970). Other possible abstinence maintenance procedures in-
clude the use of sporadic booster sessions (Hunt & Matarazzo, 1973),

long term contractional deposits, buddy system, and social pressure
groups. While the initial research in this area has not proven fruitful,
a new line of smoking control research that uses certain techniques to
maintain abstinence is severely needed and can prove to be an encouraging

adventure.

The fourth new line of smoking control research deals with incor-
porating subject variables into the smoking Eherapy model. Although the
need for therapy outcome models incorporating subject and therapist
variables as well as treatment variables has been recognized (e.g.,
Eysenck, 1973), very little research in smoking control has been sensi-
tive to individual differences (Best & Steffy, 1975). Two leading
theorists in smoking motivation, Eysenck (1973) and Ikard and Tomkins
(1973) have argued for such tailoring of treatments to individual dif-
ferences, although they differ in their model and individual difference
dimensions. The strongest evaluative research in this line of smoking
control research dealing with individual differences has been carried
out by Best (1975) and Best and Steffy (1975). Using locus of control
as the individual difference dimensions Best and Steffy (1975) cite
literature which demonstrates that "internal LOC subjects perform re-
latively better in situations perceived as potential under their control,
whereas externals perform better in chance situations". Best and Steffy
(1975) hypothesized that internal LOC subjects would perform better if
they chose their own goals during the treatment phase and external LOC
subjects when the experimenter assigned the rate of reduction. The ex-
perimenters also used groups which were either employing a satiation
technique or an environmental control technique. The results showed
that there was no interaction between LOC and the effectiveness of
whether the subject or the experimenter set the reduction rates. There
was, though, an interaction between LOC and whether or not the group
used the satiation techniques or an environmental control technique.
Best and Steffy (1975) report that internal LOC subjects performed sig-
nificantly better than external LOC subjects in the satiation group, and

externals tended to do better in the environmental control techniques.

From this finding, Best and Steffy (1975) labeled the stimulus satiation
procedure as an internal treatment focus and the situational analysis of
external events as an external treatment focus. In a follow-up study
that attempted to validate the post hoc finding of the interaction be-
tween treatment focus and locus of control, Best (1975) found a signifi-
cant interaction between the locus of control measure and whether the
technique used was a stimulus satiation aversion procedure or a situa-
tional analysis of environmental events.

The Best (1975) study is a particularly noteworthy experiment. The
study tested three treatment factors that were crossed creating a total
of eight groups. The treatment factors employed occur in and relate
treatment to the natural environment and thus were expected to contribute
to the maintenance of abstinence. Also treatments were used which were,
considered to "foster attribution of behavior change to oneself". It is
argued that changes are thus more likely to endure than behavior change
attributed to external forces. Best (1975) included a number of person-
ality and motivational measures in his study which were hypothesized to
mediate the effectiveness of the treatment factors. The locus of
control-treatment focus was previously mentioned herein. Best (1975)
also found that the only personality or motivational measure signifi-
cantly related to success outcome was perceived probability of success.
This perceptual and motivational variable has consistently shown to be
a predictor of success outcome in smoking control studies (e.g., McFall
& Hammen, 1971; Best and Steffy, 1975). The two studies by Best (1975)
and Best and Steffy (1975) are excellent examples of the new lines of

smoking control research in the 1970's. The studies take advantage of

lO

non-specific variables such as commitment and motivation, emply more
rigorous and testable experimental designs, combine and cross treatment
factors, use techniques considered to enhance generalizability and the
maintenance of abstinence behavior, and test the moderating effects of
individual differences on the efficacy of treatment factors.

Commitment Treatments

 

Smoking control research has evolved from a general operant condi-
tioning theoretical framework. The result of such a perspective has
been an abundant testing of aversive therapy techniques in modifying
smoking behavior. The testing of these techniques and the results from
the studies, though, failed to substantiate the efficacy of these tech-
niques. The new directions that smoking control research has taken in
the past ten years have been in reaction to the lack of substantiating
results in the testing of operant conditioning techniques.

These new directions are intended to be a more thorough exploration
of processes taking place during quitting behavior. The search for non-
specific treatment variables, individual difference variables, and metho-
dological issues discussed earlier indicate a direction toward an under-
standing of the individual effort which takes place during purposeful
quitting. Parallel to this direction is the increase in the use of self-
management techniques. Essentially connotating self induced or applied
treatment, self management techniques have risen to great popularity and
are now tested and used more frequently than therapist administered
techniques.

Emerging from the research on the processes of quitting behavior
and from the use of self management techniques is the repeated finding

that two specific process variables relate to the successfulness of

11

the efforts to quit (Best & Steffy, 1975). The directed energies by an
individual towards the reaching of abstinence in cigarette smoking and
the confidence and expectancy of success in these energies are thus the
necessary variables which need to be induced and explored in smoking
control research.

Most reviewers of the literature on smoking research (Lichtenstein &
Danaher, 1976; Bernstein, 1969) have expressed a rather pessimistic out-
look on the behavioral control of cigarette smoking. This view is easy
to perceive given the lack of substantiated differential effects between
treatment variables as well as between certain controls and treatment
conditions. Given a change of perspective, though, the smoking control
literature in the last decade indicates a promising venture. What must
be taken advantage of is the newly acquired information on the effective
process variables.

A more optimistic viewpoint can be aroused if one perceives the
smoking control problem as one which entails the operationalization of
the effective process variables into treatment variables. Stated in
other words, the problem is one of creating and testing treatment vari-
ables which induce higher levels of motivation and effort as well as
increase the expectancy of success and at the same time are recognized
as legitimate or actual treatments by patient and therapist alike. Given
the knowledge acquired in the literature on the effective variables in
controlling smoking, the task is one of inducing them.

Prior to this operationalizing of effective variables, a conceptual
understanding of their impact is necessary. Since these effective pro-
cess variables are mostly attitudinal and/or cognitive variables, the
linkages between these variables and behavioral change such as cigarette

smoking should be explored in a cognitive-behavioral interaction model.

12

One conceptual approach to this relationship is that of consistency
models proposed by Festinger (1957). Briefly, the model states that
individuals attempt to keep consistency among their attitudes as well as
between their behaviors and attitudes. When inconsistency exists and
is felt by an individual, motivation is aroused to reduce the inconsis-
tency. The direction this motivation takes is toward changing the atti-
tudes or the behaviors to become more consistent with one another.

This model of both attitude and behavioral change can be used to
explain the effective process variables in the smoking control literature.
Given a high level expectancy of change in a behavior, such as smoking,
consistency would demand the alignment of this attitude with the actual
behavioral change. The effort or discomfort in the act of the behavioral
change would be offset by the reduction of tension felt in the inconsis-
tency between the expectancy of not smoking and the current smoking be-
havior. The behavior change becomes a source of comfort due to its
consistency with the expectancy. Likewise, individual effort and moti-
vation would act on the same principle. Given any kind or direction of
effort to quit smoking, tension would be constantly aroused between the
actual behavior of smoking and the manifested desire to quit the current
smoking behavior. The reduction of this tension then becomes a source
of comfort which balances out the discomfort inherent in the quitting
behavior. The effort to quit and the expectancy of quitting are thus
not the direct cause of quitting. Instead, the tension that they arouse
and its subsequent reduction are directly responsible for the greater
tendency of individuals to quit.

Presuming the above model to be correct, the problem becomes one

of influencing and inducing high levels of these process variables.

13

Just being in a controlled atmosphere (McFall & Hammen, 1971), self-
monitoring one's smoking behavior (Frederikson, Epstein, & Kosevsky, 1975),
and being told you are expected to quit (Bernstein, 1970) seems to be
enough to induce a feeling of inconsistency as evidence has shown in the
efficacy of certain control groups. These manipulations, though, are in
themselves not strong enough to demonstrate differential effects or over-
come the ever present high relapse rate experienced in the predominance
of studies. The search then becomes one of finding treatment variables
which will induce high levels of expectancy of quitting and effort to do
so.

Such a direction can be taken if we put the problem into the con-
sistency model framework. The problem becomes one of finding a behavior
or set of behaviors which are parallel and tied directly to the attitude
of high expectancy of success and the effort of quitting. Two such be-
haviors which are hypothesized to induce these behaviors are that of
announcement of quitting to others and to oneself. These behaviors,
though, must be both continual and actual. They must be believable and
realistic. Such announcements of quitting, if done continually and with
credibility would lead to an extreme high level of consistency with the
belief that one will actually quit and with the effort and motivation to
do 80.. Given the correctness of the above reasoning, the therapeutic
concern then becomes one of operationalizing these announcements of
quitting.

The first of these two types of interventions, announcement of
quitting to others, is taken in this study to be social commitment.
Commitment is taken to mean the "pledging or binding of the individual

to behavioral acts" (Kiesler & Sakumura, 1966, p. 349). Social

14

commitment in this study is the act of pledging to other significant
persons in one's social circle of acquaintances that one will carry out

a set of behavioral acts, namely, quitting smoking cigarettes. It is also
understood in this study that social commitment is a voluntary act.

It is the contention in this study that the immediate result of a
social commitment to quit smoking is an increased expectancy of quitting
and also an increase in energizing behavior or motivation to actually
quit. This.notion is born from the acceptance that "one infers one's
attitude from one's behavior" (Pollak & Cummings, 1976). The behavioral
commitment becomes a foundation for subsequent related perspectives.

"In effect, one's commitment may provide an anchor by which beliefs,
attitudes, and behavioral implications are marshalled or organized"
(Pollak & Cummings, 1976). The behavioral social commitment to stop
smoking becomes a salient factor around which subsequent beliefs,
attitudes, and behaviors towards the act of cigarette smoking become
strongly founded upon and hence consistent with.

The path of the process resulting from the therapeutic interven-
tion of social commitment is now traceable. Given the task of committing
oneself socially to the act of quitting smoking, and a resultant feeling
of ownership or responsibility for this commitment, and the actual doing
of the social commitment, it is expected that an individual's constella-
tion of related beliefs, attitudes, and behaviors will become consistent
with the commitment. The strongest of these are the beliefs that one
will be successful and actually quit, that one's attitude towards the
whole process of quitting and subsequently retaining abstinence will be
favorable, and that one's behavior will be energized toward the act of

quitting and once quit, retaining abstinence.

15

The effects of social variables on cigarette smoking behavior have
been explored somewhat in the smoking control literature. Some experi-
menter-therapists view social commitment as a form of social pressure
where the threat of withdrawal of social reinforcement is the explanation
for the effect of social commitment (e.g., Bernstein, 1970). These
type of studies situate the individaul in a group of strangers under con-
trolled conditions. The main problem with this approach to social comr
mitment is the use of non-significant others rather than close acquaint-
aces who are the greater source of the most potent reinforcers to an in-
dividual. Another problem is that the majority of cigarette smoking by
the individuals is outside of the group context. The same criticisms
can be applied to therapist-induced social pressure techniques (Bernstein,
1970; McFall & Hammen, 1971). An interesting variant on social pressure
and commitment type studies was a study done by Harris and Lichtenstein
(1971). The experimenters employed a 2 x 2 experimental design using
two social pressure treatment variables. One treatment variable was low
social vs. high social treatment where high social treatment consisted
of a friendly atmosphere between experimenter and subject, high expectancy
of success, and verbal reinforcement given to the subject by the experi-
menter. The other treatment variable was individual vs. group applica-
tion of techniques. The results of the study indicated that the low
social subjects were smoking a significantly greater percent of baseline
than the high social subjects (p < .001). Also there were indications
that the relapse rate for high social treatment groups was less than for
low social treatment groups. There were no significant differences be-
tween the individual and group conditions. This study, while increasing

the understanding of the relationship between social variables and

16

smoking control, does not indicate which of the manipulated social var-
iables were effective. Also, the previous criticisms concerning the
limited source and situation of the social pressure apply as well to this
study.

The manipulated social variable in this study is the act of a social
commitment to one's acquaintances outside the therapeutic environment.

The commitment of abstinence announced to those individuals psychologically
and socially close to the subject and in the social environment the
individual is operating in is expected to be stronger than in a controlled
and limited atmosphere. It is thus the first hypothesis that social
commitment to quit smoking will lead to abstinence. Furthermore, this
abstinence is hypothesized to endure more so than abstinence reached
without the social commitment.

The second treatment to be explored in this study is that of goal
setting. Goal setting is a self commitment that takes the form of be-
havioral or operant end points formulated and acted on in a continuous
process. Goal setting is identified as a conscious intention (Locke, 1968)
containing behavioral or operant guideposts. Like social commitment, goal
setting is an announcement of a behavioral end, but instead of the be-
havior being announced to others it is an announcement to oneself.

The process by which goal setting effects the outcome of quitting
behavior is similar to that of social commitment. The setting of goals
is a behavioral act by which related behaviors, attitudes, and expec-
tancies can be a source of dissonance. The actual setting of goals leads
to such consistent attitudes as greater willingness to quit and a dislike
of the act of smoking. Also, expectancies of success and favorableness

of not smoking tend to be adjusted in alignment with the behavior of

17

goal setting. These new behaviors, attitudes, and expectancies become
a strong cognitive pivotal point which are dissonant with the act of
smoking. Friction from this dissonance results in manifested tension in
the individual which if released becomes a source of positive comfort.
The reduction of this tension is what the individual becomes motivated
towards. But of strongest influence is that the tension reduction ex-
perienced from the new alignment of the behaviors, attitudes, and ex-
pectancies is a source of positive energy or comfort which offsets the
tension and discomfort experienced from the act of quitting smoking.
Thus, both the reaching and the maintenance of abstinence from smoking
should result from the act of setting goals to reduce and ultimately
discontinue smoking behavior.

Smoking control literature has explored the effects of explicitly
stated behavioral objectives or goals somewhat. Flaxman (1974) hypo-
thesized that abrupt cessation is necessary when an individual has grad-
ually reduced consumption to a point of 12 cigarettes. Although Flaxman
(1974) did not confirm his hypothesis he did conclude that setting a
specific time goal for quitting resulted in less relapse than immediate
quittong or no specified time goal. Another study which evaluated ex-
plicitly stated behavioral objectives was by Best (1975). Best (1975)
had recognized that the setting of behavioral objectives as stepping
stones toward the overall objective of abstinence may "foster attribu-
tion of behavior change to oneself." He further hypothesized that there
would be an interaction between personality type as defined by a locus
of control measure and whether or not the behavioral objectives were set

by the therapist or client. Best (1975) did not find any empirical

18

support for his hypothesis. He also found no differences between the

two groups. It is possible that the manipulation of client vs. therapist
goal setting was not strong enough. Best (1975) could have detected
greater client self-attribution of change by using the manipulation of

no goal setting vs. goal setting.

Although some of the studies on smoking control have suggested an
effect of individual goal setting, no experimenter has manipulated the
variable of goal setting using the two levels of individually set goals
in the form of gradual behavioral objectives vs. no individually set
goals. It is the intent of this study to do so. Thus, the second set
of hypotheses to be tested is that the setting of continuously reducing
goals will lead to abstinence more than the absence of these goals, and
also that abstinence reached by those having set these goals will be more
enduring than abstinence reached without the use of the goals.

In summary, the two treatment variables to be manipulated and ev-
aluated for smoking control efficacy, social commitment and goal setting,
are both behavioral commitments which are expected to induce strong ex-
pectancies and beliefs and high levels of motivation to cease smoking.
These induced attitudes and behaviors are expected, in combination with
the reaching and maintenance of abstinence, to counter the unpleasantness
and the tension accompanying the actual quitting of smoking. The hypo-
theses for this study are that both goal setting and social commitment
will bring about higher quit rates and maintain higher abstinence rates

in treatment conditions more than in conditions without them.

METHOD

Subjects

Subjects were recruited from introductory psychology courses, student
newspaper advertisement, and bulletin board notes at a large midwestern
university. Interested subjects initially met during an orientation
meeting where, among other agenda items, information was collected that
was used for later assignment to four groups. These four groups met
exactly one week from the orientation meeting. A total of 37 subjects
arrived at the orientation meeting. At the next set of meeting 29 per-
sons (15 males and 14 females) came. There were no dropouts for the re-
mainder of the study. One subject was randomly eliminated to create
equal numbers in the four groups.

Procedure

All subjects initially met in one of a number of orientation meetings
according to their sign up time, or their preferred time, or for the ex-
perimenter's purpose of equal cell sizes. All of the orientation groups,
and hence the subjects, were given the same meeting format.

At the orientation meeting the subjects were told that the study
was intended to help people quit smoking by using graduate withdrawal
techniques. Subjects were informed that they will be taught these tech-
niques in exactly one week from the current day and time.

The subjects were then told the techniques were meant to be used on
the subjects' own time. They were told the purpose of these meetings

was to exchange information only. Subjects were told that the length of

19

20

the study was exactly 14 days long. They were notified that the first

week, from the first meeting to the next meeting (which was exactly one
week from that day), was their baseline week. They were told the base-
line week will be shortly explained. In the next six days after their

baseline week they will self administer the techniques.

Subjects were also reminded of the five dollar fee for the study.
The fee was said to be a symbol of the subjects' commitment to the study
and to their quitting Cigarettes.

Subjects were then handed a folder containing two consent forms,

a questionnaire and answer sheet, and instructions on how to compute
their baseline. Subjects were requested to fill out the standard sub-
ject consent form. They then filled out a more extensive consent form.
They were requested to initial a number of statements about the fee and
the confidentiality of the study.

All subjects were instructed on two methods of counting their
smoking rate. One method was called continuous monitoring. The method
entailed the counting of each cigarette smoked by marking the day and
time in a paper paid at the time of lighting the cigarette. The second
method was called transaction monitoring (see Appendix A). Essentially,
the method entailed the monitoring of the number of cigarettes that come
into and out of an individual's possession and the date of these trans-
actions. Examples of incoming transactions were: buying, borrowing,
finding, and the current possession of cigarettes. Examples of outgoing
cigarettes were losing, loaning, destroying, and number in possession
at end of duration. The difference between incoming and outgoing cig-
arettes was an indirect measure of cigarettes smoked. Both methods were

employed during the baseline and the treatment week and then only

21

transaction monitoring was used subsequently.

All the subjects were reminded of the $5 fee that was to be collected
at the next meeting. They were told to think about whether they really
wanted to quit enough to invest their time, energy, and money into this
project between these two meetings. A questionnaire was then filled out
by the subjects.

Starting two days prior to the second meeting, the admission meeting,
subjects were Contacted by phone about their new time (if any) for the
admission meeting.

One week after the orientation meeting, four groups met separately
for the admission meeting where the directions for the techniques were
given. Before written instructions were handed out, all subjects were
given a standardized lecturette. The lecture states some basic facts on
cigarette smoking directed not at the harmful effects of smoking, but at
the reasons why people quit. The lecture's theme was that stopping
smoking is easy but staying off of cigarettes is hard. They were told
that 80% of the people who do quit, start smoking again within three
months after the time they stopped. They were also told that the tech-
niques given to them were not just meant to get them to quit but that
they were meant to help people stay abstinent once they do quit. The
subjects were all told that the plan (pseudoname given to all groups)
was set up so they will quit somewhere between the fifth and sixth day.
They were told that if they do wish to quit before the fifth day to go
ahead with it and that they would be still a part of the study.

Subjects were then handed out one of four instruction manuals. The
instruction manuals corresponded to the four treatment conditions de-

scribed below and in Appendix B and C. These instructions were read

22

aloud and questions were urged. At the end of the reading, subjects were
reminded of the purpose of the study, their commitment, and the next
meeting in exactly seven days.

Design and Treatments

 

The experimental design as a 2 x 2 x 5 fixed effects design with re-
peated measures on the last factor. The dependent variable was measured
at the following times: at the end of the first week (baseline), at the
end of the treatment phase (one week), and again at one week, one month,
and a three month post treatment. Two dependent variables were measured,
one was the rate of cigarette smoking per day and the other was whether
or not the subject remained abstinent from smoking.

Goal Settinngreatment

 

Those persons who set goals daily set their goals for the next day's
number of cigarettes smoked for each day of the week. 'They charted and
graphed these goals on the Progress Chart along with the number of cig-
arettes smoked on the previous day. Those who did not set daily goals
only charted and graphed the number of cigarettes they smoked for the
previous day. See Appendix C for further details on the directions for
goal setting.

Social Commitment Manipulation

 

The second variable, which was crossed with the goal setting var-
iable, was whether or not a person announced publicly his/her quitting
or having quit cigarettes. Operationally, those in the social commit-
ment group received a set of buttons and instructions, encouragement,
and signed commitments to wear the buttons. One button read: "PRAISE
ME, I'M QUITTING," and the other, "PRAISE ME, I'VE QUIT." They were

instructed to wear the first button during the time they were reducing

23

their cigarette consumption and to wear the second button at least up

to two months after they had quit. Subjects in both conditions (buttons

or no buttons) were told that they were not "non-smokers" until at least
three months after they had quit. They were all reminded that 80% of those
persons who quit return to smoking within three months.

The two independent variables were crossed creating four treatment
groups: one group had neither goal setting nor social commitment, one
group had goal setting, another had social commitment, and the fourth
group had both goal setting and social commitment. All groups used the
transaction monitoring procedure described earlier and kept a physical
record of their smoking rate each day on the progress chart, which simply
was a line graph of their smoking rates. All groups were instructed on
an additional technique to help them quit. The technique is called
hierarchical reduction (Manston & McFall,1971). The technique entails
a listing of the most difficult prompts at which to quit smoking. A
prompt is described as a person, place, or event that precedes the
smoking of cigarettes. Examples of prompts given included coffee
drinking, driving a car, and attending a party. The instructions given
were to reduce cigarette smoking in two prompts each day (see Appendix
B). Thus, all groups used the transaction monitoring procedure, the
progress chart, and the hierarchical reduction technique. The groups,
though, differed on the two manipulated treatment variables of goal
setting and social commitment.

Subjects were all contacted by phone the day prior to their next
meeting which was one week after their admission meeting. Commitment to
attend the meeting was obtained over the phone whether or not they had

stopped or reduced their smoking. All subjects arrived at their

24

respective meetings. Data on smoking rate was collected, a standardized
lecturette was given to all groups similar to the one given at the
orientation meeting, and written remarks were requested and collected
pertaining to the experience of the individual during the past week.
Also, a questionnaire was given out containing items intended to check

on the carrying out of the instructions. The response to these items
were used as manipulation checks on the independent variables as reported
in the results section. Subjects were thanked for their involvement,
reminded about the follow-up telephone calls, and told that although

the groups were not going to meet again, the study would continue for

the next three months. Subjects who did not quit at this time were told
to continue their treatments until they did so. Subjects were also asked
if they expected to move within the next three months and if so, new
addresses were needed.

Follow-up telephone calls were given at the one week, one month, and
three month treatment termination periods for the purpose of collecting
data on the dependent variables. (No attempt was made to verify these
self reports.) Since the three month termination period was still in the
academic year (Spring) and address changes were acquired there was no

trouble in collecting all the subjects' data.

RESULTS

Dependent Variable Reliability

The dependent variable used in this study was the number of cigarettes
smoked per day per person. The dependent variable was measured by two
self reporting procedures. As described in the previous section, the two
procedures are a continuous counting procedure and a transaction counting
procedure. Both methods were used during the baseline and treatment
periods, then only transaction counting was employed for the remainder
of the study.

A Pearson product moment correlation was calculated between the
alternative means of measurement to ascertain the degree of reliability
in the measurements. A correlation of .92 was obtained indicating a
very strong relationship and hence consistency between the two measure-
ment techniques. Although the data used for this correlation was obtained
only during the baseline and treatment period, it is assumed that the
degree of correlation would have been upheld during the remainder of the
study had both methods been subsequently employed.

Manipulation Checks

 

To determine whether the actual manipulated independent variables
were performed by the subjects and led to their immediate intentions a
follow-up questionnaire was employed on the last day of treatment. To
determine whether goals were actually set and used by those subjects in
goal setting groups three goal setting questions were asked of all sub-

jects. The participants were asked if they set goals and recorded them,

25

26

the degree of their efforts to reach them, and the degree of awareness

of the goals in their efforts to reduce and ultimately quit cigarettes.
To check on the goal setting manipulation, a'2 x 2 analysis of variance
was performed with goal setting and social praise as the independent
variables and the subject's mean response for the three goals setting
manipulation check items as the dependent variable. The only significant
effect on this dependent variable was that of goal setting (p < .05).
This result indicates that those individuals in the goal setting group
had set goals and consciously directed themselves toward these goals.

The second independent variable, social commitment, was checked by
the use of two items in the questionnaire used at treatment termination.
One item asked for the respondent's percentage of friends who knew of the
person's immediate intentions of quitting. The second question asked
for the percentage of these people who knew of these intentions who had
also praised or made positive remarks about the subject's intentions of
quitting cigarette smoking. A 2 x 2 analysis of variance was carried
out where the independent variables were the two treatment variables and
the dependent variable was the product of the above two items. The
dependent variable was the product of the above two items. The dependent
variable was thus the percentage of one's friends who praised the indi-
vidual for the effort of quitting smoking cigarettes. Both independent
variables were statistically significant with goal setting having an
approximate F probability of .018 and social praise an approximate F
probability of .036. The interaction was not significant. The results
demonstrate that those individuals in either the goal setting group or
the social commitment group or the combination group had significantly
a greater percentage of friends praising their efforts than those in the

group with neither of the manipulated treatment variables.

27

Treatment Effects

An initial regression analysis of baseline rate on rate of smoking
at post-treatment and followbup was performed to determine if baseline
would be worth controlling for in further analysis. The regression co-
efficient was significant at less than .005. The statistically signifi-
cant regression effect of baseline on later smoking rate indicated a
usefulness of an analysis of covariance with baseline as the covariate.

Before an analysis of covariance could be performed, the assumption
that there was no covariate-factor interaction was tested. Essentially,
the assumption is that the regression for baseline on later smoking rate
is homogeneous across treatment groups. The assumption was tested by
computing Box's M statistic. The statistic value was non-significant,
indicating the appropriateness of analysis of covariance.

Table 1 presents the analysis of covariance table with baseline as

the covariate.

Table 1

Analysis of Covariance Table

 

 

Adjusted
_ . Adjusted , Mean

Source SS DF Square F .p
Social 1399.4 1399.4 11.12 .003
Goal 240.9 1 240.9~ 1.9 .180
Social x Goal 89.6 ' 1 89.6 .71 .407

Error I 2893.6 23 125.8
Time ' 945.2 3 315.1 8.98 .000
Social.x Time 174.8 3 58.3 1.66 .183
G051 x Time 70.4 3 23.5 .67 .574
Social x Goal x Time 262.5 3 87.5 2.5 .067

Error II 2490.4 71 . 35.1 f

28

The main effect for time was significant indicating that there were dif-
ferences between time periods on the dependent variables across the treat-
ment groups. The only other significant main effect was that of social
commitment which was significant at the .01 level. Before any interpre;
tation can be given to the above significant main effects, the marginally
significant three way interaction of goal setting-social commitment-and
time (p_< .10) must be explored.

Figure 1 shows the mean number of cigarettes smoked for each of the
four groups at the five time periods. The trend seems to be that rate
of smoking was extremely reduced at the end of treatment and then a re-
lapse or an increase in smoking rate took place across all the groups.
This rate or rescidivism, though, was seemingly differential. Separate
simple effects analysis at each time period using analysis of covariance
with the two treatment variables as independent variable was undertaken
to explain the three-way interaction between goal setting-social commit-
ment-and time. At the end of treatment the simple effect of goal setting
was significant at the .05 level. At one week follow-up the simple effect
of social commitment was significant at the .01 level. Simple effects
analysis at the one month follow-up resulted in a significant effect of
social commitment at the .01 level. There were no significant effects
at the three month follow-up.

Post hoc Tukcy Tests on the adjusted means were performed to
determine if there were any significant differences between the groups
at each time period other than those indicated by the simple effects.
At the end of treatment the combination of groups receiving the treatment
variable compared to the control group was statistically significant at
the .01 level. Also, at the end of treatment each of the goal setting

groups had significantly lower smoking rates than the control group.

29

At one week follow-up the group with both social commitment and goal
setting was significantly different from the goal setting group and from
the control group. At one month followeup, the social commitment group
was significantly different from the goal setting group and from the
control group. These individual comparisons among adjusted group means
reinforces the notion that at the end of treatment the goal setting and
commitment groups were not differentially effective but were significantly
more effective than the control. In addition, the comparisons illustrate
that the social commitment groups had a more lasting effect at one week
and one month follow-up than did the goal setting and control groups.

An argument can be made that the dependent variable behaves as a
dichotomous variable. That is, that an individual was considered to be
smoking or not to be smoking. One source of data that supports this was
the smoking rates for each of the subjects. 0f the 12 subjects who quit
at one point and started smoking again within the three month study
period, 11 went back to smoking greater than 80% of their baseline rate.
Table 2 presents the number of persons abstinent in each group at each

period of time.

Table 2

Number of Persons Abstinent

Time Period

 

TREATMENT ONE WEEK ONE MONTH THREE MONTH

 

 

 

.QRQHP BASELINE TERMINATION FOLLOW-UP FOLLOW-UP FOLLOW-UP
Control 0 1 0 0 0
Goal & Social 0 4 5 2 3
Goal 0 4 1 2 3

Social 0 3 4 4 3

30

A x2 test was performed on a 2x2 matrix with one dimension being
abstinent versus non-abstinent and the second dimension being the four
groups at the three month follow-up. The x2 was non-significant at the
.05 level. 'A second x2 was performed collapsing the treatment groups
into one level. The proportion of persons abstinent at the three month
follOwbup in the treatment groups was significantly different than the

proportion of persons abstinent in the control group.

DISCUSSION

Two of the major difficulties in cigarette smoking research were
handled successfully in this study. The first is the unreliability of
self monitoring of smoking rate (Frederikstein, Epstein, & Kosevsky,
1975) and the second is the often high drop-out rate (Mees, 1966). In
this study there were no drop-outs after the baseline period. As to the
former problem, the solution seems to have been the combination of the
use of two monitoring procedures with one of them being the transaction
monitoring procedure. This study was the first to have used the trans-
action monitoring procedure and to have assessed its reliability. The
successfulness of the method was based on self reports by subjects that
the method had high face validity and was extremely easy to do. Also,
the reliability was probably increased due to the use of two monitoring
procedures used simultaneously. Subjects were probably aware that the
reliability of their reports would be assessed due to the use of two
procedures. Falsification, though, was still a possibility. Further use
and evaluation of the transaction monitoring procedure is necessary.

As to the second common problem in smoking control research, that of
high drop-out rate, the solution was again a divergence from traditional
methods employed in smoking research. A method that has been tested and
found to not only reduce drop-out rate but also effect smoking outcome
(Elliot & Tighe, 1968) is the use of security deposits ranging from $10
to $60 where the money is returned upon completion of the training program.

This method may be somewhat effective in reducing drop-out rates but could

31

 

32

be a severe problem with recruiting subjects, especially in a university
setting. Also, disputes over refunds could take place. An alternative
method of reducing drop-out was employed in this study and was seemingly
succesSful. A flat 5 dollar fee for participation was administratively
simple and effective. It was expected that a nominal fee such as 5
dollars would result in a greater feeling of choice and ownership in the
participation of the study. The fee acted as a behavioral commitment to
participate in the study. Further research comparing fees with deposits
and controls should be explored.

Another methodological procedure which had positive results was the
use of a shorter one week time period for treatment. The usual treatment
length is four weeks or greater (Bernstein, 1969). Since the average
abstinence rate at a three month follow-up period is in the 20-30% range
for all reported smoking control studies, and in this study, 9 out of
28 subjects (32%) were abstinent at the three month follow-up it seems
that the short treatment length did not compromise the treatment effective-
ness. The short treatment period may also account for the low drop-out
rate.

The validity checks on the treatment variables positively support
the idea that the directions for carrying out the treatment manipulations
were actually carried out by the subjects as intended. An alternative
explanation, though, may be that there was faking on the post-treatment
follow-up questionnaire, especially for the goal-setting group. It is
possible that individuals who were instructed to set goals, but did not,
stated they did for socially desirable reasons. Interesting results
appeared for the validity check on social commitment. Both independent

variables of goal-setting and social commitment were statistically

 

33

significant when per cent of total friends who praised one's efforts to
quit smoking was used as the dependent variable. It would seem that both
the directions to tell one's friends one is quitting and the self setting
of the behavioral commitments (goal-setting) leads to the notifying of
acquaintances and a positive response from them. If percentage of
friends who respond favorably to one's quitting smoking is the effective
variable than the two treatment variables used in this study share the
common variance of the effective variable. If such is the case, then
differences between the effects of the treatment variables would be
diluted.

An indication that percentage of friends who respond favorably to
one's quitting smoking is related to the successfulness of reaching and
maintaining abstinence is the fact the correlation between per cent of
baseline rate at three months and per cent of friends responding
favorably is -.559, which is statistically significant (pm: .05).

As is common in the preponderance of smoking control studies
(Bernstein, 1969) this study demonstrated that participation in a smoking
control study does lead to reduction or abstinence in the short run. At
treatment termination 12 of 28 (43%) persons were abstinent. But, as
is common in other smoking control studies, the rate of rescidivism was
high. As indicated by the marginally significant interaction of Social
Commitment x Goal Setting x Time, this rate of rescidivism may be dif-
ferent for the different treatment groups. Figure 1 illustrates graph-
ically these rates of rescidivism. The control group is the only group
to relapse back to the baseline average and did so by the one month
follow-up. From the simple effects analysis the three-way interaction

can be explained by the significant simple effect of goal-setting at

22

20

19

16

14

12

10

 

 

l l I z I
I I , I I ~ I
Baseline End of 1 Week 1 Month 3 Month
Treatment Follow-up Follow-up Follow—up

No Goals, No Social Commitment

 

-------- Goals, No Social Commitment

. . . . No Goals, Social Commitment
-.-.-. Goals, Social Commitment
Figure 1

Adjusted Group Mean Cigarettes Smoked Per Day

 

35

treatment termination and from the significant simple effect of social
commitment at one week and one month follow-up. The goal setting effect
can be concluded to be more immediate and more temporary than the social
commitment effect. As was noted by the validity check on the manipulation,
both treatments led to increased support from acquaintances. This sup-
port from acquaintances is positively and strongly related to rate of
smoking at the three month follow-up. These results indicate that goal
setting may effect not only amount of social support but also another
process that leads to more immediate effects but wears away quickly
and may even increase the rate of rescidivism. For example, goal setting
may lead to a strong commitment to stop at a given time (e.g. treatment
termination) but once an individual does reach the goal the motivation
to stay abstinent is reduced. The effect of the social commitment on
the other hand, can be explained by the singular process of social sup-
port which would directly relate to probability of successful abstinence.

Although the results of this study do not show any clear and
conclusive support for the initial hypotheses there is some evidence for
their credence. The significant main effect for social commitment shows
that for the average across time, the groups with social commitment
smoked less cigarettes than those groups without social commitment.
Since the groups were homogeneous at baseline, from end of treatment on,
the social commitment groups averaged less cigarettes consumed. This
result suggests that social commitment did have a directional effect on
the reading of abstinence and the rate of rescidivism.

From the manipulation checks described and analyzed earlier, it is
apparent that both the social commitment and the goal setting treatment

led to a high percentage of social acquaintances knowledgeable and

 

36

positive about the subject's effort to quit smoking. Support for this
social variable is held in the significant correlation (-.559) between
per cent of friends responding favorable and per cent of baseline at the
three month follow-up. If percentage of friends with positive attitudes
about the quitting behavior is the effective variable then further
evidence would be that any of the groups except the control should exhibit
effectiveness. Using the Tukey’ test there was a statistically signifi-
cant differences between the control group and the combination of the
other three groups at only treatment termination. In addition, the
control did distinguish itself from the combination of other groups at
three month follow~up on number of persons abstinent. This latter result
reflected the fact that the control group had no persons abstinent at the
end of the three month follow-up period and the remainder of the treatment
groups had each three persons abstinent at this late follow-up period.
These results lead to the conclusion that the treatment groups did have

a positive effect above that of the control group. It also seems that

the effect of treatment groups was through the process of developing
social support and that this support increased the probability of success-
ful abstinence.

The above remarks suggest that there may be one effective process
variable running through the non-control groups: that of a behavioral
commitment publicly announced and supported. Strong reservations must
be made for this conclusion for there may be certain contaminating
effects in this study. The first possible outside variable which may
alternatively explain the trend difference between the control group and
the remainder of the treatment groups is order effects. Although, the

order of the meeting times of the groups was randomly assigned, the

37

control group was the first group to have met. It was possible that the
quality of experimenter interaction increased after exposure with the
control and improved in the remainder of the three groups. If experi-
menter-subject rapport or interaction had an effect on outcome then the
order effect may be contaminating. Another possible explanation for the
low performance of the control group is that of experimenter bias. The
same experimenter was used for each group and was aware of the hypotheses.
It is possible that the congrol group was adversely influenced through
suggestion or expectation unconsciously communicated by the experimenter.
Thus the control may be distinguished from the other three groups not
from the manipulated treatment variables but from some order or experi-
menter effects.

This study was highly successful in terms of methodology and admin-
istrative innovations but lacked strong statistical evidence for support
of the efficacy of the treatment variables. Trend differences to exist
but are subject to alternative explanations. This line of research in
behavioral commitment, both in the form of social announcement and goal
setting, is in need of further research with emphasis on subject variable
moderators and measurement of the effective process variables. Further
investigation of social support is warranted. Much larger sample sizes

should be employed to determine clear differences and effects.

APPENDICES

 

APPENDIX A

DIRECTIONS FOR TRANSACTION MONITORING

 

To Use The Record Book: The number of cigarettes smoked is

counted by logging in your transactions with cigarettes. It is

important that you keep an accurate count of cigarettes smoked.

The Mechanics:

 

During the week do the following:

1)

2)

3)

4)

Put the dates for the week at the top of the page.

Under the column "day" write the first day's date. Every time

you have a transaction with cigarettes log it in. Transactions

are the number of cigarettes that come into your disposal and

out of your disposal. Cigarettes are not directly counted by

when you smoke them. There are five major transactions:

Cigarettes that are

+a) started out with at beginning of week (counted at DE-LITE
TIME).

+b) bought

+c) borrowed

-d) loaned

-e) left over at end of the week (at next DE-LITE TIME).

There may be other transactions such as lost or broken cig-

arettes.

Write each transaction at the time they happen. THIS MEANS

KEEPING THE RECORD BOOK WITH YOU AT ALL TIMES.

38

 

5)

6)

7)

39

Each day of the week write in all transactions.

At your next DE-LITE TIME add all the pluses and subtract the
minuses to get a total for the week.

The number of cigarettes you have left over at the end of the

week will be the number you start the next week with.

 

 

APPENDIX B

DIRECTIONS FOR PROMPTS CONTROL

Mechanics:

1)

2)

3)

4)

5)

6)

On a separate sheet of paper list as many prompts as possible that
you feel are influential in your smoking. List at least fifteen.
You may want to "break apart" some of these prompts to make them
specific. For example, "at work" can be broken down into: "while
working at desk," "at coffee breaks," "during lunch," etc. Below

is a sample list which you may decide to choose from.

after eating walking watching TV
with coffee working at desk break from work
driving car ' drinking (alcohol) house cleaning
talking drinking (non-alcohol) doing hobby
reading after physical activity while waiting
while writing at parties in bed

after waking up listening to music‘

After making your list, choose ten of these prompts which you feel
would be difficult to stop smoking in and that you do frequently.
List these ten prompts in order of difficulty from 1 (the most
difficult) to 10 (the least difficult).

Write these prompts in order, from top to bottom, on the PROMPTS
TABLE.

The first day of the SMOKERS DE-LITE PLAN do pp£_smoke in the 9th
or 10th prompt categories.

The second day do not smoke at all in the 7th, 8th, 9th, or 10th

prompt categories.

40

7)

8)

9)

41

Each successive day limit your smoking by the next two higher prompt
categories.

By the end of the 5th day you should not be smoking in any of these
top ten prompt categories.

Use the PROMPT TABLE by Xing out the prompts in the appropriate

weeks.

 

 

BIBLIOGRAPHY

BIBLIOGRAPHY

Azrin, H.H., & Powell, J. Behavioral engineering: The reduction of
smoking behavior by a conditioning apparatus and procedure. Journal
‘2; Applied Behavior Analysis, 1968, 1, 193-200.

Bernstein, D.A. The modification of smoking behavior: An evaluative
review. Psychological Bulletin, 1969, 11, 418-440.

Bernstein, D.A. The modification of smoking behavior: A search for
effective variables. Behavior Research and Therapy, 1970, 8, 133-146.

Best, J.A. Tailoring smoking withdrawal procedures to personality and

motivational differences. Journal pf Consulting and Clinical

 

Psychology, 1975,_43, 1-8.

 

Best, J.A., & Steffy, R.A. Smoking modification procedures for internal

and external locus of control clients. Canadian Journal pf Behavioral

 

 

Science, 1975, 1, 155-165.
Elliott, R., & Tighe, T. Breaking the cigarette habit: Effects of a

technique involving threatened loss of money. Psychological Record,

 

1968, 18, 503-513.
Eysenck, H.J. Personality and the maintenance of the smoking habit. In

W.L. Dunn, Jr. (Ed.), Smoking Behavior: Motives and Incentives.

 

 

Washington, D.C.: Winston & Sons, 1973, 113-146.

Festinger, L. ‘A_theo§y pf cognitive dissonance. Evanston, 111.: Row

 

Peterson, 1957.
Flaxman, J. Smoking cessation: Gradual vs. abrupt quitting. Paper

presented at the meeting of the Association for Advancement of

4.2

43

of Behavior Therapy, Chicago, October 1974.
Frederiksen, L.W., Epstein, L.H., & Kosevsky, B.P. Reliability and
controlling effects of three procedures for self-monitoring smoking.

Psychological Record, 1975, 25, 255-264.

 

Harris, D.E., & Lichtenstein, E. Contribution of nonspecific social
variables to a successful, behavioral treatment of smoking.
Paper presented at the meeting of the western Psychological
Association, San Francisco, April 1971.

Hunt, W.A., & Bespalec, D.A. An evaluation of current methods of modi-

fying smoking behavior. Journal pf Clinical Psychology, 1974, 30,

 

431-438.

Hunt, W.A., & Matarazzo, J.D. Habit mechanisms in smoking. In W.A.
Hunt (Ed.), Learning Mechanisms in Smoking. Chicago: Aldine,
1970, 65-102.

Ikard, F.F., & Tomkins, S. The experience of affect as a determinant
of smoking behavior: A series of validity studies. Journal pf

Abnormal Psychology, 1973, 81, 172-181.

 

Kiesler, C.A., & Sakamura, J. A test of a model for commitment. Journal

‘2: Personality and Social Psychology, 1966, 3, 349-353.

 

Levinson, B.L., Shapiro, D., Schwartz, G.E., & Tursky, B. Smoking

elimination by gradual reduction. Behavior Therapy, 1971, 2, 447-487.

 

Lichtenstein, E. Modification of smoking behavior: Good designs--in—

effective treatments. Journal pf_Consu1ting and Clinical Psychology,

 

1971, 39, 163-166.
Locke, E.A. Toward a theory of task motivation and incentives. Organi-

zational Behavior and Human Performance, 1968, 3, 157-189.

 

Logan, F.A. The smoking habit. In W.A. Hunt (Ed.), Learning Mechanisms

 

44

ip_Smoking. Chicago: Aldine, 1970, 131-145.

Logan, F.A. Self-control as habit, drive, and incentive. Journal pf
Abnormal Psychology, 1973, 81, 127-136.

Marston, R.M., & McFall, R.M. Comparison of behavior modification

approaches to smoking reduction. Journal pf Consulting and Clinical

 

Psychology, l971,_§§, 153-162.

 

McFall, R.M., & Hammen, C.L. Motivation, structure, and self-monitoring:
Role of nonspecific factors in smoking reduction. Journal pf
Consulting and Clinical Psychology, 1971, 31, 80-86.

Mees, H.L. Placebo effects in aversive control: A preliminary report.
Paper presented at the joint meeting of the washington State-Oregon
Psychological Association, Ocean Shores, wash., May 1966.

Ober, D.C. Modification of smoking behavior. Journal pf_Clinical and

 

ConsultingPsychology, 1967, 31, 104-118.

 

Pallak, M.S., & Cummings, W. Commitment and voluntary energy conserva-
tion. Personality and Social Psychology Bulletin, 1976, 2, 27-30.
Premack, D. Mechanisms of self-control. In W.A. Hunt (Ed.), Learning

Mechanisms ip Smoking. Chicago: Aldine, 1970, 107-123.

 

Powell, J.R., & Azrin, N. The effects of shock as a punisher for

cigarette smoking. Journal pf_Applied Behavior Analysis, 1968, l,

 

63-71.

Thoresen, C.E., & Mahoney, M.J. Behavioral self-control. New York:

 

Holt, 1974.
Winett, R.A. Parameters of deposit contracts in the modification of

smoking. Psychological Record, 1973, 23, 49-60.

 

Yates, A.J. Theory and practice ip_behavior therapy. New York: Wiley,

 

 

1975.

"‘IIIIIIIIIIIIIIIII