Nova Southeastern University

NSUWorks
College of Psychology: Faculty Articles College of Psychology

11-1-1993

What's In A Label? The Effects of Substance Types

and Labels on Treatment Considerations and
Stigma
John A. Cunningham
University of Toronto

Linda C. Sobell
Nova Southeastern University, sobelll@nova.edu

Virginia M. Chow

Follow this and additional works at: http://nsuworks.nova.edu/cps_facarticles

Part of the Psychology Commons

NSUWorks Citation
Cunningham, J. A., Sobell, L. C., Chow, V. M. (1993). What's In A Label? The Effects of Substance Types and Labels on Treatment
Considerations and Stigma. Journal of Studies on Alcohol, 54(6), 693-699.
Available at: http://nsuworks.nova.edu/cps_facarticles/798

This Article is brought to you for free and open access by the College of Psychology at NSUWorks. It has been accepted for inclusion in College of
Psychology: Faculty Articles by an authorized administrator of NSUWorks. For more information, please contact nsuworks@nova.edu.
What's in a Label? The Effects of SubstanceTypes and
Labelson TreatmentConsiderations
andStigma*
JOHNA. CUNNINGHAM,M.A.,LINDA C. SOBELL,PH.D.,'tANDVIRGINIA M.C. CHOW,B.SC.*
AddictionResearchFoundation,Toronto& Departmentof Psychology,Universityof Toronto,Toronto,Ontario, Canada
ABSTRACT. Visitors (N = 579) to a science center read selected
scenariosand evaluatedthe most likely outcomefor a hypothetical
substanceabuser.Respondents were randomlyassignedto one of six
scenarioconditions:a personwith one of three different substance
abuseproblems(alcohol,tobacco,or cocaine)was crossedwith two
labelsreflectinghigh or low substancedependence.Resultsindicated
that: (1) cigaretteswere viewed as a less serioussubstanceabuse
TEREOTYPICAL
BELIEFS
about
substance
abusers
are common. Dean and Poremba (1983) found that
three-quartersof the words used by respondentsto de-
scribe an "alcoholic" reflected the image of a skid row
bum. The impression of a "drug addict" is similarlyneg-
ative (Dean and Rud, 1984).
Stereotypicalimpressionsmay also affect interactions
with substanceabusers.Tootle (1987) found that recov-
ered "alcoholics" were not fully sociallyacceptedin the
work place. Ericksonand Goodstadt(1979) reportedthat
thosewith a supposedgreatercriminal record for mari-
juana possessionreceivedfewer postitiveresponses from
potentialemployers.Sobelland Sobell(1975) foundthat a
harsherpenalty was more often recommendedwhen a
crime suspectwas said to have been drunk and labeledan
alcoholic than when not so labeled.
It has also been postulatedthat the stigma associated
with beinglabeledan alcoholiccanbe a reasonfor problem
drinkers not wanting to enter treatment (Roizen, 1977;
Tuchfeldet al., 1976). In supportof this, a recent study
Received:June23, 1992. Revision:August 17, 1992.
*The views expressedin this article are thoseof the authorsand do
not necessarilyreflect thoseof the AddictionResearchFoundation.This
study was supported,in part, by a studentship from the Medical Re-
searchCouncil, Canada. Portionsof the article were presentedat the
25th annualmeetingof the Associationfor Advancementof Behavior
Therapy,New York City, November1991.
*LindaC. SobellandVirginaM.C. Choware withtheAddiction
Research Foundation, Toronto. Linda C. Sobell is also associated with
the Departmentsof Psychologyand BehaviouralScience,Universityof
Toronto.
Reprint requestsshouldbe sentto JohnA. Cunningham,M.A., Ad-
diction Research Foundation, 33 Russell Street, Toronto, Ontario, M5S
2S1, Canada.
693
problemthan were alcoholor cocaine(a personwho smokedciga-
retteswas rated as more likely to recoverfrom his problem, self-
changewas regardedas more appropriateand less stigma was
associatedwith smokingthan with the other two drugs); (2) non-
abstinentrecoveriesof all types were greetedwith skepticism;and
(3) recoverywas rated as more likely to occur from treatmentthan
from self-change.(J. Stud.Alcohol54: 693-699, 1993)
of personswho had resolvedan alcoholproblemwithout
treatmentfound that 40% reportedthey had not sought
treatmentbecauseof the stigmaof being labeled an alco-
holic (Sobellet al., 1992). Problemswith identifyingwith
the stereotypeof an alcoholichave also led somepatients
to delay seekingtreatment(Thom, 1986). Finally, 45% of
peoplein a generalpopulationsurveybelievedthat alco-
hol treatmentshouldbe conductedaway from a person's
home to protecttheir privacy (Mulford and Miller, 1961).
Stafford and Petway (1977) found that different labels
may evokedifferentlevelsof stigma.Using semanticdif-
ferentialscales,respondents rated descriptionsof an indi-
vidual labeled as an alcoholic, a drunk, or an unlabeled
target. The alcoholicwas rated as significantlyless reli-
able, more dishonestand more sick than the drunk target,
and both the alcoholicand the drunk were rated as signif-
icantly more sour,dishonest,bad, immoral, slow, weak,
selfish, hopeless,less respectableand less responsible
than the unlabeled target. Similarly, Wallston and col-
leagues(1976) found that a patient admittedto hospital
with a bleedingulcer who was describedas alcoholicwas
rated by nursesless favorablythan the same patient not
so described.These labelingeffectshave also been found
to interactwith aspectsof the target (e.g., employment
status)as well as the respondentpopulation(e.g., urban,
rural, studentor alcoholic)(Kilty, 1981; Kilty and Meen-
aghan, 1977). Stigmatizationis also reportedwith drug
abuse(Dean and Rud, 1984) and cigarettesmoking(Coo-
per and Kohn, 1989; Dion et al., 1990; Goldstein, 1991).
A problemwith the labelingresearchto date is that the
amountof reportedsubstance usehas not beencontrolled.
This is importantas differentlevelsof drinking are asso-
ciatedwith differentlabels(e.g., socialdrinker,problem
drinker, alcoholic)(McKirnan, 1977). In one studywhich
694 JOURNAL OF STUDIES
controlledfor amount, pattern and results of drinking,
targetslabeledas alcoholicwere rated as more negative
on psychologicalattributes(Cash et al., 1984). The alco-
holic label producedstigma when the drinking pattern
was describedas normal (i.e., levels similar to most sub-
jects) as well as abnormal(i.e., more alcoholconsumed,
hard liquor,drinksalone, severehangovers and a reckless
driving charge).Respondents also reportedthey wanted
less contactwith the alcoholictarget in social and work
situations.Unfortunately,the quantity of drinking in the
abnormalconditionwas very low (i.e., four drinks). The
presentstudyusedalcoholand drug dosesthat reflected
an abuseprofile.
The presentstudy had adults evaluatescenariosthat
described individuals with different substance use labels
(e.g., heavy drinker versusalcoholic)for three different
substance types(alcohol,tobacco,cocaine)andwho were
seeking either an abstinentor a nonabstinentrecovery
throughtreatmentor on their own. The followingissues
were addressed: (1) people'sbeliefsaboutthe appropriate-
nessof self-changeand moderatedrinking outcomes;(2)
the degree of credibility attributedto substanceabusers
when they make statementsabouttheir recovery;and (3)
beliefs associated with the abuse of different substances.
Method
The study was approvedby the Ethics Committeeof
the Departmentof Psychology,University of Toronto.
Subjects19 years of age or older were recruitedat the
Ontario ScienceCenter(Toronto,Canada)in responseto a
poster stating "Take part in a psychologyexperiment.
Topic:What are your beliefsaboutaddictions?"Partici-
pantswere given a questionnaire and consentform and
were asked to follow the instructions on the consent form.
Subjectsweretold that the experimentwouldtakeapprox-
imately 15 minutesand that their answerswould be con-
fidential and anonymous.
Backgrounddata and information about the respon-
dent's own substanceuse were collected. The CAGE, a
four-itemquestionnaire, was usedto evaluatewhetherthe
subjecthad ever had an alcoholproblem(Mayfield et al.,
1974; Smart et al., 1991). Nicotine dependencewas as-
sessedby askingthe numberof cigarettessmokedduring
an averageday and the numberof minutesupon waking
until smokingthe first cigarette(Heathertonet al., 1989).
For cocaine,respondents were askedif they hadever tried
the drug. After they had completedthe questionnaire,re-
spondentswere thanked for their participation and any
questionsaboutthe studywere answered.
Subjects
Of the 606 respondents who volunteeredto participate
in the study,26 were droppedbecausethey were lessthan
19 yearsof ageor failedto givetheir age. Anotherrespon-
ON ALCOHOL / NOVEMBER 1993
dentwas droppedbecauseit appearedthat he did not take
the experimentseriously(e.g., reportedthat the average
male smokersmoked483 cigarettesper day). The final
samplesize was 579.
Differences in demographicvariables across condi-
tions were assessedusing two-way analysesof variance
(ANOVA)for parametricvariablesand chi-squaretestsfor
nonparametricvariables. Since none of the variables dif-
fered significantly(p > .05), data were collapsedacross
all conditions.The respondents'mean (-+SD) age was
29.1 -+ 9.3 years(range:19 to 76), and slightlymorethan
half were women (52.8%). Educationalstatuswas quite
high; almostall had completedhigh school(95.4%) and
58.1% had some university education. Three-quarters
(72.1%) of all respondentsresidedin Canada with most
of the remaining sample (23.5%) living in the United
States.Almost half (49.3%) had neversmokedcigarettes,
23.3% were ex-smokersand 27.4% currentlysmokedcig-
arettes. Of current smokers, the mean (-+ SD) number of
cigarettessmokedper day was 15.9 -+ 11.0 and the me-
dian (mean-+SD) number of minutesupon waking to
their first cigarette was 30.0 (88.6-+ 144.9). Almost
three-quarters(72.8%) of respondentswere currentdrink-
ers with the remainderreportingabstinence(16.8% cur-
rent abstainers,10.4% lifetime abstainers).One-quarter
(24.7%) scoredtwo or more on the CAGE, which is sug-
gestiveof havinghad (or having)an alcoholproblem.Fi-
nally, 24.0% reportedhavingtried cocaineat least once.
Procedure
Respondents were randomlyassignedto one of six tar-
get scenarioconditions:the scenariosinvolved a man who
had one of three different substanceuse problems(alco-
hol, tobaccoor cocaine)crossedwith two labelsreflecting
high or low substancedependencein a between-subjects
design.Other aspectsof the target scenario(drug quan-
tity, family andjob situation)were kept constant.The fol-
lowing is an exampleof one of the scenarios(alcohol,
low-dependence label):
JohnSmith is a heavydrinker.On week nightshe usuallyhas
from five to ten beers. On weekends, John has from ten to
fifteen beersa day. Johnis the managerof a small depart-
ment store. He is married with two children, aged six and
ten. In the last year, John has thought about how being a
heavydrinker is affectinghis life. (italics added)
The substanceuse labels employedin the target sce-
nario and in the questionnairewere as follows. (1) alco-
hol: "alcoholic" or "heavy drinker"; (2) tobacco:"heavy
chain smoker" or "moderate smoker"; and (3) cocaine:
"regular cocaine user" or "social/recreational cocaine
user." While the labels varied across conditions, the
amountof reportedsubstance
use was held constant.The
levels of use were: (1) alcohol abuser:drank 5 to 10 beers
 CUNNINGHAM,
Very Alcohol
Lxkelyto
succeed
4 il
Not at all
likelyto
succeed Alcoholic Heavy
Drinker
FIGURE1. Means(-+SEs)for ratingsof the likelihoodof the target'srecoverywith treatmentvs self-change
on week nightsand 10 to 15 beersper day on weekends;
(2) smoker:smokedabout30 cigarettesa day; and (3) co-
caine user:snorted1 to 2 hits of cocainea day.
Using 7-point scales,respondents were askedto rate the
likelihoodof the target'srecoveryif he tried to resolvehis
substanceuse problem with treatmentor without (self-
change). The likelihood of recovery was similarly as-
sessedfor abstinentand nonabstinent(i.e., moderateuse)
recoverieswithin treatment and self-changerecoveries.
Respondents were alsoaskedhow likely they wouldbe to
believe statementsthat the target had resolvedhis sub-
stanceuse problemthroughabstinenceor throughnonab-
stinence.Respondentsfurther ranked what they thought
were the most appropriatetreatmentsfor the target (i.e.,
physician,AlcoholicsAnonymous,friends/family,psychi-
atrist, church,self-change,treatmentagency).Lastly, re-
spondents were askedhow likely the target would be to
experiencevocational,legal andmarital problemsbecause
of his substanceuse problemand to rate how uncomfort-
able they would feel being the target'sco-workeror invit-
ing the target to a dinnerparty.
Results
Treatmentoutcomequestions
A 2 x 3 x 2 repeated-measures ANOVAcomparedre-
spondents'ratings of the likelihoodof the target's prob-
lem resolutionthrough self-change(without treatment)
versusthrough treatment (within subjects)acrosssub-
SOBELL AND CHOW 695
[ Self-Change
I Treatment
Tobacco Cocaine
Heavy
Chain Moderate Regular
Cocaine Social/
Smoker Smoker User Recreactional
Cocaine User
(no treatment)
stancetype and label conditions(betweensubjects).Since
the distributionof responses for the questionregarding
self-changewas positivelyskewed,the analysiswas per-
formed on natural logarithmtransformeddata. The ob-
servedmeansand standarderrorsare shownin Figure1.
A main effect was found for substance use labels
(F = 5.04, 1/568 df, p = .025). The low dependence la-
bel targetwasratedas morelikely to succeedat recovery
than the high dependencelabel target (observedmeans
[-SD] = 4.1 - 0.9 and 4.0 - 1.1, respectively).There
was also a main effect of self-changeversustreatment
(F = 1030.76 1/568 df, p < .001), indicatingthat the
treatedtargetwas ratedas morelikely to succeedthanthe
targetwho attemptedto changeon his own. Finally, there
was a main effect of substance type (F = 6.62, 2/568 df,
p = .001), and a significantinteractionbetweentreatment
and substancetypes (F = 15.89, 2/568 df, p < .001).
Scheff6posthoc pairwisecomparisons foundfive signifi-
cant (p < .05) differences.In the self-changecondition,
respondents in the tobaccoconditionrated the target as
morelikely to succeedthanthosein the alcoholor cocaine
condition. In the treatmentcondition, the likelihood of the
target succeedingdid not differ significantlyacrosssub-
stancetypes (p > .05). Finally, for all substancetypes,
recoverieswith treatmentwere rated as more likely to
succeedthan thosewithouttreatment(p < .05).
A 2 x 2 x 3 x 2 repeated-measures ANOVAwas con-
ductedcomparinga differentset of questionsthat crossed
self-changeversustreatmentand abstinentversusnonabsti-
nent resolutions(within-subjectvariables)with substance
696 JOURNAL OF STUDIES ON ALCOHOL / NOVEMBER 1993

Abstinent
Recovery
I Non-Abstinent
Recovery

Very Very TREATMENT

likelyto 7 SELF-CHANGE likely
to 7
succeed succeed

3-

Not at all Not at all

likelyto 1 - -- likelyto 1
succeed Alcohol Tobacco Cocaine succeed Alcohol Tobacco Cocaine
FOURE 2. Means(-SEs) for ratingsof the likelihoodof the target'srecoverywith treatmentvs self-change(no treatment)and with abstinentvs
nonabstinent resolutions

type and label conditions(between-subject variables). posthoc testsfoundfive significantpairwisecomparisons

Natural logarithm transformeddata were used in this
analysisbecauseresponsesto the self-changequestions
(abstinentand nonabstinentresolutions)were positively
skewed. The observed means and standard errors
shownin Figure 2.
There was no significantmain effect for substanceuse
labels (F = 1.8, 1/562 df, p > .05), but abstinentreso-
lutions were rated more likely to succeedthan were
nonabstinentresolutions (F = 951.6, 1/562 df, p <
.001). Also, resolutions through treatment were rated
more likely to succeedthan were those through self-
change (without treatment) (F = 77.2, 1/562 df, p <
.001). There was a significantmain effect for substance
type (F = 15.59, 2/562 df, p < .001) and a significant
interactionbetweenabstinenceand substancetype (F =
23.3, 2/562 df, p < .001). Scheff6post hoc pairwise
comparisons revealedeight significant(p < .05) differ-
ences.When abstinencewas selected,respondents in the
tobaccoconditionrated the target more likely to succeed
at recoveryas comparedto respondents' ratingsin either
the alcohol or cocaine conditions. When nonabstinence
was selected, all substancetypes differed significantly
(p < .05) with the smokingtargetrated as the mostlikely
to recoverand the cocainetarget least likely. Abstinent
resolutionswere rated as more likely to succeedthan non-
abstinentresolutionsfor all substances(p < .05). There
was also a significantinteractionbetweentreatmentand
substance types(F = 5.85, 2/562 df, p = .003). Scheff6
(p < .05). When self-changewas chosen,respondents in
the tobaccoconditionrated the target as more likely to
succeedthan respondents who were rating the alcoholor
are cocaine target. However, when treatment was chosen,
likelihoodof successdid not differ significantly(p > .05)
by substancetype. Resolutionsthrough treatmentwere
rated as more likely to succeedthan those throughself-
changefor all substancetypes (p < .05). Finally, there
was significant interaction between self-changeversus
treatment and abstinent versus nonabstinent resolutions
(F = 74.0, 1/562 df, p < .001). Scheff6 post hoc tests
found four significantpairwise comparisons(p < .05).
Abstinent resolutionswere rated more likely to succeed
than nonabstinent resolutions and treatment was rated as
superiorto self-change.The differencebetweenabstinent
and nonabstinentresolutionswas not as great when the
target attemptedself-changecomparedto when treatment
was used.
Treatment recommendations
Respondents
were askedto rank order sevenpossible
treatments(1 = most favored; 7 = least favored). Table 1
presentsthe mean rank for each treatmentfor each sub-
stancetype. For respondentsin the alcoholcondition,AA
and a treatmentagencywere the two most favoredtreat-
ment choices; self-changewas the least favored. For
respondentsin the tobaccoandcocaineconditions,a treat-
 CUNNINGHAM,
TABLE1. Meanranktreatmentrecommendations
for eachsubstance
type
Mean rank within substance
typea
Variable Alcohol Tobacco Cocaine
Treatmentagency 2.2 2.1
Physician 4.4 3.0
AA 1.9 NAb
Friends/family 4.1 3.6
Psychiatrist 4.1 4.0
Church 5.6 5.9
By himself/herself 5.7 3.6
'Alcohol: 1 = most favored treatment; 7 = least favored treatment.
Tobacco and cocaine: 1 = most favored treatment; 6 = least favored
treatment.
bNotapplicable.
ment agency or physicianwere rated as the treatment of
choice and the church least favored.
The credibilityand stigmaof substanceabuse
A 2 x 2 x 3 repeated-measures ANOVAcomparedthe
likelihood that the target would be believed(1 = not at
all likely; 7 -- very likely) if he told the respondentthat
he had an abstinent or nonabstinent resolution (within
subjects)acrosssubstancetype and label conditions(be-
tween subjects).There was a significantmain effect for
substancetype (F = 48.0, 2/559 df, p < .001). Posthoc
Scheff6testsrevealedthat respondents in the tobaccocon-
dition rated the target as significantly(p < .05) more
likely to be believed than thosein the alcohol or cocaine
conditions (means [-SDs] = 4.5 - 1.4, 3.3 - 1.4,
3.2- 1.4, respectively). There was also a main ef-
fect for abstinence versus nonabstinence (F = 161.4,
1/559 df, p < .001). Abstinent resolutionswere more
likely to be believed than were nonabstinentresolutions
(means= 4.3 - 1.8, 3.1 - 2.0, respectively).
An overall stigma scorewas formed by combiningthe
following five questions:(1) "If John Smith's supervisor
at work knew that John had an alcohol problem, how
likely do you think this would be to negatively affect
John'scareer?"; (2) "If John Smith was your co-worker,
how uncomfortablewould his continueddrinking make
you feel?"; (3) "How uncomfortable would you feel
about inviting John Smith to a dinner party?"; (4) "How
likely do you think it would be for John Smith's wife to
leave him becauseof his alcohol problem?"; (5) "How
likely do you think it would be for JohnSmith to get in
trouble with the law becauseof his alcohol problem?"
Since each item was scoredon a 7-point scale, the com-
positescorerangedfrom 5 to 35. Higher scoresindicated
that respondentsbelieved there was more stigma associ-
ated with the target's substanceuse. A two-way ANOVA
revealed a main effect of substancetype (F -- 293.5, 2/
531 df, p < .001). Scheff6 post hoc comparisonsfound
that respondents in the tobaccoconditionsrated the target
as significantly(p < .05) lesslikely to experiencestigma
SOBELL AND CHOW 697
because of his substance use than did those in the alcohol
or cocaine conditions (means = 16.2 _+ 6.4, 28.2 - 4.6,
28.4 _ 5.0, respectively).
1.5
Discussion
3.5
NAb
4.0
Substancetype significantlyaffected respondents'rat-
3.8
5.5 ings about likelihoodof recovery.The smokerwas rated
5.5 morelikely to recoverthan the targetwho abusedalcohol
or cocaine,especiallywhen the mode of recoverywas
self-change.However, irrespectiveof substancetype,
treatmentis perceivedas moreeffectivethan self-change,
abstinentresolutionsare seenas superiorto nonabstinent
resolutions and treatment and abstinent recoveries interact
in sucha way that when self-changeis the modeof recov-
ery abstinentresolutionsare rated almostas unlikely to
succeed as are nonabstinent resolutions.
Respondents'treatment recommendationsshed further
light on their perceptionsaboutdifferent substanceprob-
lems. For alcohol, the favored treatment was Alcoiolics
Anonymous,a finding consistentwith otherstudies(Blum
et al., 1989; Caetano, 1987; Mulford and Miller, 1961)
and which probablyreflects a widespreadawarenessof
this organization(Rodin, 1981). Treatmentagencieswere
the secondmost favored choice for the target's alcohol
problem, and the most favoredchoice for smokingand
cocaine-abuseproblems.Self-changewas rankedas one of
the least favored choices for the alcohol and cocaine-abuse
target, while for the smokingtargetit had a meanrank of
fourth and had a modal rankingof one. This finding is
consistentwith widespreadmedia reportsthat many ciga-
rette smokershave stoppedon their own (Fiore et al.,
1990;Office on Smokingand Health, 1988). Collectively,
these results indicate that resolutions without treatment
are seenas moreprobablefor cigarettesmokerscompared
to individualswith alcoholor cocaineproblems.
The perceptionthat recovery with treatment is more
likely to succeedthan recovery without treatmentis in
contrastto studiesshowingthat self-changeis a common
pathway to recovery for alcohol abusers(Fillmore, 1988;
Institute of Medicine, 1990; Sobell and Sobell, 1991; So-
bell et al., 1992). Thesefindingssuggestthat respondents
were not aware of, or did not have much confidence in,
the evidenceregardingnatural recoveries.
Since smokingwas not viewed as likely to causepsy-
chosocialproblems(e.g., work, legal, social, familial)
comparedto alcoholor cocaineabuse,it is possiblethat
respondentsdid not view the use of cigarettesas sub-
stanceabuse,a positionsimilar to that taken by the Amer-
ican PsychiatricAssociationthrough 1980 (Sobell et al.,
1990). Perhapssuch attitudeswill changeas the preva-
lenceof smokersdecreases andthe remainingsmokersac-
quire a more "die-hard image" (Coambs et al., 1989).
The fact that alcohol and cocaine abuse were associated
with similar degreesof stigma,particularly in relation to
698 JOURNAL OF STUDIES ON ALCOHOL
legal trouble,was surprising.One explanationis that re-
spondents thoughtof differentlegal issuesrelatedto each
drug (e.g., drunkdriving versusdrug arrests).
An importantreflectionof stigmafound in the present
study was the credibility accordedthe target's statement
about recovery.The smokingtarget'sclaims of recovery
weresignificantlymorelikely to be believedthanthoseof
the alcoholor cocaineabusetargets.Also, assertionsof
nonabstinentrecoverieswere greetedwith more skepti-
cism than were those of abstinent resolutions. Given the
prevalencein the generalpopulationof nonabstinent
re-
coveries for alcohol abusers(Fillmore, 1988; Sobell and
Sobell, 1991; Sobell et al., 1992), the fact that nonabsti-
nent recoveriesare less credible demonstratesanothergap
betweenrespondents' beliefsand reality.
Fewer label effects were found in this study as com-
paredto pastresearch(Cashet al., 1984;Dean and Rud,
1984; Kilty, 1981; Kilty and Meenaghan,1977; Stafford
and Petway, 1977; Wallstonet al., 1976). Previousstud-
ies, however, have either failed to mention the amount
of the substance used or have indicated use of small
amounts.Cash et al (1984), for example, describedthe
targetas consuming2-4 drinksper day, while the present
studyhad the target drinking 5-10 beerson weeknights
and 10-15 beers per day on the weekends.Perhapsthe
amountreportedin the presentstudywas so heavythat it
obscuredany label effects. Alternatively, it may be that
providingextensivedrinkingor druguseinformationmin-
imizesthe impactof the label. It is also possiblethat the
scenariosusedin the presentstudy(i.e., married, middle
classwith a family) are not consonant with that of an "al-
coholic," a "heavy chain smoker,"or a "regular cocaine
user." Finally, attitudestowards substanceabusersmay
have changedover time, leadingto the lack of label ef-
fects seen in this study.
Futurestudiesmay wish to explorethe influenceof dif-
ferenttarget (e.g., low versushigh socioeconomic status)
and respondent(e.g., educationalstatus)characteristics
on treatmentratings.Also, the disparitybetweenrespon-
dents'perceptionsof how recoveriesoccurand actualout-
comes suggeststhat researchmight focus on how such
beliefs could be changed.
Acknowledgments
The authors wish to thank Drs. Jonathan L. Freedman and Mark B.
Sobell for their help in this project.
References
GLUM, T.C., ROMAN,P.M. AND BENNETT,N. Public imagesof alco-
holism:Data froma Georgiasurvey.J. Stud.Alcohol50: 5-14, 1989.
CAETANO,R. Public opinionsabout alcoholismand its treatment.J.
Stud. Alcohol 48: 153-160, 1987.
/ NOVEMBER 1993
CASH, T.F., GRIDDELL,D.W., GILLEN, B. AND MACKINNON, C. When
alcoholicsare not anonymous: Socioperceptualeffectsof labelingand
drinking pattern. J. Stud. Alcohol 45: 272-275, 1984.
COAMBS,R.B., KOZLOWSKI,L.T. AND FERRENCE,R.G. The future of
tobaccouse and smokingresearch.In: NEY, T. ANDGALE, A. (Eds.)
Smokingand Human Behavior,New York: JohnWiley & Sons,Inc.,
1989, pp. 337-348.
COOPER, W.H. ANDKOHN, P.M. The socialimageof the youngfemale
smoker. Brit. J. Addict. 84: 935-941, 1989.
DEAN, J.C. AND POREMBA, G.A. The alcoholicstigmaand the disease
concept.Int. J. Addict 18: 739-751, 1983.
DEAN, J.C. ANDRUD, R. The drug addictand the stigmaof addiction.
Int. J. Addict. 19: 859-869, 1984.
DION, K.L., DION, K.K., COAMBS, R. AND KOZLOWSKI, L. Smokers
and Drinkers:A Tale of Two Stereotypes,
Ottawa, Canada:Canadian
PsychologicalAssociation,May, 1990.
ERICKSON,P.G. AND GOODSTADT, M.S. Legal stigma for marijuana
possession.
Criminology17: 208-216, 1979.
FILLMORE, K.M. Alcohol Use Across the Life Course: A Critical Re-
view of 70 Years of InternationalLongitudinalResearch,Toronto:
Addiction Research Foundation, 1988.
FIGRE, M.C., NOVOTNY, T.E., PIERCE,J.P., GIOVINO, G.A., HATZIAN-
DREU, E.J., NEWCOMB,P.A., SURAWICZ, T.S. AND DAVIS, R.M.
Methods used to quit smoking in the United States..lAMA 263:
2760-2765, 1990.
GOLDSTEIN, J. The stigmatizationof smokers:An empirical investiga-
tion. J. Drug Educ. 21: 167-182, 1991.
HEATHERTON,T.F., KOZLOWSKI,L.T., FRECKER,e.c., RICKERT, W.
ANDROBINSON, J. Measuringthe heavinessof smoking:Using self-
reportedtime to the first cigaretteof the day and numberof ciga-
rettessmokedper day. Brit. J. Addict. 84: 791-800, 1989.
INSTITUTEOF MEDICINE.Broadeningthe Baseof Treatmentfor Alcohol
Problems,WashingtonD.C.: National AcademyPress, 1990.
KILTY, K.M. Drinking statusand stigmatization.Amer. J. Drug Alco-
hol Abuse 8: 107-116, 1981.
KILTY, K.M. ANDMEENAGHAN,T.M. Drinking status,labelingand so-
cial rejection. J. Soc. Psychol.102: 93-104, 1977.
MCKIRNAN,D.J. A communityapproachto the recognitionof alcohol
abuse:The drinking normsof three Montrealcommunities.Canad.J.
Behav. Sci. 9: 108-122, 1977.
MAYFIELD,D., MCLEOD, G. AND HALL, P. The CAGE questionnaire:
Validation of a new alcoholismscreeninginstrument.Amer. J. Psy-
chiat. 131: 1121-1123, 1974.
MULFORD, H.A. AND MILLER, D.E. Public definition of the alcoholic.
Q. J. Stud. Alcohol 22: 312-320, 1961.
OFFICEON SMOKINGANDHEALTH. The Health Consequences of Smok-
ing. Nicotine Addiction: A Report of the SurgeonGeneral, 1988,
DHHS Publication No. (CDC) 88-8406, Washington:Government
Printing Office, 1988.
RODIN, M.B. Alcoholism as a folk disease:The paradoxof beliefs and
choiceof therapyin an urbanAmericancommunity.J. Stud. Alcohol
42: 822-835, 1981.
ROIZEN,R. Barriers to AlcoholismTreatment,Berkeley,Calif.: Alcohol
ResearchGroup, 1977.
SMART,R.G., ADLAF, E.M. AND KNOKE, D. Use of the CAGE scale in
a populationsurveyof drinking. J. Stud. Alcohol 52: 593-596, 1991.
SOBELL,L.C. AND SOBELL,M.B. Drunkenness,a "special circum-
stance" in crimes of violence: Sometimes. Int. J. Addict. 10: 869-
882, 1975.
SOBELL,L.C. AND SOBELL,M.B. Cognitive mediatorsof natural re-
coveriesfrom alcohol problems:Implicationsof treatment.Paper
presentedat the AnnualMeeting of the Associationfor Advancement
of BehaviorTherapy, New York, N.Y., November 1991.
SOBELL,L.C., SOBELL,M.B., KOZLOWSKI,L.T. AND TONEA, TTO, T.
Alcohol or tobacco research versus alcohol and tobacco research.
Brit. J. Addict. 85: 263-269, 1990.
 CUNNINGHAM, SOBELL AND CHOW 699

SOBELL,L.C., SOBELL,M.B. ANDTONEATTO,T. Recoveryfrom alco-
hol problemswithout treatment. In: HEATHER,$., MILLER, W.R.
AND GREELEY,J. (Eds.) Self-Control and Addictive Behaviors, New
York: Maxwell Macmillan, 1992, pp. 198-242.
STAFFORD,R.A. AND PETWAY, J.M. Stigmatization of men and
womenproblemdrinkersand their spouses: Differentialperception
and leveling of sex differences.J. Stud. Alcohol 38: 2109-2121,
1977.
THOM, B. Sex differencesin help-seekingfor alcohol problems--
1. The barriers to help-seeking. Brit. J. Addict. 81: 777-788,
1986.
TOOTLE,D.M. Social acceptanceof the recoveringalcoholicin the
workplace:A researchnote. J. Drug Issues17: 273-279, 1987.
TUCHFELD,B.S., SIMUEL,J.B., SCHMITT,M.L. ET AL. Changesin Pat-
terns of Alcohol Use without the Aid of Formal Treatment. An Ex-
ploratoryStudyof FormerProblemDrinkers,Final Report.Prepared
for the National Institute on Alcohol Abuse and Alcoholism,Publi-
cation No. PB 259-202, Springfield,Va.: National TechnicalInfor-
mation Services, 1976.
WALLSTON,K.A., WALLSTON,B.S. AND DEVELLIS, B.M. Effect of a
negativestereotypeon nurses'attitudestowardan alcoholicpatient.
J. Stud. Alcohol 37: 659-665, 1976.

CONTROVERSIES IN THE ADDICTIONS FIELD:

VOLUME ONE

Edited by Ruth C. Engs

This book presentsopposingsides of importantissues by leading authoritiesin the addictionsfield. There are nine
sections concerning:the nature of alcoholismand drug addictions;public policies to prevent alcohol abuse; the
minimumdrinkingage; mandatorydrug testingin the workplace;restrictionson beverageadvertising;the effective-
ness of warning labels in preventingalcohol abuse; abstinence for pregnant women; effectivenessof school based
alcohol/drugeducationprograms;abstinenceas the only treatmentfor alcoholism.
Forewordand introductionby Ruth C. Engs. Chaptersare contributedby: Reid K. Hesterand Nancy Sheehy;Donald
W. Goodwin;Ting-Kai Li; Robert D. Myers; HowardT. Blane; Herbert Fingarette;Karol L. Kumpfer,Eric R Trunnell
and Henry O. Whiteside;Robin Room; Dwight B. Heath; David J. Hanson; RichardJ. Goeman; Robert L. DuPont;
Thomas J. Delaney,Jr; KimberlyNeuendorf;Hal Shoup; Donna Polowchena;David Pittman;SterlingClarren; Ernest
Abel and Robert Sokol; Stuart Fors; Louis Gliksman and Cynthia Smythe; Gail Milgram; John Wallace; Martha
Sanchez-Craig.

1990 210 pages Paper ISBN 0-8403-6057-6

$9.60 (plus $2.50 shipping & handling)
Order from:

American Council on Alcoholism, Inc.

5024 Campbell Boulevard,Suite H
Baltimore, MD 21236-6950

Telephone: (410) 931-9393

Fax: (410) 931-4585