Addictive Behaviors 39 (2014) 1784–1788

Contents lists available at ScienceDirect

Addictive Behaviors

Comparing college smokers' and dual users' expectancies towardscigarette smoking

Joshua C. Gottlieb ⁎, Lee M. Cohen, Aaron K. HaslamTexas Tech University, USA

⁎ Corresponding author at: Department of Psycholog42051, Lubbock, TX 79409, USA. Tel.: +1 806 834 3644.

E-mail address: josh.gottlieb@ttu.edu (J.C. Gottlieb).

http://dx.doi.org/10.1016/j.addbeh.2014.07.0150306-4603/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Available online 17 July 2014

Keywords:Dual usersExpectanciesTobaccoCollege students

Background: As no agreed upon definition exists for dual use (i.e., individuals who concurrently use more thanone form of tobacco), this population remains largely unstudied in the substance use literature, despite increasesin smokeless tobacco use among young adults. Individuals 18–25 years of age report the highest rates of smoke-less tobacco use, dual use, and cigarette use. The current study compared the smoking outcome expectancies ofcollege student dual users to those who reported only smoking cigarettes.Methods: The Short Form of the Smoking Consequences Questionnairewas used to examine potential differences

in positive or negative expectations regarding cigarette use.Results: Data from this study suggest that smokers believe that smoking will lead to greater positive conse-quences (“cigarettes taste good”), negative reinforcement (“cigarettes helpmedealwith anger”), andweight/ap-petite reduction (“smoking controlsmy appetite”) when compared to dual users. Conversely, dual users believedthat smoking would lead to greater negative consequences (e.g., “smoking is taking years off of my life”).Discussion: These results may help to explain why some smokers choose not to use smokeless tobacco productsfor harm reduction or smoking cessation purposes, as well as why increases are being observed in smokeless to-bacco rates among young adults.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The use of tobacco products is a major public health concern in theUnited States and abroad. It remains the leading cause of preventabledeath worldwide and is associated with several serious illnessesincluding lung cancer, ischemic heart disease, and emphysema(Adhikari, Kahende, Malarcher, Pechacek, & Tong, 2008; World HealthOrganization [WHO], 2011). Despite the well-known health risks asso-ciated with tobacco use, many continue to smoke cigarettes and/oruse smokeless tobacco products (e.g., chewing tobacco, snuff, snus). In2012, an estimated 22.0% of Americans 18 years or older reportedcurrent cigarette use, while 3.6% reported current use of one or moresmokeless tobacco products (Centers for Disease Control & Prevention[CDC], 2014). Current use is defined as having smoked or usedsmokeless tobacco in the past 30 days.

While adult smoking rates have declined over the past 15 years,individuals 18–25 years of age have the highest rates of tobacco use.In 2012, nearly early one-third of young adults 18–25 were smokers(31.8%) and 5.5% used smokeless tobacco (CDC, 2014). Males tendedto smoke more than females (36.6% and 27.1%, respectively), and usemore smokeless tobacco (10.5% of males and 0.5% of females; CDC,

y, Texas Tech University, Box

2014). Alarmingly, nearly half (46.1%) of all new smokeless tobaccousers and over three-fourths (86.9%) of smokers initiate use beforeage 18 (CDC, 2014; Substance Abuse & Mental Health ServicesAdministration [SAMHSA], 2011).

Several explanationsmay account for the increase in smokeless tobac-co use among teens and young adults including, smoking bans and newflavors of smokeless tobacco products (e.g., cherry, apple blend, peach,and grape) clearly targeting this group (Alpert, Koh, & Connolly, 2008;Oliver, Jensen, Vogel, Anderson, & Hatsukami, 2013; Widome, Brock,Klein, & Forster, 2012). Furthermore, the addition of snus to theAmerican market has made tobacco use significantly easier to concealgiven that the user swallows saliva produced by the product rather thanspit it out (Foulds, Ramstrom, Burke, & Fagerström, 2003; Galanti,Wickholm, & Gilljam, 2001). The advertising of smokeless tobacco prod-ucts (including snus) has also increased, so larger numbers of our youthare being introduced to an addictive product where use can be easily hid-den from authority figures (Curry, Pederson, & Stryker, 2011).

These trends are disturbing as individuals are becoming dependenton nicotine earlier and are less likely to successfully quit after a cessa-tion attempt (Lando, Haddock, Robinson, Klesges, & Talcott, 2000).One reason for this is that young smokers have been substituting theircigarettes with smokeless tobacco products (rather than turning to anon-tobacco alternative), or they become dual users, thereby increasingtheir overall nicotine intake (Hatsukami, Ebbert, Feuer, Stepanov, &Hecht, 2007; Tomar, Alpert, & Connolly, 2010).

1785J.C. Gottlieb et al. / Addictive Behaviors 39 (2014) 1784–1788

Dual users are thosewho concurrently usemore than one formof to-bacco. Aside from the well-known risks of tobacco use, dual users faceadditional risks when compared to those who use either of the tobaccoproducts in isolation including, heavier alcohol consumption, and great-er levels of risk taking behavior (Klesges et al., 2011). Demographically,dual users tend to be Caucasian males, between the age of 18 and25 years, who live in the southern United States (McClave-Regan &Berkowitz, 2011). In 2012, prevalence estimates for dual use were10.1% of young adults 18–25 years old and 3.7% of all adults (CDC,2014). One study examining smokers, smokeless tobacco users, anddual users, found that dual users reported the greatest severity of nico-tinewithdrawal symptoms, followed by smokers, and smokeless tobac-co users (Post, Gilljam, Rosendahl, Bremberg, & Galanti, 2010).Furthermore, the same research group found that smokeless tobaccoand dual users reported symptoms of tobacco dependence that were2–5 times greater than cigarette smokers (Post et al., 2010).

1.1. Tobacco outcome expectancies

Smoking outcome expectancies are important to consider whenstudying an individual's smoking behavior, motivation to continuesmoking, and the likelihood of successful smoking cessation (Brandon,Juliano, & Copeland, 1999; Copeland & Brandon, 2000). Additionally,they are predictive of initiation, relapse, and levels of consumption(Brandon & Baker, 1991; Brandon et al., 1999; Copeland, Brandon, &Quinn, 1995; Rose, Chassin, Presson, & Sherman, 1996). Smokingoutcome expectancies can generally be grouped into the following cat-egories: 1) positive reinforcement smoking expectancies, 2) negativeconsequences smoking expectancies, 3) negative reinforcementsmoking expectancies, and 4) appetite/weight control smoking expec-tancies (Brandon & Baker, 1991).

Positive smoking expectancies include the facilitation of social situa-tions, and enjoyment of the flavor of the product used (Hendricks &Brandon, 2005; Morrell, Song, & Halpern-Felsher, 2010; Mullennix,Kilbey, Fisicaro, Farnsworth, & Torrento, 2003; Myers, McCarthy, Mac-Pherson, & Brown, 2003).While smokers may enjoy the positive effectsof smoking they are also aware of the negative consequences associatedwith continued use including the numerous health consequences,smelling like smoke, and having bad breath (Glock, Unz, & Kovacs,2012; Hendricks & Brandon, 2005). Smokers are able to rationalize con-tinued use via the positive expectancies they hold, despite concurrentlyholding negative expectancies (Glock et al., 2012).

There are also smoking outcome expectancies relevant to negativereinforcement. These expectancies have been found to be associatedwith trait worry and the expectation that smoking will reduce overallnegative affect. Furthermore, negative affect reduction expectanciesare also thought to induce a positive mood and are associated withsmoking behavior as well as nicotine dependence (Brandon, Wetter, &Baker, 1996; Downey&Kilbey, 1995; Peasley-Miklus,McLeish, Schmidt,& Zvolensky, 2012). Finally, expectancies regarding appetite/weight re-duction from smoking have been shown to be amotivating factor for in-dividuals to smoke as well, especially among young women (Adams,Baillie, & Copeland, 2011; Brandon & Baker, 1991). Women tend toinitiate smoking for weight control reasons and have stronger beliefsregarding the appetite suppressing properties of nicotine whencompared to men (Copeland & Carney, 2003; Copeland et al., 1995;French & Jeffery, 1995).

Much less is known regarding the role of smokeless tobaccooutcome expectancies. A recent study found that positive expectanciestowards smokeless tobacco are predictors of current use (Gottlieb,Cohen, Demarree, Treloar, & McCarthy, 2013). Health consequences re-lated to smokeless tobacco use were also found to be lower amongsmokeless tobacco users when compared to non-tobacco users andsmokers (Gottlieb et al., 2013). Another research group found thatsmokeless tobacco use was positively correlated with expectationsthat smokeless tobacco could control one's mood, as well as curb the

need to smoke (Wiium&Aarø, 2011). Given that nicotine is the primarypsychoactive substance present in both products it makes sense thatthere is a substantial overlap between smoking and smokeless tobaccooutcome expectancies, but differences also exist. One primary differ-ence is that while weight loss is a significant concern for many smokers,smokeless tobacco does not appear to be used for this purpose (Gerend,Boyle, Peterson, & Hatsukami, 1998).

While it is evident that little is known regarding the role of smoke-less tobacco outcome expectancies, even less is known about the expec-tancies of dual users of cigarettes and smokeless tobacco. Taking intoconsideration that dual use can be significantly more harmful to the in-dividual (McClave-Regan & Berkowitz, 2011; Noonan & Duffy, 2014)and dual users appear to be more difficult to treat when compared tousers of either product alone (Post et al., 2010; Rosendahl, Galanti, &Gilljam, 2008;Wetter et al., 2002), it is important to gain a better under-standing of this largely understudied group. Thus, the present study isdesigned to examine any differences between smokers and dual users'responses to a measure of smoking outcome expectancies.

It is hypothesized that smokerswill report higher positive reinforce-ment outcome expectancies for smoking when compared to dual users,given that smokers use only one tobacco product, whereas dual usersmay experience asmuch (ormore) positive reinforcement from smoke-less tobacco. Regarding negative consequence as a result of smoking, itis predicted that dual users will expect greater negative outcomeswhen compared to smokers, as the former may chose to use smokelesstobacco as a means of harm reduction and substitute such use a portionof the time they have urges to smoke.We also hypothesize that smokerswill display greater negative reinforcement outcome expectancieswhen compared to dual users as smokingmaybeused as the onlymeth-od for reducing negative affect, where dual users may decide to use asmokeless tobacco product instead of smoking. Finally, consideringthat a relationship between smokeless tobacco use and appetite/weightreduction has not yet been established and significantly moremen thanwomen use smokeless tobacco, it is hypothesized that smokers will re-port higher outcome expectancies related to appetite/weight reduction.

2. Method

2.1. Participants

This analysis was part of a larger study (N= 968) that developed ameasure for smokeless tobacco outcome expectancies among youngadults (Gottlieb et al., 2013). Participants in the current study were306 undergraduate students enrolled a large, public university in thesouthern United States. The other 662 students were not included inthe analyses for the current study due to the fact that not all phases ofdata collection for the development of the measure asked about fre-quency of use. Of the 306 students included, there was a fairly even dis-tribution of dual users (n = 140) and those who reported use ofcigarettes, but not smokeless tobacco (n = 166).

Cigar and pipe tobacco users were excluded from participation, andelectronic cigarette use was not assessed for, as these devices were notcommonly used at the time of assessment. The mean age was found tobe just under 20 years old (M = 19.89; SD = 3.70) and more females(58.2%) than males. The sample self-identified as 71.2% Caucasian,16.3% Hispanic or Latino, 3.9% African American, 2.9% Asian Americanor Pacific Islander, 0.7% Native American, and 4.9% identified as biracialor “other.” The noted demographic data are generally representative ofthe city in which the study was conducted, however it is important tomention that African Americans are underrepresented in this sample.See Table 1 for complete demographic information.

2.2. Procedure

The Human Research Protection Program (Institutional ReviewBoard) of the university where the study was conducted approved the

Table 1Participant demographics.

Smokers(n = 166)

Dual users(n = 140)

Combined(N = 306)

% Male 13.9 75.0 41.8% Female 86.1 25.0 58.2

Mean age (Standard Deviation) 19.36 (2.51) 20.52 (4.67) 19.89 (3.70)% African American 3.0 5.0 3.9% Asian American 1.2 5.0 2.9% Caucasian 67.5 75.7 71.2% Hispanic 21.7 10.0 16.3% Native American .6 .7 0.7% Biracial or other 6.0 3.5 4.9

1786 J.C. Gottlieb et al. / Addictive Behaviors 39 (2014) 1784–1788

procedures used for this study and all who participated did so voluntar-ily. Potential participants were able to choose from a variety of studiesavailable in the Department of Psychology via Sona ExperimentManagement System, a web-based human subject pool managementsystem. Students who signed up for the current study were given alink to an on-line questionnaire that included the measures describedbelow. Data reported for the current study were collected in 2012.

Following completion of the on-line questionnaire, students weredirected to another link where they entered their name so that coursecredit could be awarded for their participation while also keepingtheir responses anonymous. The on-line questionnaire consisted of abackground questionnaire that collected information regarding age,gender, race/ethnicity, and information related to their tobacco use(e.g., type of tobacco product used). The questionnaire also includedthe Short Form of the Smoking Consequences Questionnaire. Based onpast 30-day use, participants were assigned to one of two groups ac-cording to the type of tobacco product(s) used (i.e., dual users or ciga-rette smokers).

2.3. Measures

2.3.1. Background questionnaireA questionnaire designed for the present study was used in order to

collect personal information relevant to the purposes of the study. Allinformation was asked in a manner so that participants could not beidentified via their responses in order to assure anonymity and partici-pant honesty. Basic demographic questions were asked includingparticipant age, gender, and race/ethnic background. Additionally,frequency of tobacco use was assessed.

2.3.2. Smoking expectanciesThe Short Form of the Smoking Consequences Questionnaire (S-

SCQ; Myers et al., 2003) was developed as a brief measure to determinesmoking outcome expectancies in young adult and adolescent popula-tions. The S-SCQ consists of 21-items and four factors: 1) negative con-sequences, 2) positive reinforcement, 3) negative reinforcement, and 4)appetite-weight control. The original Smoking Consequences Question-naire (SCQ; Brandon & Baker, 1991) consists of 50-items and containsthe same four factors as the short form. Items are rated based on likeli-hood of occurrence if the individual smoked on a 10-point Likert-typescale (0 = completely unlikely to 9 = completely likely). The purposeof the SCQ is to measure an individual's subjective expected utility ofsmoking (Brandon & Baker, 1991). When compared to the SCQ (α =.96), internal consistency for the S-SCQ is excellent (α = .93).

The S-SCQwas also found to correlate highly with the full scale (r=.94, p b .01). This was also found to be the case for each factor as well:negative consequences (r = .78, p b .01), positive reinforcement (r =.86, p b .01), negative reinforcement (r = .99, p b .01), while the appe-tite/weight control factor included the same five items as the full SCQ.Confirmatory factor analyses show good fit for the four-factor structureof the S-SCQ (χ2 = 338.12 (183), SRMSR = .08, CFI = .92, TLI = .92).

2.4. Data analysis

The data that were used to test the hypotheses were not normallydistributed, indicating that there were both ceiling and floor effects de-pending on the subscale used in the S-SCQ. As a result, non-parametricMann–Whitney U tests with continuity correction were used forassessing all hypotheses. This statistical procedure does not examinethe group mean locations as is done in a t-test or analysis of variance(ANOVA), but instead examines the rank order of score values fromwithin each group without assuming an underlying distribution.Mann–Whitney U tests were conducted using R (V 3.0.2; RDevelopment Core Team, 2013). Measures of effect sizes were calculat-ed by hand using the following equation:

θ ¼ U= n1 � n2ð Þ ðEq:1Þ

where θ can be interpreted as the overlap between the two distribu-tions. When θ approaches 0 or 1 it is indicative of a little overlapbetween the populations and suggests differences between groups. Ifθ approaches .5 then there is little evidence that any true differenceexists between the groups (Newcombe, 2006).

3. Results

3.1. Gender and ethnic make up

Given that only 0.5% of females aged18–25 years are current smoke-less tobacco users (compared to 10.5% ofmales), we expected that therewould be a significant difference in gender between dual users andsmokers (CDC, 2014) which was found to be the case. Results indicatedthat there were significantly more males than females in the group ofdual users, and more females than males in the smoking group (χ2 =116.69, p b .001) which may suggest that a greater number of femalesmokers have never used a smokeless tobacco product when comparedto their male counterparts.

The racial makeup of the dual user and smoker groups were alsofound to be significantly different (χ2 = 12.62, p= .05). These findingsare consistent with national demographic data indicating that the vastmajority of those who report use of smokeless tobacco identify asCaucasian (CDC, 2014). Given the observed national prevalence rates,it is not surprising that our sample of dual users were more likely toidentify as Caucasian when compared to our sample of cigarettesmokers.

3.2. Smoking expectancies

As stated, Mann–Whitney U tests were conducted for the primaryhypotheses, as the data collected from the S-SCQwas not normally dis-tributed. We hypothesized that smokers would report greater positivereinforcement outcome expectancies when compared to dual users,and that dual users would expect greater negative consequenceswhen compared to smokers. Both of these hypotheses were confirmed(U = 9894, p = .988, θ = .426; U = 13462.5, p = .993, θ = .579, re-spectively). Further, we expected to observe that smokers would reportgreater negative reinforcement outcome expectancies and greaterappetite/weight control expectancies when compared to dualusers. Once again, both of these hypotheses were confirmed (U =9316, p= .001, θ= .401; U= 10,053, p= .02, θ= .433, respectively).See Table 2 for descriptive statistics by factor.

To assess the relationship between number of days smoked andnumber of days smokeless tobacco products were used and the S-SCQsubscale scores, Spearman's rho and Kendall tau correlations were con-ducted given the ordinal nature of the data and the fact that the datawere not normally distributed. The number of days smoked per weekwas not correlated with any of the S-SCQ expectancy subscales at thep b .05 level. There was, however, a significant positive relationship

Table 2Statistics for dual users and smokers on the S-SCQ (by factor).

First quartile Median Third quartile

Positive reinforcementDual users 0 2.1 5.2Smokers 0.8 3.9 6Negative consequencesDual users 7 8.75 9Smokers 6.25 7.875 9Appetite–weight controlDual users 0 0 3.9Smokers 0 1.6 4.25Negative reinforcementDual users 0 1.571 5.714Smokers 1.25 4.714 6.714

1787J.C. Gottlieb et al. / Addictive Behaviors 39 (2014) 1784–1788

observed between number of days participants reported usingsmokeless tobacco and the positive reinforcement factor score (ρ =.35, p b .001;τ= .299, p b .001,n= 306), indicating that as positive ex-pectancies increased so did the number of days used. Additionally, a sig-nificant negative relationship was observed between the number ofdays participants reported using smokeless tobacco and the negative re-inforcement factor score (ρ= − .363, p b .001; τ = − .3, p b .001, n =282) indicating that as negative expectancies increased, the number ofdays used decreased. Twenty-four participants who did not providecomplete data regarding the frequency of their smokeless tobacco usewere excluded from these analyses.

4. Discussion

The current study sought to examine the smoking outcome expec-tancies of college students who use different tobacco products. Bystudying tobacco users' expectancies, wemay gain insight intowhy col-lege students initiate and maintain a behavior that is known globally tocompromise an individual's health status. Dual users are especially im-portant to study given the lack of research among this group and sincemany are using smokeless tobacco products to replace (or at leasthelp them reduce) cigarette smoking. Gaining insight as to whysubstituting smokeless tobacco for cigarettes does not work for somecollege students (but does for others) may prove useful in the develop-ment of improved treatment packages for tobacco use disorder.

We examined positive and negative outcome expectancies of dualusers regarding the use of cigarettes when compared to those whoonly smoke using the S-SCQ, a valid and reliable measure (Myerset al., 2003). Itwas predicted that smokerswould report greater positivereinforcement, negative reinforcement, and appetite/weight controloutcome expectancies when compared to dual users. Conversely, itwas predicted that dual users would report a greater likelihood of neg-ative consequences as a result of their cigarette smoking when com-pared to smokers. All of our predictions were confirmed.

Data from this study suggest that smokers believe that smoking willlead to greater positive consequences (e.g., “when I smoke, the taste ispleasant”) when compared to their dual using counterparts. Thismakes sense conceptually since smokers only use cigarettes whiledual users may shift between products. It also appears that dual usersbelieve that smoking will lead to greater negative consequences(e.g., “smoking is taking years off of my life”) when compared to theirsmoking counterparts. This may be indicative of the fact that somedual users are trying to reduce their use of cigarettes and switch exclu-sively to smokeless tobacco products as a means of harm reduction(Rodu & Phillips, 2008).

Additionally, smokers were found to hold stronger beliefs thatsmoking would relieve negative affect (e.g., “cigarettes help me dealwith anger”) when compared to dual users. It is possible that smokersbelieve this more strongly because they turn to only one tobacco prod-uct to cope with negative affect. Conversely, dual users may use

cigarettes or a smokeless tobacco product to combat their negative af-fect. Finally, it was not surprising to observe that smokers held a stron-ger belief that smoking would lead to appetite/weight reduction(e.g., “smoking controls my appetite”) when compared to dual users.Since there is no evidence to support that smokeless tobacco users usein order to control weight, it may be that this observation is bestaccounted for by a gender effect (i.e., the smoking only group was pre-dominately female, whereas the dual user group was predominatelymale). This notion is consistent with prior research indicating thatwomen, more than men, smoke for this reason (Adams et al., 2011;Copeland & Carney, 2003).

Despite the potential utility of the findings, certain limitationsshould be consideredwhen interpreting the results. First, and as alludedto earlier, groups differed significantly by gender and race. As such, it ispossible that the differences observed are at least partially accounted forby the gender or racial makeup of each group rather than the tobaccoproduct used. In the future, studies should take these factors into ac-count when recruiting participants. This said, the demographics of thegroup of dual users in the present study resembled rates of dual use na-tionally, however, it is unclear why the smoking group predominatelyconsisted of female participants. One potential explanation is that uni-versity populations tend to consist of more females than males(Morrell, Cohen, & Dempsey, 2008).

Another limitation is that we did not assess participants' reasons forbeing a dual user. Specifically, we did not assess which tobacco productwas preferred by the dual users who participated in the current study,nor did we assess the reasons for the use of multiple products. In addi-tion,we did not assess for how long individuals have beenusing tobaccoproducts or assess which individuals primarily smoked cigarettes com-pared to those who primarily use smokeless tobacco. This informationmay better elucidate the findings from this study. Future studies shouldexamine the specific roles for the use of each tobacco product amongdual users, as well as why college students who choose to use a singletobacco product have the preference they do. Further, differences be-tween dual users who primarily use smokeless tobacco versus thosewho primarily smoke cigarettes should be examined, including any dif-ferences in expectancies across levels of use. Gaining a better under-standing of such differences may help treatment providers distinguishwhether an individual will be better suited for participation in asmoking cessation group, a smokeless tobacco cessation program, or aprotocol designed to addresses dual use.

In sum, we found that dual users were more likely to believe thatsmoking cigarettes would lead to negative consequences when com-pared to smokers. We also found that smokers believe that smokingwill lead to greater positive reinforcement, negative reinforcement,and hold a greater belief that smoking will serve as an appetite/weightsuppressant when compared to dual users. Overall the findings of thepresent study suggest that because smokers choose to use only oneform of tobacco, they are more likely to have favorable expectancies re-garding smoking than college students who use multiple tobacco prod-ucts. It is feasible that those who use two or more forms of tobacco holdless favorable expectancies regarding smoking since they have one ormore alternate tobacco products to turn to in order to satisfy their crav-ings for nicotine, while still receiving the positive sensations that are as-sociated with tobacco use. By examining tobacco users' outcomeexpectancies, researchers can better understand what ultimately leadsan individual to use the substance, which can lead to improved treat-ment programs and outcomes specifically tailored for college students.

Role of funding sourcesThere were no funding sources for the research leading to the current manuscript.

ContributorsJoshua C. Gottlieb and Lee M. Cohen designed the study. Joshua C. Gottlieb conducted

literatures searches relevant to the study and prepared articles for the various sections ofthemanuscript. Aaron K. Haslam conducted the statistical analyses. Joshua C. Gottlieb andLeeM. Cohen wrote the first draft of themanuscript, and Aaron K. Haslam helped edit the

1788 J.C. Gottlieb et al. / Addictive Behaviors 39 (2014) 1784–1788

final document. All three authors: Joshua C. Gottlieb, Lee M. Cohen, and Aaron K. Haslamcontributed to and have approved the final version of the manuscript.

Conflict of interestJoshua C. Gottlieb, LeeM. Cohen, and Aaron K. Haslam have no conflicts of interest, fi-

nancial or otherwise.

References

Adams, C. E., Baillie, L. E., & Copeland, A. L. (2011). The smoking-related weight and eatingepisodes test (SWEET): Development and preliminary validation. Nicotine & TobaccoResearch, 13(11), 1123–1131. http://dx.doi.org/10.1093/ntr/ntr162.

Adhikari, B., Kahende, J., Malarcher, A., Pechacek, T., & Tong, V. (2008). Smoking-attributable mortality, years of potential life lost, and productivity losses — UnitedStates, 2000–2004. Morbidity and Mortality Weekly Report, 57(45), 1226–1228.

Alpert, H. R., Koh, H., & Connolly, G. N. (2008). Free nicotine content and strategicmarket-ing of moist snuff tobacco products in the United States: 2000–2006. Tobacco Control,17(5), 332–338. http://dx.doi.org/10.1136/tc.2008.025247.

Brandon, T. H., & Baker, T. B. (1991). The Smoking Consequences Questionnaire: The sub-jective expected utility of smoking in college students. Psychological Assessment, 3(3),484–491. http://dx.doi.org/10.1037/1040-3590.3.3.484.

Brandon, T. H., Juliano, L. M., & Copeland, A. L. (1999). Expectancies in tobacco smoking. InI. Kirsch (Ed.), How expectancies shape experience (pp. 263–299). Washington, DC:American Psychological Association.

Brandon, T. H., Wetter, D. W., & Baker, T. B. (1996). Affect, expectancies, urges, andsmoking: Do they conform to models of drug motivation and relapse? Experimentaland Clinical Psychopharmacology, 4(1), 29–36. http://dx.doi.org/10.1037/1064-1297.4.1.29.

Centers for Disease Control and Prevention (2014). The health consequences of smoking —50 years of progress. A report of the Surgeon General. Atlanta, GA: U.S. Department ofHealth and Human Services.

Copeland, A. L., & Brandon, T. H. (2000). Testing the causal role of expectancies in smokingmotivation and behavior. Addictive Behaviors, 25(3), 445–449. http://dx.doi.org/10.1037/1040-3590.7.4.484.

Copeland, A. L., Brandon, T. H., & Quinn, E. P. (1995). The Smoking ConsequencesQuestionnaire-Adult: Measurement of smoking outcome expectancies of experi-enced smokers. Psychological Assessment, 7(4), 484–494. http://dx.doi.org/10.1037/1040-3590.7.4.484.

Copeland, A. L., & Carney, C. E. (2003). Smoking expectancies as mediators between die-tary restraint and disinhibition and smoking in college women. Experimental andClinical Psychopharmacology, 11(3), 247–251. http://dx.doi.org/10.1037/1064-1297.11.3.247.

Curry, L. E., Pederson, L. L., & Stryker, J. E. (2011). The changing marketing of smokelesstobacco in magazine advertisements. Nicotine & Tobacco Research, 13(7), 540–547.http://dx.doi.org/10.1093/ntr/ntr038.

Downey, K. K., & Kilbey, M. M. (1995). Relationship between nicotine and alcohol expec-tancies and substance dependence. Experimental and Clinical Psychopharmacology,3(2), 174–182. http://dx.doi.org/10.1037/1064-1297.3.2.174.

Foulds, J., Ramstrom, L., Burke, M., & Fagerström, K. (2003). Effect of smokeless tobacco(snus) on smoking and public health in Sweden. Tobacco Control, 12(4), 349–359.http://dx.doi.org/10.1136/tc.12.4.349.

French, S. A., & Jeffery, R. W. (1995). Weight concerns and smoking: A literature review.Annals of Behavioral Medicine, 17(3), 234–244. http://dx.doi.org/10.1007/BF02903918.

Galanti, M. R., Wickholm, S., & Gilljam, H. (2001). Between harm and dangers: Oral snuffuse, cigarette smoking and problem behaviours in a survey of Swedish male adoles-cents. European Journal of Public Health, 11(3), 340–345. http://dx.doi.org/10.1093/eurpub/11.3.340.

Gerend, M.A., Boyle, R. G., Peterson, C. B., & Hatsukami, D. K. (1998). Behavior and weightcontrol among women using smokeless tobacco, cigarettes, and normal controls.Addictive Behaviors, 23(2), 171–178. http://dx.doi.org/10.1016/S0306-4603(97)00023-3.

Glock, S., Unz, D., & Kovacs, C. (2012). Beyond fear appeals: Contradicting positivesmoking outcome expectancies to influence smokers' implicit attitudes, perception,and behavior. Addictive Behaviors, 37(4), 548–551. http://dx.doi.org/10.1016/j.addbeh.2011.11.032.

Gottlieb, J. C., Cohen, L. M., Demarree, K. G., Treloar, H. R., & McCarthy, D.M. (2013). TheDevelopment and Psychometric Evaluation of the Smokeless Tobacco ExpectanciesScale (STES). Psychological Assessment, 25(3), 997–1001. http://dx.doi.org/10.1037/a0032256.

Hatsukami, D. K., Ebbert, J. O., Feuer, R. M., Stepanov, I., & Hecht, S. S. (2007). Changingsmokeless tobacco products new tobacco-delivery systems. American Journal of Pre-ventive Medicine, 33(6 Suppl.), S368–S378. http://dx.doi.org/10.1016/j.amepre.2007.09.005.

Hendricks, P.S., & Brandon, T. H. (2005). Smoking expectancy associates among collegesmokers. Addictive Behaviors, 30(2), 235–245. http://dx.doi.org/10.1016/j.addbeh.2004.05.006.

Klesges, R. C., Ebbert, J. O., Morgan, G. D., Sherrill-Mittleman, D., Asfar, T., Talcott, W. G., et al.(2011). Impact of differing definitions of dual tobacco use: Implications for studyingdual use and a call for operational definitions. Nicotine & Tobacco Research, 13(7),523–531. http://dx.doi.org/10.1093/ntr/ntr032.

Lando, H. A., Haddock, C. K., Robinson, L. A., Klesges, R. C., & Talcott, G. W. (2000). Ethnicdifferences in patterns and correlates of age of initiation in a population of Air Forcerecruits. Nicotine & Tobacco Research, 2, 337–344. http://dx.doi.org/10.1080/14622200020028065.

McClave-Regan, A. K., & Berkowitz, J. (2011). Smokers who are also using smokeless to-bacco products in the US: A national assessment of characteristics, behaviours andbeliefs of ‘dual users’. Tobacco Control, 20(3), 239–242. http://dx.doi.org/10.1136/tc.2010.039115.

Morrell, H. E. R., Cohen, L. M., & Dempsey, J. P. (2008). Smoking prevalence and awarenessamong undergraduate and health care students. American Journal on Addictions,17(3), 181–186. http://dx.doi.org/10.1080/10550490802019899.

Morrell, H. E., Song, A. V., & Halpern-Felsher, B.L. (2010). Predicting adolescent percep-tions of the risks and benefits of cigarette smoking: A longitudinal investigation.Health Psychology, 29(6), 610–617. http://dx.doi.org/10.1037/a0021237.

Mullennix, J. W., Kilbey, M. M., Fisicaro, S. A., Farnsworth, L. M., & Torrento, R. (2003). Therepresentation of smoking expectancies by single-word adjectives and their relation-ship to smoking characteristics.Nicotine & Tobacco Research, 5(5), 681–694. http://dx.doi.org/10.1080/1462220031000158654.

Myers, M. G., McCarthy, D.M., MacPherson, L., & Brown, S. A. (2003). Constructing a shortform of the Smoking Consequences Questionnaire with adolescents and youngadults. Psychological Assessment, 15(2), 163–172. http://dx.doi.org/10.1037/1040-3590.15.2.163.

Newcombe, R. G. (2006). Confidence intervals for an effect size measure based on theMann–Whitney statistic. Part 1: General issues and tail-area-based methods.Statistics in Medicine, 25(4), 543–557. http://dx.doi.org/10.1002/sim.2323.

Noonan, D., & Duffy, S. A. (2014). Factors associated with smokeless tobacco use and dualuse among blue collar workers. Public Health Nursing, 31(1), 19–27. http://dx.doi.org/10.1111/phn.12095.

Oliver, A. J., Jensen, J. A., Vogel, R. I., Anderson, A. J., & Hatsukami, D. K. (2013). Flavoredand nonflavored smokeless tobacco products: Rate, pattern of use, and effects.Nicotine & Tobacco Research, 15(1), 88–92. http://dx.doi.org/10.1093/ntr/nts093.

Peasley-Miklus, C. E., McLeish, A.C., Schmidt, N.B., & Zvolensky, M. J. (2012). An examina-tion of smoking outcome expectancies, smoking motives and trait worry in a sampleof treatment-seeking smokers. Addictive Behaviors, 37(4), 407–413. http://dx.doi.org/10.1016/j.addbeh.2011.11.037.

Post, A., Gilljam, H., Rosendahl, I., Bremberg, S., & Galanti, M. R. (2010). Symptoms of nic-otine dependence in a cohort of Swedish youths: A comparison between smokers,smokeless tobacco users and dual tobacco users. Addiction, 105(4), 740–746. http://dx.doi.org/10.1111/j.1360-0443.2009.02852.x.

R Development Core Team (2013). R: A language and environment for statistical computing(version 3.0.2). Vienna, Austria: R Foundation for Statistical Computing (Retrievedfrom http://www.R-project.org).

Rodu, B., & Phillips, C. V. (2008). Switching to smokeless tobacco as a smoking cessationmethod: Evidence from the 2000 National Health Interview Survey. Harm ReductionJournal, 5, 18. http://dx.doi.org/10.1186/1477-7517-5-18.

Rose, J. S., Chassin, L., Presson, C. C., & Sherman, S. J. (1996). Prospective predictors of quitattempts and smoking cessation in young adults. Health Psychology, 15(4), 261–268.http://dx.doi.org/10.1037/0278-6133.15.4.261.

Rosendahl, K. I., Galanti, M. R., & Gilljam, H. (2008). Trajectories of smokeless tobacco useand of cigarette smoking in a cohort of Swedish adolescents: Differences and implica-tions. Nicotine & Tobacco Research, 10(6), 1021–1027. http://dx.doi.org/10.1080/14622200802097522.

Substance Abuse andMental Health Services Administration (2011). Results from the 2010National Survey on Drug Use and Health: Summary of national findings.HHS PublicationNo. 11-46580. Rockville, MD: Substance Abuse and Mental Health ServicesAdministration.

Tomar, S. L., Alpert, H. R., & Connolly, G. N. (2010). Patterns of dual use of cigarettes andsmokeless tobacco among USmales: Findings from national surveys. Tobacco Control,19(2), 104–109. http://dx.doi.org/10.1136/tc.2009.031070.

Wetter, D. W., McClure, J. B., de Moor, C., Cofta-Gunn, L., Cummings, S., Cinciripini, P.M.,et al. (2002). Concomitant use of cigarettes and smokeless tobacco: Prevalence, cor-relates, and predictors of tobacco cessation. Preventive Medicine, 34(6), 638–648.http://dx.doi.org/10.1006/pmed.2002.1032.

Widome, R., Brock, B., Klein, E. G., & Forster, J. L. (2012). Smokeless tobacco advertising atthe point of sale: Prevalence, placement, and demographic correlates. Nicotine &Tobacco Research, 14(2), 217–223. http://dx.doi.org/10.1093/ntr/ntr188.

Wiium, N., & Aarø, L. E. (2011). Outcome expectations and use of smokeless tobacco(snus): A cross-sectional study among young Norwegian snus users. ScandinavianJournal of Psychology, 52(1), 64–70. http://dx.doi.org/10.1111/j.1467-9450.2010.00823.x.

World Health Organization (2011). WHO report on the global tobacco epidemic, 2011:Warning about the dangers of tobacco. (Geneva, Switzerland).