7 page powerpoint slide

BRAVOET AL. 203 Can I Use Marijuana Safely? An Examination of Distal Antecedents, Marijuana ProtectiveBehavioral Strategies, and Marijuana Outcomes ADRIAN J.BRA VO, PH.D.,aMARK A. PRINCE, PH.D.,bMATTHEW R. PEARSON, PH.D.,a,* & THE MARIJUANA OUTCOMES STUDY TEAM † aCenter on Alcoholism, Substance Abuse, and Addictions, University of New Mexico, Albuquerque, New MexicobDepartment of Psycholo gy, Colorado State University, Fort Collins, Colorado 203 ABSTRACT. Objecti ve:Gi ven the high pre valence of marijuana use among college students, it is imperati veto determine the factors that may reduce risk of problematic marijuana use and/or the de velopment of cannabis use disorder. Weexamined marijuana protecti vebehavioral strategies (PBS) as a proximal predictor of marijuana-related outcomes and amediator of the associations between other known risk/protec- ti ve factors and marijuana-related outcomes. Method:Using data from a sample of 2,129 past-month marijuana users, collected from 11 uni versities in the United States, weexamined marijuana PBS use as a mediator of the effects of sex, age at ﬁrst use, impulsivity-like traits, and marijuana use moti ves on marijuana use frequency and marijuana- related consequences.

Results:Marijuana PBS wasidentiﬁed as arobust negati vepredictor of marijuana use frequenc yand marijuana-related consequences. Further, Marijuana PBS use fully or partially mediated the effects of sex, premeditation, perse verance, coping moti ves, en- hancement moti ves, conformity moti ves, and expansion moti veson marijuana outcomes. Conclusions:Our results suggest that marijuana PBS use is agood candidate to be considered as amechanism bywhich marijuan auser smoderate thei rmarijuan aus ean dattenuat ethei rrisk of ex periencing marijuana-related consequences. (J. Stud. Alcohol Drugs, 78, 203–212, 2017) Recei ved: July 15, 2016. Revision: No vember 4, 2016.

Adrian J.Bra vois supported bya training grant (T32-AA018108) from the National Institute on Alcohol Abuse and Alcoholism (NIAAA). Matthew R. Pearson is supported bya career de velopment grant from the NIAAA (K01-AA023233). The views expressed in this article are those of the authors and not of the National Institute on Alcohol Abuse or Alcoholism. *Correspondence may be sent to Matthew R. Pearson at the Center on Alcoholism, Substance Abuse, and Addictions, Uni versity of New Mexico, 2650 Yale Blvd SE MSC 11-6280, Albuquerque, NM 87106, or via email at:

[email protected]. †This project was completed bythe Marijuana Outcomes Study Team (MOST), which includes the following in vestigators (in alphabetical order): Amber Anthenien, Uni versity of Houston; Adrian J. Bra vo,Uni versity of Ne wMexico; Bradle yT. Conner ,Colorado State R AT ES OF MARIJUANA USE AND cannabis use dis- order peak during traditional college years (ages 18– 25) in the U.S. (Farmer et al., 2015). In fact, about 30% of college students report past-year pre valence of marijuana use, and nearly 10% meet diagnostic criteria for cannabis use disorder (Caldeira et al., 2008; Johnston et al., 2015). In arecent study across 11 different U.S. uni versities, Pe arson and colleagues (2017) found that between 15.5% and 38.7% (M = 26.2%) of college students report using marijuana in the past month. Further ,marijuana-related negati veconsequences are pre valent with marijuana users ex periencing approximately eight distinct negati veconse- quences monthly (Pearson et al., 2017). Although there are se ve ral known risk (e.g., earlier age at ﬁrst use and coping moti ves) and protecti vefactors (e.g., self-regulation and female sex) of problematic marijuana use, research needs to go be yond examining onl ydistal antecedents 1and consider more proximal behaviors that ma yincrease or decrease the negati veconsequences from using marijuana.

Wi thin the present study, wefocus on marijuana protec- ti ve behavioral strategies (PBS; Pedersen et al., 2016) as a proximal factor expected to a) relate to both frequency of marijuana use and marijuana-related negati veconsequenc- es, and b) account for the effects of se veral known risk/ protecti vefactors of problematic marijuana use.

Stemming from a harm reduction focus in the alcohol ﬁeld, much research has been conducted examining the use of alcohol PBS, deﬁned as “behaviors that are used im- 1We refer to these variables as “distal antecedents” to marijuana- related outcomes to distinguish them from more “proximal antecedents,” which tend to be variables that are less stable, more malleable, and presumed to be more proximal in a causal chain leading to marijuana-related outcomes.

Uni versity; Christopher J.Correia, Auburn Uni versity; Robert D.Dvorak, Uni versity of Central Florida; Gregory A. Egerton, Uni versity at Buf falo; John T.P.Hustad ,Pe nnsylvania State Uni versity College of Medicine; Ta tyana Kholodk ov,Uni versity of Wyoming; Kevin King, Uni versity of Wa shington; Bruce S. Liese, Uni versity of Kansas; Bryan G. Messina, Auburn Uni versity; James G. Murph y,The Uni versity of Memphis; Clayton Neighbors, Uni versity of Houston; Xuan-Thanh Nguyen, Uni versity of California, Los Angeles; Jamie E. Parnes, Colorado State Uni versity; Mat- the wR. Pearson, Uni versity of Ne wMexico; Eric R. Pedersen, RAND; Mark A. Prince, Colorado State Uni versity; Sharon A. Radomski, Uni versity at Buf falo; Lara A. Ra y,Uni versity of California, Los Angeles; and Jennifer P.

Read, Uni versity at Buf falo. 204JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / MARCH 2017 mediately prior to, during, and/or after drinking that reduce alcohol use, intoxication, and/or alcohol-related harm” (Pearson, 2013, p. 1030). Examples of alcohol PBS identi- ﬁ ed in previous work (Protecti veBehavior Strategies Sur vey; Martens et al., 2005) include limiting/stopping drinking (e.g., “Stop drinking at a predetermined time”), manner of drinking (e.g., “Drink slowl y,rather than gulp or chug”), and serious harm reduction (e.g., “Know where your drink has been at all times”). Within the college student alcohol literature, increasing evidence suggests that PBS use is a robust protecti vefactor associated with lo weralcohol-related consequences (Pearson, 2013; Prince et al., 2013). Moreo ver, alcohol PBS use has been found to mediate the effects of multiple inter ventions (i.e., increased PBS use wasrelated to lo we r alcohol outcomes; Barnett et al., 2007; Dvorak et al., 2015, 2016; Larimer et al., 2007; Murph yet al., 2012), pro- viding some evidence that PBS use is aproximal mechanism of changing one’s alcohol use and related outcomes (Prince et al., 2013). Recentl y,Pe dersen and colleagues de veloped the Protec- ti ve Behavioral Strategies for Marijuana (PBSM) scale to probe marijuana PBS use (Pedersen et al., 2016). Similar to alcohol PBS, marijuana PBS include strategies that arerelated to limiting marijuana intake bysetting consump- tion limits (e.g., “Having a set amount of times you take a hit of a marijuana joint”), avoiding behaviors that lead to more intoxication than one would like (e.g., “Avo id us- ing marijuana in concentrated form [e.g., hashish, hashish/ honey oil, kief, marijuana butter/oil] to avoid getting too high”), and avoiding serious harm from impaired driving (e.g., “Use a designated dri ver [i.e., someone who has not used] after using marijuana”). In addition, marijuana PBS include behaviors that minimize potential problems in interpersonal relationships (e.g., “Avo id use while spend- ing time with family”), reduce problems at work or school (“ Av oid using marijuana before work or school”), and pre- ve nt legal troubles (e.g., “Avo id possibilities of legal reper- cussions [e.g., smoke in a safe place like home, avoiding having marijuana with youwh ere you might get searched, etc.]”). Taken together, if individuals effecti vely use PBS, the ycould simultaneousl ydecrease large amounts of m arijuana use and reduce the likelihood of experienc- ing negati vemarijuana-related consequences. In suppor t of this notion ,th e PBSM wasfound to ha veasimple, single-factor structure and correlate negati vely with vari- ous indicators of marijuana in volvement among past-month users (e.g., marijuana use frequenc yand consequences).

Although this study provides preliminar yevidence that marijuana PBS use is an important protecti vefactor that reduces marijuana-related harms (e.g., social-impersonal consequences, impaired control, and risk behaviors), it is the only quantitati vestudy to date that has examined mari- juana PBS use and replication is warranted.

Be yond just examining the PBS use-alcohol outcomes link, multiple researchers ha ve examined whether alcohol PBS use mediates the associations between more distal antecedents and alcohol-related outcomes (Pearson, 2013;Prince et al., 2013). Recentl y,Bra voand colleagues (2015, 2016) were able to replicate the protecti veeffect of alcohol PBS use on alcohol outcomes, as well as most of the direct effects of antecedent variables on alcohol PBS use, includ- ing age at drinking onset (positi ve), drinking moti ves (neg- ati ve), and impulsivity-like traits 2(negati ve; Bra voet al., 2015, 2016). Similar to what has been found in the alcohol ﬁeld, male sex (Johnston et al., 2015), earlier age at ﬁrst use (Chen et al., 2009), impulsivity-like traits (Kaiser et al., 2012; Robinson et al., 2014; Wardell et al., 2016), and marijuana use moti ves (Simons et al., 1998; Zvolensky et al., 2007) ha veall been shown to be risk factors associated with increased marijuana use, consequences, and/or depen- dence. Ho wever,give n the nascence of the marijuana PBS ﬁeld, no studies ha ve examined whether marijuana PBS use mediates the associations between these antecedents and marijuana-related outcomes.

Purpose of study The purpose of the present study is to extend research on the associations between distal antecedents, marijuana PBS use, and marijuana outcomes among college student marijuana users. Although a large alcohol PBS ﬁ eld sug- gests that alcohol PBS use mediates the effects of a wide range of distal antecedents on alcohol-related outcomes, it is unknown whether marijuana PBS use would operate similarl y.Based on mediation ﬁndings in the alcohol PBS literature (Bra voet al., 2015, 2016) and preliminary ﬁnd- ings that marijuana PBS use is a protecti vefactor associ- ated with reduced marijuana-related harm (Pedersen et al., 2016), weexpected that the associations between distal an- tecedents (i.e., sex, age at ﬁrst use, impulsivity-like traits, and marijuana use moti ves) and marijuana outcomes (i.e., marijuana use frequency and consequences) would be me- diated bymarijuana PBS use. Generall y,we expected that “protecti ve”factors (e.g., female sex, premeditation, and perse verance) that are associated with greater marijuana PBS use would in turn be associated with less frequent marijuana use and fe werrelated consequences, whereas “risk” factors (e.g., younger age at ﬁrst use, coping mo- ti ve s) associated with less marijuana PBS use would in turn be associated with greater marijuana use frequency and more related consequences. 2Impulsivity-like traits are distinct constructs that contribute to impulsi vebehaviors: sensation seeking, urgency (positi veand negati ve), planning, and persistence. These facets ha vebeen shown to be distinct traits of impulsivity and demonstrate different aspects of risky behaviors and clinical utility (Smith et al., 2007; Whiteside& Ly nam, 2001). BRAVOET AL. 205 Method Pa rticipants and procedure Pa rticipants were college students recruited from Psychol- og y Department Participant Pools at 11 participating uni- ve rsities in the United States between fall 2015 and spring 2016. Participants completed an online sur vey examining the correlates of marijuana use among college students and recei ved research participation credit on completion. For the present study, data only from students who consumed marijuana at least 1 day in the previous month (n = 2,129) we re included in the ﬁnal analysis from a larger sample (n = 8,141, 66.9% female; see Pearson et al., 2017, for more information about the larger sample). Among college stu- dent marijuana users, the majority of participants identiﬁed as being either White, non-Hispanic (n = 1,285; 60.4%), or of Hispanic/Latino ethnicity (n = 390; 18.3%),were female (n = 1,260; 59.2%), and reported a mean age at 19.95 (SD = 3.66) years. The study was appro vedby the institutional review boards at the participating institutions.

Measures Ageat ﬁrs tuse .Age at ﬁrst use wasassessed with asingle item: “Ho wold wereyouthe ﬁrst time youused marijuana?” Impulsivity-like traits .Impuls ivity-lik etraits were assessed using the 59-item UPPS-P ,wh ich combines the 14-item Posi- ti ve Urgency Measure (Cyders et al., 2007) with the 45-item Urgenc yPremeditation Perse verance Sensation Seeking Impulsi veBehavior Scale (Whiteside & Lynam, 2001). The UPPS-P assesses ﬁve distinct impulsivity-like traits (each individually examined in our models): negati veurgency (12 items; α= .87), premeditation (11 items; α= .84), perse ver- ance (10 items; α= .79), sensation seeking (12 items; α= .83); and positi veurgenc y(14 items; α=.92). All items were measured on a 4-point response scale (1 = strongly disagree, 2 = disagree ,3 = agree, 4 = strongl yag ree). Items wereaver- aged for each trait and higher scores on premeditation and p erse verance represent less impulsivity ,wh ereas higher scores on positi veurgenc y,negati veurgenc y,and sensation seeking represent more impulsivity. An examination of the psycho- metric properties of the measure re vealed that the UPPS-P ex hibited good psychometric properties and is an accurate and valid measure of impulsivity (Lynam et al., 2006).

Marijuana use motives. Marijuana use motiveswe re as- sessed using the 25-item Marijuana Moti ves Questionnaire (MMQ; Simons et al., 1998). The MMQ assesses ﬁve dis- tinct marijuana moti ves: enhancement (ﬁ veitems; α= .86), conformity (ﬁ veitems; α= .89), expansion (ﬁ veitems; α= .92), coping (ﬁve items;α= .89), and social moti ves (ﬁve items; α= .87). Each item is rated on a 5-point Likert-type scale (1 =almost never/never to5 = almost al ways/al ways).

Items wereaveraged for each distinct moti veso that higher scores indicate greater endorsement of a speciﬁc marijuana use moti ve. An examination of the psychometric properties of the measure re vealed that the MMQ exhibited good psy- chometric properties and is an accurate and valid measure of marijuana use moti ves (Zvolensky et al., 2007).

Marijuana protective behavioral strategies. Marijuana PBS use was assessed using the 39-item version of the PBSM (Pedersen et al., 2016) scale. This measure consists of behaviors individuals repor tengaging in while using mari- juana to reduce marijuana-related problems (e.g., “a void us- ing marijuana in public places”). Items are rated on a6-point Likert-type scale (1 = neverto 6 = alwa ys). Items were av eraged so that a higher score indicates higher marijuana PBS use (α = .96).Within their initial psychometric study, Pe dersen and colleagues found support for criterion-related va lidity byﬁnding moderate (negati ve)relationship s between PBS and marijuana outcomes (e.g., cannabis use disorder symptoms) as well as con vergent validity byﬁnding moder - ate (positi ve) associations between the PBSM and the most widel yused PBS measure (the Protecti veBehavior Strategies Sur vey ; Martens et al., 2005) in the alcohol literature.

Marijuana use frequency. Marijuana use frequencywas assessed using ahigh-deﬁnition measure patterned from the Dail yDrinking Questionnaire (Collins et al., 1985). Speciﬁ - call y,each da yof the week wasbroken down into six 4-hour bl ocks of time (midnight–4 A.M., 4 A.M.–8 A.M.,8 A.M.–noon, etc.), and participants were asked to report at which times they used marijuana during a “typical week” in the past 30 days. Wecalculated typical frequency of marijuana use by summing the total number of time blocks for which they reported using during the typical week (range: 0–42). The initial measure has shown adequate reliability and validity in previous research (Dvorak & Da y,2014).

Marijuana-related consequences. Marijuana-related consequences were assessed using achecklist version of the 50-item Marijuana Consequences Questionnaire (MACQ; Simons, et al., 2012). The MACQ assesses eight domains of marijuana consequences experienced in the past 30 days (0 = no,1 = yes): social-interpersonal consequences (six items), impaired control (six items), negati veself-perception (ﬁve items), self-care (nine items), risk behaviors (eight items), academic/occupational consequences (ﬁve items), physical dependence (four items), and blackout use (se ven items). Wesummed all items to create a marijuana-related problems total score reﬂecti veof the number of distinct problems experienced in the past 30 days (α = .96). Previ- ous research supports the test-retest reliability, con vergent, and discriminant validity of the MACQ as a measure of marijuana-related problems (Simons et al., 2012).

Statistical analysis To determine which distal constructs uniquely (i.e., con- trolling for other distal antecedents) relate to marijuana PBS 206JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / MARCH 2017 use and marijuana outcomes, weconducted acomprehensi ve structural equation model (SEM) in which the proposed distal antecedents (i.e., age at ﬁrst use, sex, impulsivity-like traits, and moti ves) we re modeled as predictors of marijuana- related consequences via marijuana PBS use and marijuana use frequenc y(e.g., age at ﬁrst use %marijuana PBS use % marijuana use frequency %marijuana-related consequenc- es). Totest the comprehensi vemodel, structural equation modeling using Mplus 7.4 (Muthén &Muthén, 1998–2012) wa s conducted.

To evaluate overall model ﬁt, weused model ﬁt criteria suggested byHu and Bentler (1999) including the com- parati veﬁt index (CFI) > .95, Tucker–Lewis index (TLI) > .95, root mean square error of approximation (RMSEA) < .06, and standardized root mean square residual (SRMR) < .08. Toreduce the complexity of the model, wefollo wed the item-to-construct balance approach described byLittle and colleagues (2002) bycreating parcels for impulsivity- lik etraits, marijuana PBS use, and marijuana-related consequences. Weﬁrst conﬁrm ed and then extracted a single factor in exploratory factor analyses for each latent construct, sorted the items from highest to lo west factor loadings, and created three to ﬁ vebalanced parcels bypair- ing items with the highest factor loadings with items with the lo westfactor loadings. A supplementary table of the correlations among the parcels and items used as indicators of the latent factors in the model are available from the au- thors on request. We examined the total, direct, and indirect effects of each predictor variable on outcomes using bias-corrected bootstrapped estimates (Efron &Tibshirani, 1993) based on 10,000 bootstrapped samples, which provides apo we rful test of mediation (Fritz & MacKinnon, 2007) and is robust to small departures from normality (Erceg-Hur n &Mirosevich, 2008). Parameters were estimated using maximum likelihood estimation, and missing data were handled using full infor- mation maximum likelihood ,wh ich is more efﬁ cient and has less bias than alternati veprocedures (Enders, 2001; Enders & Bandalos, 2001). Gi ven our large sample size (i.e., large statistical po wer), statistical signiﬁcance was determined by 99% bias-corrected bootstrapped conﬁdence inter vals that do not contain zero. Moreo ver,give nour large statistical po wer we place emphasis on effect sizes of signiﬁcant direct and indirect results.

Results All bi variate correlations and descripti vestatistics among distal antecedents, marijuana PBS use, and marijuana out- comes in the comprehensi vemodel are summarized in Table 1. All total, direct, and indirect associations are shown in Ta- bl e2. After item parceling, the SEM provided an acceptable ﬁt to the data based on most ﬁt indices (CFI = .928; TLI = .919; RMSEA = .046, 90% CI [.045, .047], SRMR = .049). The signiﬁcant model chi-square, ' 2(1, 223) = 6,746.04,p < .001, would suggest poor model ﬁt; ho wever,the model chi-square is highly sensiti veto sample size (Jöresk og& Sörbom, 1993; Kline, 1998).

Marijuana PBS use and marijuana-related outcomes Marijuana PBS use wassigniﬁcantl ynegati vely associated with marijuana use frequency () = -.47) and marijuana-re- lated consequences () = -.26), and marijuana use frequency wa ssigniﬁcantl ypositi vely associated with marijuana-related consequences () = .19). As expected bythese direct associa- tions, marijuan ause frequen cysigniﬁcantl ymediated the as- sociations between marijuana PBS use and marijuana related consequences (indirect )= -.09) accounting for 25.38% of the total effect. TABLE 1. Bi variate correlations among distal antecedents, marijuana PBS use, and marijuana outcomes in the comprehensi vemodel Va riable 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Mor % SD 1. Sex 60% – 2. Age at ﬁrst use .04 16.45 5.99 3. Premeditation .02.05 2.87 0.40 4. Perse verance -.02.04 .47 2.89 0.40 5. Sensation seeking -.24-.03 -.04 .21 2.85 0.47 6. Positi ve urgency -.12-.02-.31 -.32 .19 2.06 0.51 7. Negati veurgency .06-.02 -.25 -.35 .10 .66 2.38 0.49 8. Social moti ves .04 -.04 .03 .02 .06 .18 .13 2.67 1.05 9. Coping moti ves .07-.02 -.02 -.09-.05 .26 .30 .53 2.20 1.05 10. Enhancement moti ves -.04 -.05 .04 .05 .18.02 .05 .49 .28 3.65 0.98 11. Conformity moti ves -.09-.02-.10 -.12 -.05.31 .17 .33 .35 -.07 1.48 0.77 12. Expansion moti ves -.07 -.05 .01 -.04.09 .19 .13 .43 .50 .35 .29 2.44 1.16 13. Marijuana PBS use .14 .10 .22 .19-.02-.14 -.11 -.13 -.22 -.14 -.06-.21 4.15 0.94 14. Marijuana use frequency -.13 -.06-.04-.06 .03 .03 .02 .16 .24 .23-.03.28 -.48 5.76 6.92 15. Marijuana consequences -.14 -.10 -.16 -.17.02.21 .25 .15 .28 .11 .18 .21 -.40 .35 8.91 7.83 Notes: Sexwas coded 0 = male,1 =female. Signiﬁcant correlations are in boldtypeface for emphasis and were determined bya 99% bias-corrected boot- strapped conﬁdence inter val (based on 10,000 bootstrapped samples) that does not contain zero. BRAVOET AL. 207 Direct and indirect ef fects of sex on marijuana-related outcomes Sex was signiﬁcantly positi vely associated with marijua- na PBS use () = .16;women reported higher PBS use) and signiﬁcantly negati vely associated with both marijuana use frequency () = -.07) and marijuana-related consequences () = -.10; women reported lo wer marijuana use frequency and consequences). As could be expected bythese direct associations, marijuana PBS use signiﬁcantly mediated the associations between sex and marijuana use frequency (in- direct )= -.08) accounting for 53.42% of the total effect, and the associations between se xand marijuana-related consequences (indirect )= -.04), accounting for 24.60% of the total effect. Further, the double-mediated association (i.e., sex %marijuana PBS use %marijuana use frequency % marijuana-related problems) was signiﬁcant (indirect ) = -.01), accounting for 8.38% of the total effect.

Direct and indirect ef fects of age at ﬁrs t use on marijuana- re lated outcomes Despite signiﬁcant (although weak) bi variate correla- tions, age at ﬁrst use did not ha vea signiﬁcant direct as- sociation with marijuana PBS use, marijuana use frequenc y, or marijuana-related consequences when examined in the comprehensi vemodel.

Direct and indirect ef fects of impulsivity-like traits on marijuana-related outcomes Of the ﬁve impulsivity-lik etraits, onl ypremeditation () = .18) and perse verance () = .11)were signiﬁcantly as- sociated with marijuana PBS use; ho wever, neither were directly signiﬁcantly associated with marijuana outcomes.

Marijuana PBS use signiﬁcantly mediated the associations between premeditation and marijuana use frequency (indi- rect )= -.08; fully mediated) and the associations between premeditation and marijuana-related consequences (indi- rect )= -.05), accounting for 45.71% of the total effect.

Further, the double-mediated association (i.e., premedita- tion %marijuana PBS use %marijuana use frequency % marijuana-related problems) was signiﬁcant (indirect )= -.02), accounting for 15.67% of the total effect. Similarl y, marijuana PBS use signiﬁ cantly mediated the associations between perse verance and marijuana use frequenc y(in- direct )= -.05) accounting for 57.96% of the total effect, and the associations between perse verance and marijuana- related consequences (indirect )= -.03), accounting for 55.66% of the total effect. Ho wever, the double-mediated association (i.e., perse verance %marijuana PBS use % marijuana use frequenc y% marijuana-related problems) wa s not signiﬁcant. The only signiﬁcant direct association between impulsivity-lik etraits and marijuana outcomes wa s found between negati veurgency and marijuana-related consequences () = .23). Positi ve urgenc yand sensation seeking did not ha veuniquely signiﬁcant direct associa- tions with marijuana PBS use, marijuana use frequenc y,or marijuana-related consequences (Table 2).

Marijuana use motives Coping moti veswa s signiﬁcant lynegati vely associated with marijuana PBS use () = -.22) and signiﬁcantly posi- ti ve ly associated with both marijuana use frequency () = .10) and marijuana-related consequences () = .11). Marijua- na PBS use signiﬁcantly mediated the associations between coping moti ves and marijuana use frequency (indirect )= .10) accounting for 49.83% of the total effect, and the asso- ciations between coping moti ves and marijuana-related con- sequences (indirect )= .06), accounting for 26.82% of the total effect. Further, the double-mediated association (i.e., coping moti ves% marijuana PBS use %marijuana use frequency %marijuana-related problems) was signiﬁcant (indirect )= .02), accounting for 9.04% of the total effect.

Expansion moti veswa s signiﬁcant lynegati vely associ- ated with marijuana PBS use () = -.13) and signiﬁcantly positi vely associated with marijuana use frequency () = .13). Marijuana PBS use signiﬁcantl ymediated the as- sociations between expansion moti ves and marijuana use frequency (indirect )= .06) accounting for 32.49% of the total effect, and the associations between expansion mo- t i ve sand marijuana-related consequences (indirect )= .03), accounting for 47.01% of the total effect. Further, the double-mediated association (i.e., expansion moti ves% marijuana PBS use %marijuana use frequency %mari- juana-related problems) was signiﬁcant (indirect )= .01), accounting for 16.42% of the total effect. Counterintuiti vely, conformity moti veswe re signiﬁcant- ly positi vely associated with marijuana PBS use () = .10) and marijuana-related consequences () = .12) but signiﬁ- cantly negati vely associated with marijuana use frequency () = -.10). Marijuana PBS use signiﬁcantly mediated the associations between conformity moti vesand marijuana use frequency (indirect )= -.05) accounting for 32% of the total effect, and the associations between coping mo- ti ve sand marijuana-related consequences (indirect )= -.03), accounting for 42.62% of the total effect. Ho wever, the double-mediated association (i.e., conformity moti ves % marijuana PBS use %marijuana use frequenc y% marijuana-related problems) was not signiﬁcant. Enhance- ment moti ves had a uniquely signiﬁcant positi vedirect ef- fect on marijuana use frequency () = .08) but did not have uniquely signiﬁcant direct associations with marijuana PBS use and marijuana-related consequences. Social moti ves did not ha veuniquely signiﬁcant direct associations with marijuana PBS use, marijuana use frequenc y,and marijua- na-related consequences. 208JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / MARCH 2017 TABLE 2. Summaryof total, indirect, and direct effects of distal antecedents and marijuana protecti vebehavioral strategy (PBS) use on marijuana outcomes in a comprehensi vemodel Outcomevariables Marijuana Marijuana use Marijuana related PBS use frequency consequences Predictor variable )[99% CI] )[99% CI] )[99% CI] Sex To tal .16 [.10, .22] -.14 [-.20, -.08] -.17 [-.23, -.11] To tal indirect a – –-.08 [-.11, -.05] -.07 [-.09, -.04] Marijuana PBS use ––-.08 [-.11, -.05] -.04 [-.06, -.02] Marijuana use frequency ––––-.01 [-.02, -.00] Marijuana PBS use – marijuana use frequency ––––-.01 [-.02, -.01] Direct .16 [.10, .22] -.07 [-.12, -.01] -.10 [-.16, -.04] Age at ﬁrst use To tal .07 [-.07, .21] -.04 [-.15, .08] -.08 [-.19, .03] To tal indirect a – –-.03 [-.10, .03] -.03 [-.08, .03] Marijuana PBS use ––-.03 [-.10, .03] -.02 [-.05, .02] Marijuana use frequency ––––-.00 [-.01, .01] Marijuana PBS use – marijuana use frequency ––––-.01 [-.02, .01] Direct .07 [-.07, .21] -.00 [-.06, .06] -.06 [-.12, .01] Premeditation To tal .18 [.09, .27]-.04 [-.12, .05] -.10 [-2.10, .10] To tal indirect a – –-.08 [-.13, -.04] -.05 [-.01, -.02] Marijuana use frequency ––––.01 [-1.27, .37] Marijuana PBS use – marijuana use frequency ––––-.02 [-.03, -.01] Direct .18 [.09, .27].05 [-.03, .12] -.05 [-.15, .05] Pe rse verance To tal .11 [.01, .20]-.09 [-.18, .01] -.05 [-.15, .05] To tal indirect a – –-.05 [-.10, -.00] -.04 [-.08, -.01] Marijuana PBS use ––-.05 [-.10, -.00] -.03 [-.05, -.00] Marijuana use frequency ––––-.01 [-.02, .01] Marijuana PBS use – marijuana use frequency ––––-.01 [-.02, .00] Direct .11 [.01, .20]-.04 [-.12, .05] -.01 [-.10, .09] Sensation seeking To tal .01 [-.06, .09] -.01 [-.08, .07] -.02 [-.11, .06] To tal indirect a – –-.01 [-.04, .03] -.01 [-.01, .01] Marijuana PBS use ––-.01 [-.04, .03] .00 [-.02, .02] Marijuana use frequency ––––-.00 [-.42, .34] Marijuana PBS use – marijuana use frequency –––– -.00 [-.01, .01] Direct .01 [-.06, .09] -.00 [-.07, .07] -.02 [-.10, .06] Po siti ve urgency To tal -.02 [-.12, .09] -.03 [-.13, .08] -.08 [-.19, .03] To tal indirect a – –.01 [-.04, .06] .00 [-.04, .04] Marijuana PBS use ––.01 [-.04, .06] .00 [-.02, .03] Marijuana use frequency ––––-.01 [-.02, .01] Marijuana PBS use – marijuana use frequency ––––.00 [-.01, .01] Direct -.02 [-.12, .09] -.03 [-.12, .06] -.08 [-.18, .03] Negati veurgency To tal .05 [-.05, .16] -.06 [-.16, .04] .21 [.09, .32] To tal indirect a – –-.03 [-.07, .02] -.02 [-.06, .02] Marijuana PBS use ––-.03 [-.07, .02] -.01 [-.04, .01] Marijuana use frequency ––––-.01 [-.02, .01] Marijuana PBS use – marijuana use frequency ––––-.01 [-.01, .01] Direct .05 [-.05, .16] -.03 [-.13, .06] .23 [.13, .33] Social moti ves To tal -.03 [-.11, .06] .00 [-.09, .09] -.02 [-.11, .07] To tal indirect a – –.01 [-.03, .05] .01 [-.04, .04] Marijuana PBS use ––.01 [-.03, .05] .01 [-.02, .03] Marijuana use frequency ––––-.00 [-.02, .01] Marijuana PBS use – marijuana use frequency ––––.00 [-.01, .01] Direct -.03 [-.11, .06] -.01 [-.09, .07] -.03 [-.11, .05] Table continued BRAVOET AL. 209 TABLE 2.Continued Outcomevariables Marijuana Marijuana use Marijuana related PBS use frequency consequences Predictor variable )[99% CI] )[99% CI] )[99% CI] Coping moti ves To tal -.22 [-.30, -.14] .20 [.12, .29] .21 [.13, .29] To tal indirect a – –.10 [.07, .14] .09 [.06, .13] Marijuana PBS use ––.10 [.07, .14] .06 [.03, .08] Marijuana use frequency ––––.02 [.00, .04] Marijuana PBS use – marijuana use frequency ––––.02 [.01, .03] Direct -.22 [-.30, -.14] .10 [.03, .18] .11 [.04, .19] Enhancement moti ves To tal -.06 [-.16, .03].11 [.03, .19].04 [-.05, .12] To tal indirect a – –.03 [-.02, .08] .04 [.00, .07] Marijuana PBS use ––.03 [-.02, .08] .02 [-.01, .04] Marijuana use frequency ––––.02 [-.00, .03] Marijuana PBS use – marijuana use frequency ––––.01 [-.00, .01] Direct -.06 [-.16, .03].08 [.01, .16]-.00 [-.08, .08] Conformity moti ves To tal .10 [.02, .19] -.15 [-.23, -.07] .06 [-.04, .16] To tal indirect a – –-.05 [-.09, -.01] -.05 [-.09, -.02] Marijuana PBS use ––-.05 [-.09, -.01] -.03 [-.05, -.00] Marijuana use frequency ––––-.02 [-.03, -.00] Marijuana PBS use – marijuana use frequency ––––-.01 [-.02, .00] Direct .10 [.02, .19] -.10 [-.18, -.03] .12 [.03, .21] Expansion moti ves To tal -.13 [-.19, -.06] .19 [.11, .27] .07 [.00, .14] To tal indirect a – –.06 [.03, .10] .07 [.04, .10] Marijuana PBS use ––.06 [.03, .10] .03 [.01, .06] Marijuana use frequency ––––.02 [.01, .04] Marijuana PBS use – marijuana use frequency ––––.01 [.00, .02] Direct -.13 [-.19, -.06] .13 [.06, .20] .00 [-.07, .07] PBS use To tal ––-.47 [-.53, -.41] -.35 [-.41, -.28] To tal indirect (marijuana use frequency) ––––-.09 [-.13, -.05] Direct ––-.47 [.53, -.41] -.26 [-.34, -.17] Notes: Signiﬁcant associations are in boldtypeface for emphasis and were determined bya99% bias-corrected unstandardized bootstrapped conﬁ- dence inter val(based on 10,000 bootstrapped samples) that does not contain zero. aReﬂects the combined indirect associations via marijuana PBS use ,marijuana use frequenc y,marijuana PBS use via marijuana use frequenc y. Discussion The present study examined the direct effects of se veral risk and protecti vefactors of marijuana related outcomes, and examined marijuana PBS use as a potential mediator of the associations between these risk/protecti vefactors and marijuana-related outcomes. As the legal status of marijuana use has begun to shift, researchers ha vebegun to question wh ich scientiﬁc ﬁndings that ha vebeen explored with other substances will translate to marijuana. Forex ample, there is a large body of literature establishing a robust negati ve relationship between PBS use and alcohol use and related consequences (Pearson, 2013; Prince et al., 2013); ho wever, research on PBS use aimed at reducing marijuana outcomes is in its infanc y.As such, it was an important ﬁrst step to establish that our ﬁndings are consistent with Pedersen et al.

(2016) in that marijuana PBS use wasassociated with lo wer marijuana usefrequenc yand experiencing fe wermarijuana- related consequences. Next, a principle aim of the present study was to exam- ine whether certain risk and protecti vefactors for alcohol outcomes ser vethe same function for marijuana outcomes through examining direct associations in the context of a comprehensi veSEM. Wefound signiﬁcant direct effects for (a) sex, coping moti ves, conformity moti ves, and marijuana PBS use for both marijuana use frequenc yand marijuana-re- lated consequences ;(b) enhancement mot ives and expansion moti ves for marijuana use frequency but not marijuana- related consequences; and (c) negati veurgenc yfor marijua- na-related consequences but not marijuana use frequenc y. 210JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / MARCH 2017 Fo r age at ﬁrst use, premeditation, perse verance, sensation seeking, positi veurgenc y,and social moti ves, neither the direct effects to marijuana use frequency nor marijuana- related consequences were signiﬁcant. This pattern of ﬁnd- ings could be explained, in part, byour implementation of a comprehensi vemodel (Bra voet al., 2016). Indeed, for a direct effect to remain signiﬁcant it would need to explain unique variance in the outcome over and abo vethe other predictor variables. Ho wever,give nour large sample size, we had sufﬁcient po wer to run a comprehensi vemodel, which has the advantage of testing man yeffects simultaneousl yand providing insights into the unique strength of each predictor va riable while controlling for the other predictor variables.

To summarize our direct effects, nearly all effects were in the expected direction with risk factors being associated with greater marijuana use frequency and more marijuana- related consequences and protecti vefactors being associated with less frequent marijuana use and fe wermarijuana-related consequences. Consistent with previous research,conformity moti veswa s actually associated with less marijuana use (Zvolensky et al., 2007). Moreo ver,we found that risk fac- tors for marijuana use (e.g., coping moti ves) we re associated with using fe wer marijuana PBS and protecti vefactors for marijuana use (e.g., premeditation) were associated with us- ing more marijuana PBS. Importantl y,marijuana PBS use partially mediated the associations between se veral risk/protecti vefactors and marijuana-related outcomes. Speciﬁcall y,marijuana PBS fully or partially mediated the effects of sex, premedita- tion, perse verance, coping moti ves, enhancement moti ves, conformity moti ves, and expansion moti ves on marijuana outcomes. These results are consistent with recent replication attempts in the alcohol PBS literature demonstrating alcohol PB Sus eas amediator betwee nag eat drinking onset, drink- ing moti ves, impulsivity-like traits, and alcohol outcomes (Bra voet al., 2015, 2016). Takentogether ,marijuana PBS is not only a robust predictor of marijuana use frequency and marijuana-related consequences but also a good candidate to be considered as a mechanism bywhich marijuana us- ers moderate their marijuana use and attenuate their risk of ex periencing marijuana-related consequences.

Clinical implications The present study has a number of important clinical implications. Based on the current sample, wecan identify certain characteristics that might be particularl ygood targets for marijuana PBS-based inter vention efforts. Forex ample, our results indicate that female college students tend to use marijuana less often, experience fe wermarijuana-related consequences, and use more marijuana PBS. On the one hand ,these results suggest that female college students seek- ing help for their marijuana use may be recepti veto mari- juana PBS in the context of a marijuana inter vention. On the other hand ,it ma ybe men whowo uld beneﬁ tmos tfro m increasing their marijuana PBS use, gi ven the strong nega- ti ve association between marijuana PBS use and marijuana use and related consequences. Similarl y,those who tend to use marijuana for coping moti vesor expansion moti vesma y also beneﬁ t from an inter vention targeting marijuana PBS use.

Limitations The present study should be considered in light of its limitations. First, this was a cross-sectional study, which precludes an ycausal inferences to be drawn from the results.

Second ,the present sample, although large, ma ynot be rep- resentati veof all college student marijuana users nationwide as the sample was collected from participant pools from 11 uni versities. Third, unlike many alcohol studies, welooked at frequency of marijuana use but not quantity of marijuana use. This decision was made in part because of the chal- lenges of standardizing the quantity of marijuana use. Last, marijuana PBS use as measured bythe PBSM is focused more on avoidance strategies than the most common alcohol measure (i.e., the Protecti veBehavior Strategies Sur vey; Martens et al., 2005), which ma ywe aken direct comparisons between marijuana PBS studies and alcohol PBS studies.

Conclusions The present study identiﬁ ed marijuana PBS as a robust predictor of marijuana use frequency and marijuana-related consequences, as well as establishing preliminary evidence for marijuana PBS as a mediator of a variety of risk/protec- ti ve factors of marijuana outcomes. These results ha vea va riety of clinical implications, including supporting the potential beneﬁt of a marijuana PBS-focused inter vention for coll egestudent marijuana users. The present study iden- ti ﬁed that most (but not all) distal antecedents of marijuana outcomes were signiﬁcant in the comprehensi vemodel, suggesting that future studies should continue to elucidate total and unique effects of speciﬁc antecedents on marijuana- related outcomes. Although the present study provides some preliminary insights into the role of marijuana PBS use in the relationship between distal antecedents and marijuana use frequenc yand marijuana-related consequences, more re- search is needed to replicate and extend the current ﬁndings.

References Barnett, N.P., Murph y,J.G., Colb y,S. , & Monti, P. (2007). Ef- ﬁcacy of counselor vs. computer-deli vered inter vention with mandated college students. Addictive Behaviors, 32, 2529–2548. doi:10.1016/j.

addbeh.2007.06.017 Bra vo, A. J., Prince, A., & Pearson, R. (2015). Does the how mediate the why? A multiple replication examination of drinking mo- BRAVOET AL. 211 tive s, alcohol protecti vebehavioral strategies, and alcohol outcomes.

Jo urnal of Studies on Alcohol and Drugs, 76, 872–883. doi:10.15288/ jsad.2015.76.872 Bra vo, A. J., Prince, A., & Pearson, R. (2016). A multiple replica- tion examination of distal antecedents to alcohol protecti vebehav- ioral strategies. Journal of Studies on Alcohol and Drugs, 77, 958–967.

doi:10.15288/jsad.2016.77.958 Caldeira, K. , Arria, A. , O’Grady, K. E., Vincent, K. B., & Wish, E.

D. (2008). The occurrence of cannabis use disorders and other cannabis- related problems among ﬁrst-year college students. Addictive Behaviors, 33, 397–411. doi:10.1016/j.addbeh.2007.10.001 Chen, C.-Y., Stor r,C. L., &Anthon y,J.C. (2009). Early-onset drug use and risk for drug dependence problems. Addictive Behaviors, 34,319–322.

doi:10.1016/j.addbeh.2008.10.021 Collins, R. L., Parks, G. A., &Marlatt, G. A. (1985). Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol. Journal of Consulting and Clinical Psycholo gy, 53, 189–200. doi:10.1037/0022-006X.53.2.189 Cyders, A., Smith, G. T., Spillane, N.S., Fischer, S., Annus, A. , & Peterson, C. (2007). Inte gration of impulsivity and positi vemood to predict risk ybehavior: De velopment and validation of amea- sur eof positi veurgenc y.Psychological Assessment, 19, 107–118.

doi:10.1037/1040-3590.19.1.107 Dvorak, R. D., & Day,A. (2014). Marijuana and self-regulation: Exam- ining likelihood and intensity of use and problems. Addictive Behaviors, 39, 709–712. doi:10.1016/j.addbeh.2013.11.001 Dvorak, R. D.,Pe arson, R., Neighbors, C., & Martens, P.(2015).

Fi tting in and standing out: Increasing the use of alcohol protecti ve behavioral strategies with a deviance regulation inter vention. Journal of Consulting and Clinical Psycholo gy,83, 482–493. doi:10.1037/ a0038902 Dvorak, R. D.,Pe arson, R., Neighbors, C., Martens, P.,Ste venson, B.

L., & Kuvaas,N.J.(2016). A road pa ved with safe intentions: Increas- ing intentions to use alcohol protecti vebehavioral strate gies via Devi- ance Regulation Theor y.Health Psycholo gy,35, 604–613. doi:10.1037/ hea0000327 Efron, B., & Tibshirani, R. J.(1993). An introduction to the bootstrap. Boca Raton, FL: CRC Press.

Enders, C. K. (2001). The impact of nonnormality on full informa- tion maximum-likelihood estimation for structural equation models with missing data. Psychological Methods, 6,352–370.

doi:10.1037/1082-989X.6.4.352 Enders, C. K., & Bandalos, D.L. (2001). The relati veperformance of full information maximum likelihood estimation for missing data in structur - al equation models. Structural Equation Modeling: AMultidisciplinary Jo urnal, 8, 430–457. doi:10.1207/S15328007SEM0803_5 Erceg-Hurn, D. , &Mirosevich, V. (2008). Moder nrobust sta- tistical methods: An easy w a y to maximize the accurac yand po we rof your research. American Psychologist, 63, 591–601.

doi:10.1037/0003-066X.63.7.591 Fa rm er, R. F., Ko sty, D.B., Seele y,J.R., Duncan, S. C., Lynsk ey, T. , Rohde, P., . . . Lewinsohn, P. (2015). Natural course of can- nabis use disorders. Psychological Medicine, 45, 63–72. doi:10.1017/ S003329171400107X Fr itz, S., &Mackinnon, D.P.(2007). Required sample size to detect the mediated effect. PsychologicalScience,18, 233–239.

doi:10.1111/j.1467-9280.2007.01882.x Hu, L., & Bentler, P. (1999). Cutoff criteria for ﬁt inde xes in co vari- ance structure analysis: Con ventional criteria versus new alternati ves.

Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55.

doi:10.1080/10705519909540118 Johnston, L. D., O’Malle y,P. , Bachman, J.G., Schulenberg, J.E., & Miech, R. A. (2015). Monitoring the Future national survey results on drug use, 1975–2014:

Volume 2, Colle gestudents and adults ages 19–55. Ann Arbor, MI: Institute for Social Research, The Uni versity of Michigan.

Jöresk og, K. G., & Sörbom, D.(1993). LISREL 8: User’s guide. Chicago, IL: Scientiﬁc Software.

Kaiser, A. J.,Milich, R., Lynam, D.R., &Charnigo, R. J.(2012). Negati ve urgenc y,distress tolerance, and substance abuse among college students.

Addictive Behaviors, 37, 1075–1083. doi:10.1016/j.addbeh.2012.04.017 Kline, R. B.(1998). Principles and practice of structural equation model- ing. New York, NY: Guilford Press.

Larimer , E., Lee, C. , Kilmer ,J. R., Fabiano, P. , Stark, C.

B. , Geisner, I. , . . . Neighbors, C. (2007). Personalized mailed feedback for college drinking pre vention: Arandomized clinical trial. Journal of Consulting and Clinical Psycholo gy, 75, 285–293.

doi:10.1037/0022-006X.75.2.285 Little, T.D.,Cunningham, W.A., Shahar ,G., &Wi daman, K. F.(2002). To parcel or not to parcel: Exploring the question, weighing the merits.

Structural Equation Modeling: AMultidisciplinary Journal, 9, 151–173.

doi:10.1207/S15328007SEM0902_1 Ly nam, D.R., Smith, G. T., Whiteside, S. P., & Cyders, A. (2006). The UPPS-P: Assessing ﬁ ve personality pathways to impulsive behavior.

We st Laf ayette, IN: Purdue Uni versity.

Martens, P.,Fe rrier, A. G., Sheeh y, J., Corbett, K., Anderson, D.A., & Simmons, A. (2005). De velopment of the Protecti veBehavioral Strat- eg ies Sur vey.Jo urnal of Studies on Alcohol, 66, 698–705. doi:10.15288/ jsa.2005.66.698 Murph y,J.G., Dennhardt, A. A., Skidmore, J.R., Borsari, B.,Barnett, N.P., Colb y,S. , &Martens, P.(2012). Arandomized controlled trial of a behavioral economic supplement to brief moti vational inter ventions for college drinking. Journal of Consulting and Clinical Psycholo gy,80, 876–886. doi:10.1037/a0028763 Muthén, L. K., &Muthén, B.O. (1998–2012). Mplus user’sguide (7th ed.).

Los Angeles, CA: Authors.

Pe arson, R. (2013). Use of alcohol protecti vebehavioral strategies among college students: A critical revie w.Clinical Psycholo gyRevie w, 33, 1025–1040. doi:10.1016/j.cpr.2013.08.006 Pe arson, R., Liese, B.S., Dvorak, R. D., & Marijuana Outcomes Study Te am. (2017). College student marijuana in volvement: Perceptions, use, an dconsequences across 11 coll egecampuses. Addictive Behaviors, 66, 83–89. doi:10.1016/j.addbeh.2016.10.019 Pe dersen, E. R., Hummer, J.

F., Rinker, D.V.,Tray lor, Z. K., & Neighbors, C. (2016). Measuring protecti vebehavioral strategies for marijuana use among young adults. Journal of Studies on Alcohol and Drugs, 77, 441–450. doi:10.15288/jsad.2016.77.441 Prince, A., Care y,K. B., & Maisto, S. A. (2013). Protecti vebehavioral strategies for reducing alcohol in volvement: A review of the method- ological issues. Addictive Behaviors, 38, 2343–2351. doi:10.1016/j.

addbeh.2013.03.010 Robinson, J. , Ladd, B.O., & Anderson, K. G. (2014). When you see it, let it be: Urgenc y,mindfulness and adolescent substance use. Addictive Behaviors, 39, 1038–1041. doi:10.1016/j.addbeh.2014.02.011 Simons, J., Correia, C. J., Care y,K. B., & Borsari, B.E. (1998). Validating a ﬁ ve-factor marijuana moti vesmeasure: Relations with use, problems, and alcohol moti ves. Journal of Counseling Psycholo gy, 45, 265–273.

doi:10.1037/0022-0167.45.3.265 Simons, J.S., Dvorak, R. D.,Merrill, J.E., &Read ,J. P. (2012). Dimensions and se verity of marijuana consequences: De velopment and validation of the Marijuana Consequences Questionnaire (MACQ). Addictive Behav- iors, 37, 613–621. doi:10.1016/j.addbeh.2012.01.008 Smith, G. T.,Fi scher ,S., Cyders, A., Annus, A. , Spillane, N.

S., & McCarth y,D. (2007). On the validity and utility of dis- criminating among impulsivity-like traits. Assessment, 14,155–170.

doi:10.1177/1073191106295527 212JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / MARCH 2017 Wardell, J. D.,Strang, N. , &Hendershot, C. S. (2016). Negati veurgenc y mediates the relationship between childhood maltreatment and problems with alcohol and cannabis in late adolescence. Addictive Behaviors, 56, 1–7. doi:10.1016/j.addbeh.2016.01.003 Whiteside, S. P., & Lynam, D.R. (2001). The Five Factor Model and impul- sivity: Using a structural model of personality to understand impulsiv- ity.

Personality and Individual Dif fere nces, 30, 669–689. doi:10.1016/ S0191-8869(00)00064-7 Zvolensk y, J., Vu janovic, A. A., Bernstein, A., Bonn-Miller, O., Marshall, E. C., & Leyro, T. (2007). Marijuana use moti ves: A conﬁrm atory test and evaluation among young adult marijuana users.

Addictive Behaviors, 32, 3122–3130. doi:10.1016/j.addbeh.2007.06.010