Research Article 49 Ved ABSTRACT College student cannabis use is at an all-time high. Although frequent heavy cannabis use is related to cannabis problems, perceived risk of cannabis use is rapidly decreasing. Yet, it is unknown whether specific domains of risk perceptions (general and domain-specific risk, risk to others and personal risk) are related to more cannabis use or related problems. Thus, among 130 undergraduates who reported past-month cannabis use, the present study conducted secondary analyses to test whether, for both perceived risk to others and perceived personal risk: (1) general perceived risk was associated with cannabis-related outcomes (i.e., use, negative consequences, cannabis use disorder (CUD) symptoms, motivation to change), (2) seven specific domains of perceived risk were related to cannabis outcomes, and (3) domain-specific perceived risk was related to cannabis use frequency. General perceived risk to others was negatively associated with cannabis use frequency whereas general perceived personal risk was positively associated with cannabis-related negative consequences, CUD symptoms, and importance and readiness to change. Greater legal and withdrawal/dependence risks were uniquely related to several outcomes (e.g., CUD symptoms). Participants who used cannabis frequently perceived more personal risk in most risk domains and less general risk to others than those who used infrequently. Findings suggest personal risk is an important component to consider when assessing perceived risk of cannabis use and focusing on both general and domain-specific risks may provide valuable insight for future prevention and intervention efforts. Key words: = perceived risk; undergraduates; cannabis; motivation to change; college students Cannabis use among US undergraduate students is reaching some of the highest levels ever recorded. In 2021, 40.3% of undergraduates reported past-year and 24.2% past-month cannabis use (Patrick et al., 2019). Daily or near daily cannabis use rates also remain high (Patrick et al., 2019), which is concerning given heavier use is associated with increased likelihood and severity of unwanted physical and psychosocial outcomes, including cannabis use disorder (CUD; Kirstyn N. Smith-LeCavalier1, Paige M. Morris2, Mary E. Larimer3, Julia D. Buckner2, Katherine Walukevich-Dienst3 1University of Washington, Department of Psychology 2Louisiana State University, Department of Psychology 3University of Washington, Department of Psychiatry and Behavioral Sciences Cannabis 2024, Volume 7 (1) © Author(s) 2024 researchmj.org 10.26828/cannabis/2024/000194 General and Domain-Specific Perceived Risk Demonstrate Unique Associations with Cannabis Use, Negative Outcomes, and Motivation to Change among Undergraduate Students Corresponding Author: Kirstyn N. Smith-LeCavalier, University of Washington, Department of Psychology, 119A Guthrie Hall, Box 351525, Seattle, WA 98195, USA. Email: kirstynl@uw.edu. Cannabis, A Publication of the Research Society on Marijuana 50 Caldeira et al., 2008; Gunn et al., 2020), poorer mental health (Keith et al., 2015), and worse academic outcomes (Suerken et al., 2016). Despite known risks of cannabis, perceived risk of regular use (i.e., one’s perceptions of the negative effects of using substances; Danseco et al., 1999) has rapidly decreased over the past 20 years and is at some of the lowest levels ever recorded among undergraduates (Lipari & Jean- Francois, 2016). This is particularly notable compared to trends in alcohol and tobacco risk perceptions, which remain relatively stable (Waddell, 2022; Lipari & Jean-Francois, 2016). Rapid decrease in perceived risk of cannabis use is especially troubling, as perceived risk is a critical determinant of health-related behavior (Janz & Becker, 1984; Kasten et al., 2019), contributes to motivation to change risky behaviors (Kasten et al., 2019), and prospectively predicts changes in cannabis use (Azofeifa et al., 2016; Bachman et al., 1998; Bachman et al., 1988; Compton et al., 2016). Among undergraduates, perceived risk may protect against initiating cannabis use (D’Silva et al., 2020; Hanauer et al., 2021). However, few undergraduates report believing regular cannabis use confers “great risk” of harm (Lipari & Jean- Francois, 2016), and some evidence suggests more frequent cannabis use is associated with decreases in risk perception over time (Grevenstein et al., 2015). Students who experience negative consequences due to their cannabis use still report low perceived risk, with no difference in risk perception between those who had and had not experienced certain cannabis- related negative consequences (Kilmer et al., 2007). As such, a more detailed understanding of how undergraduates conceptualize risk of cannabis is needed, particularly among those using frequently. Existing research has examined perceived risk to others (i.e., how much others risk harming themselves from using cannabis) and perceived personal risk (i.e., how much an individual risks harming themselves at their current rate of cannabis use). Most population-based studies assess perceived risk to others (Azofeifa et al., 2016; Bachman et al., 1998; Bachman et al., 1988; Compton et al., 2016; Grevenstein et al., 2015; Lipari & Jean-Francois, 2016) whereas research on perceived personal risk is limited. Some studies found perceived personal risk is higher among undergraduates who use more frequently compared to those who use less frequently (O'Callaghan et al., 2006) and is cross-sectionally associated with cannabis-related negative consequences among adults who use cannabis (Magnan & Ladd, 2019). However, other studies did not find associations between perceived personal risk and use frequency (Kilmer et al., 2007; Magnan & Ladd, 2019) or the experience of negative consequences (Kilmer et al., 2007). Given these inconsistencies, additional research is needed to better understand and explain discrepancies. This is particularly important when considering how perceived risk may be useful to inform cannabis prevention and intervention programs, and how perceived risk to others versus personal risk may maintain varying salience for individuals. There is also considerable variability in perceived risk across different domains of risk (e.g., physical harm, dependence, legal risks; O'Callaghan et al., 2006). Although only 30.4% of undergraduates reported believing regular cannabis use puts the user at great risk for harm generally, over 50% reported regular use puts the user at great risk for physical dependence, finding it hard to stop using, and performing worse at school/work. Thus, undergraduates may perceive specific aspects of cannabis use as risky, which may obfuscate effects on use patterns when only examining general risk. The Current Study The current study sought to expand prior work (Kilmer et al., 2007; Magnan & Ladd, 2019; O'Callaghan et al., 2006) on general vs domain- specific perceived risk to self and others. First, we examined associations between general perceived risk to others and general perceived personal risk with cannabis outcomes (i.e., past 3-month cannabis use frequency, cannabis-related negative consequences, CUD symptoms, motivation to change). We hypothesized general perceived risk to others would be negatively associated with cannabis use, negative consequences, and CUD symptoms, general perceived personal risk would be positively associated with these outcomes, and both variables would be positively associated with motivation to change. Second, we tested whether seven domains of perceived risk (i.e., productivity, Perceived Risk, Cannabis Use, and Outcomes 51 lower energy, memory loss or cognitive impairment, problems at school/work, physical health problems, legal problems, dependence/withdrawal) to others and self were cross-sectionally associated with general perceived risk and outcomes. Consistent with prior work (O'Callaghan et al., 2006), we hypothesized dependence/withdrawal and problems at school/work would emerge as significant predictors of cannabis outcomes. Third, as some prior work found differences in perceived risk by use frequency (e.g., Okaneku et al., 2015), we tested whether domains of risk differed by use frequency. Compared to students who use less than weekly, undergraduates who use cannabis weekly or more experience more negative consequences and CUD symptoms (Buckner et al., 2008; Burdzovic Andreas et al., 2021) and are more likely to meet criteria for CUD (Burdzovic Andreas et al., 2021). Thus, we hypothesized participants who engaged in cannabis use weekly or more (compared to less frequently) would rate perceived personal risk domains as higher, but risk to others as lower. The present aims were tested through secondary data analyses from a study that developed and tested problem-focused personalized feedback (PFI) against brief personalized normative feedback (PNF; Morris & Buckner, 2023; Walukevich-Dienst et al., 2021; Walukevich-Dienst et al., 2019). Neither domains of risk nor baseline associations between risk domains and outcomes were examined as part of the parent study primary aims. Participants who received an extended problem-focused intervention were asked to reflect on and rate their perceived risk of cannabis as part of the intervention.1 METHODS Participants Participants were from a sample of 268 undergraduates recruited for the parent intervention trial (Morris & Buckner, 2023; Walukevich-Dienst et al., 2021; Walukevich- Dienst et al., 2019). For the parent study, eligible participants were current undergraduate students at Louisiana State University who reported past-month cannabis use and at least one cannabis-related problem in the past three months. The current study utilized baseline data from 130 undergraduates (47.8% of total sample, 76.2% female, Mage=19.8 years, SD=1.3) who completed questions about domains of perceived risk as part of their intervention and passed attention check questions (described in Procedures below). Of participants, 73.1% identified as non-Hispanic/Latin White, 14.6% Black, 2.3% Asian, and 3.1% multiracial; 6.9% Hispanic/Latin. Procedures Participants were recruited through the psychology department’s online research pool or on-campus flyers. The parent study was advertised as a two-part study on cannabis use rather than an intervention study to recruit participants with a range of motivation to change. Interested participants first completed an online screening survey to determine eligibility. Eligible participants were immediately directed to the online baseline survey and randomized to the online PFI condition or PNF-only condition. The analytic sample for the current study includes participants randomized to the PFI condition, as only PFI participants answered questions about domain-specific perceived risks. The PFI condition included PNF on cannabis use and related problems and brief psychoeducation modules on seven empirically informed domains of risk (see Measures). Upon starting each module, participants were asked to rate domain- specific perceived risk (see Measures). Participants also indicated which of 10 DSM-5 symptoms of CUD they experienced in the past year during the dependence/withdrawal module.2 Participants received personalized feedback on CUD based on number of endorsed CUD symptoms. Intervention modules were presented in a randomized order to control for order presentation effects. More information about the intervention and procedures can be found in Walukevich-Dienst et al. (2019) and Walukevich- Dienst et al. (2021). 1Intervention findings indicated no main effect of condition on cannabis use frequency, consequences, or ratings of general perceived risk (Walukevich-Dienst, 2019). Further information can be found in Morris & Buckner, 2023, Walukevich-Dienst et al., 2021, and Walukevich-Dienst et al., 2019. 2 Continued use despite having persistent or recurrent social and interpersonal problems was not included due to a programming error. Cannabis, A Publication of the Research Society on Marijuana 52 Psychology pool participants received research credits and non-psychology pool participants were compensated $10 for baseline and $20 for follow-up. The study was approved by the university's institutional review board and we obtained a Certificate of Confidentiality from the National Institute of Mental Health. Informed consent was obtained prior to data collection and all procedures maintained adherence to APA ethical guidelines for research with human subjects (Sales & Folkman, 2000). Measures Marijuana Use The Marijuana Use Form (MUF; Buckner et al., 2007) is an 11-item measure used to assess past 3-month cannabis use frequency (0=none or less, 6=3 or 4 times a week, 10=21 times per week or more). In addition to the categorical MUF outcome score, a categorical measure of use frequency was created to test whether perceived risk differed between participants who used frequently (i.e., once a week or more) or infrequently (i.e., less than once a week). Marijuana Problems The modified 30-item Marijuana Problems Scale (Stephens et al., 2000; Walukevich-Dienst et al., 2019) assessed past 3-month cannabis- related problems. Participants rated each problem from 0 (no problem) to 2 (serious problem) and items were converted to a count score of cannabis-related problems. The 30-item modified version demonstrated excellent internal consistency (α=0.96). Perceived Risk of Cannabis Use General perceived risk was measured using the perceived risk item from the Monitoring the Future Project (Schulenberg et al., 2021) which was modified to specify using “regularly” as using cannabis once a week or more per prior work (O'Callaghan et al., 2006). Participants were asked to rate general perceived risk to others (i.e., “How much do you think people risk harming themselves physically or in other ways if they use marijuana regularly [once a week or more]?”) and general perceived personal risk (i.e., “How much do you think you risk harming yourself physically or in other ways if you use marijuana at your current rate of use?”) from 1 (no risk) to 4 (great risk). Additionally, using the same scale, participants rated domain-specific perceived risk to others (e.g., “How much do you think people risk having lower energy if they use marijuana regularly?”) and domain-specific perceived personal risk (e.g., How much do you think you risk having lower energy if you use marijuana at your current rate of use?”) for all seven domains (i.e., productivity, lower energy, cognitive impairment, problems at school/work, physical health problems, legal problems, dependence and withdrawal). Domains were empirically informed through prior work identifying areas of low perceived risk (O'Callaghan et al., 2006) and frequent cannabis-related problems among undergraduates (Buckner et al., 2010). Cannabis Use Disorder (CUD) Symptoms CUD Symptoms were measured by asking participants whether they had experienced (0=no, 1=yes) 10 different symptoms of CUD in the past year.2 Responses were converted to a count score of the ten items (α=0.77). Symptoms were derived from DSM-5 criteria of CUD (e.g., “In the past year, have you used marijuana in larger amounts or for longer periods of time than you meant to?”). Number of CUD symptoms was significantly, positively associated with past 3-month cannabis use frequency (r=.44, p<.001) and negative consequences (r=.52, p<.001). Motivation to Change Cannabis Use Motivation to Change Rulers (Buckner et al., 2016) were used to assess readiness (0=not ready to change to 10=trying to change), importance (0=not important to 10=very important), and confidence (0=not at all confident to 10=most confident) to change. Rulers were based on work by Miller and Rollnick (2013) and shown to be associated with changes in cannabis use in prior work (Gates et al., 2012; Walukevich-Dienst et al., 2021). To detect careless responding, three attention check questions were included in both Perceived Risk, Cannabis Use, and Outcomes 53 baseline and follow-up surveys (e.g., “Please select ‘strongly agree’ as your answer to this question”). Participants (n=2) were excluded from data analysis if they failed attention check by answering two or more attention check questions incorrectly (Meade & Craig, 2012). Data analyses Analyses were conducted using SPSS version 29. First, we examined descriptive statistics and bivariate correlations among measures. Second, we conducted 14 two-step hierarchical multiple regression analyses for each independent variable (IV): (1) perceived risk to others domains and (2) perceived personal risk domains on the following dependent variables (DV): general perceived risk to others, general perceived personal risk, cannabis use frequency, cannabis-related negative consequences, CUD symptoms, and readiness, importance, and confidence to change. Notably, both independent variables were only examined as predictors of their respective general perceived risk DVs. In step one, sex assigned at birth and age were entered as covariates. Use frequency was also entered as a covariate in step one for all models except the use frequency model. In step two, the seven risk domains were entered simultaneously as IVs. We conducted separate models for each DV and computed squared semi-partial correlations (sr2) as effect size indices. Third, we used a one- way analysis of covariance (ANCOVA) model to test differences in perceived risk by use frequency, controlling for age and sex assigned at birth, using a Bonferoni-corrected p-value (<.003) to correct for multiple comparisons. RESULTS Descriptive Statistics and Bivariate Correlations On average, participants used cannabis approximately twice per week and experienced 8.00 negative consequences (SD=5.08) in the past 3-months. Average use was comparable to the defined “regular use” frequency (i.e., once or more a week) specified for perceived risk to others. Importance (M=4.6, SD=3.03) and readiness (M=3.25, SD=3.12) to change were low, whereas confidence to change was high (M=8.18, DS=2.42). Nearly 85% of participants reported no-to-slight perceived risk of regular use to others, whereas nearly 94% reported no- to-slight perceived personal risk. Risk to others and self was highest for legal problems and lowest for physical health problems. On average, participants rated risk to others as having slight-to-moderate risk across domains, whereas personal risk was rated as no-to-slight risk across domains. Descriptive statistics are provided in Table 1. Correlations between use frequency, negative consequences, CUD symptoms, and general and domain-specific perceived risk are displayed in Table 2. Both general perceived personal risk and the majority of perceived personal risk domains were significantly, positively associated with use frequency, negative consequences, and CUD symptoms. Perceived personal risk variables were not associated with most risk to others variables, with a few exceptions. Only a few perceived risk to others variables were associated with use frequency, consequences, and CUD symptoms. For example, use frequency was significantly, negatively associated with legal and dependence/withdrawal risk to others domains. For motivation to change variables, general perceived personal risk was significantly, positively associated with readiness (r=.23, p=.007) and importance (r=.24, p=.005) to change. Cognitive risk to others was significantly, positively associated with readiness (r=.22, p=.013) and importance (r=.23, p=.008) to change. Personal risk of lower productivity was positively related to importance (r=.23, p=.030), whereas confidence to change was negatively associated with five risk domains for both personal risk and risk to others (i.e., low energy, cognitive, school/work problems, physical health, dependence/withdrawal; rs= -.17 to -.28, ps<.05). Cannabis, A Publication of the Research Society on Marijuana 54 Table 1. Means, Standard Deviations, and Frequencies of General Perceived Risk and Risk Domains Item Mean SD No risk (%) Slight risk (%) Moderate risk (%) Great risk (%) Perceived Risk to Others General perceived risk 1.73 0.81 46.2 38.5 11.5 3.8 Less productive 2.52 0.78 9.4 37.5 44.5 8.6 Lower energy 2.44 0.80 11.7 40.6 39.8 7.8 Cognitive impairment 2.25 0.78 15.6 49.2 29.7 5.5 Problems at school or work 2.35 0.84 17.1 37.2 39.5 6.2 Physical health problems 2.05 0.91 33.3 34.1 27.1 5.4 Legal problems 2.64 0.92 13.2 27.1 42.6 17.1 Dependence and withdrawal 2.30 0.9 21.3 36.2 33.9 8.7 Perceived Personal Risk General perceived risk 1.37 0.65 70.8 23.1 4.6 1.5 Less productive 1.84 0.85 38.8 44.2 10.9 6.2 Lower energy 1.77 0.83 43.4 41.9 9.3 5.4 Cognitive impairment 1.71 0.75 45.7 39.5 13.2 1.6 Problems at school or work 1.60 0.75 55.0 30.2 14.0 0.8 Physical health problems 1.54 0.71 56.6 34.1 7.8 1.6 Legal problems 1.82 0.92 45.0 35.7 11.6 7.8 Dependence and withdrawal 1.59 0.78 55.8 32.6 8.5 3.1 Table 2. Correlations between Use Frequency, Negative Consequences, CUD Symptoms, and General and Domain-Specific Perceived Risk 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 1. Use Frequency 1 - 2. Consequences -.01 1 - 3. CUD Symptoms .44** .52** 1 - 4. General RiskP .11 .31** .27* 1 - 5. General RiskO -.28* .18 .01 .50** 1 - 6. ProductivityP .33** .24* .42** .26* -.001 1 - 7. Low EnergyP .42** .21 .40** .27* -.04 .78** 1 - 8. CognitiveP .51** 0.2 .45** .30** -.01 .59** .67** 1 - 9. School/WorkP .31** .22 .39** .23* .05 .68** .73** .60** 1 - 10. Physical HealthP .27* .24* .31** .38** .18 .64** .63** .62** .65** 1 - 11. LegalP .32** -.01 .19 .13 -.13 .59** .55** .50** .47** .46** 1 - 12. DependenceP .39** .29** .42** .30** .02 .61** .58** .67** .59** .66** .58** 1 - 13. ProductivityO .01 .17 .11 .15 .15 .26* .21 .22 .12 .18 .09 .19 1 - 14. Low EnergyO -.17 .16 -.04 .17 .26* .11 .20 .07 .04 .15 -.10 .04 .61** 1 - 15. CognitiveO -.09 .28* .06 .22 .32** .14 .20 .28* .17 .28* .03 .20 .54** .45** 1 - 16. School/WorkO -.21* .28* .04 .17 .26* -.01 .11 .09 .19 .12 -.16 .08 .42** .50** .54** 1 - 17. Physical HealthO -.18 .20 .04 .25* .43** .08 .14 .11 .20 .40** .01 .21 .31** .41** .55** .50** 1 - 18. LegalO -.10 -.08 -.18 .12 .13 .11 .09 .08 .04 .05 .44** .16 .30** .27 .33** .21 .27* 1 - 19. DependenceO -.29** .16 -.13 .19 .30** -.03 .00 .01 .01 .10 -.12 .06 .34** .39** .38** .56** .48** .33** 1 Note. bolded p<.05, * p<.01, ** p<.001, O risk to others, P personal risk, CUD = cannabis use disorder Perceived Risk, Cannabis Use, and Outcomes 55 Regression Results Predictors of General Perceived Risk (Table 3) Model 1. Risk to Others Domains Predicting General Perceived Risk to Others. The general perceived risk to others model examined which domains of perceived risk to others predicted general perceived risk to others. Step one of the model accounted for significant variance in general perceived risk, F(3,122)=5.75, p <.001, R2=0.12. In step two, the model remained significant, F(10,115)=4.48, p <.001, R2=0.28, and accounted for a significant increase in R2 , ΔF(7,115)=3.58, p=.002, ΔR2=0.16. Perceived risk of physical health problems was significantly, positively associated with general perceived risk (sr2=0.065), whereas past 3-month use frequency was significantly, negatively associated with general perceived risk (sr2=0.063). Model 2. Personal Risk Domains Predicting General Perceived Personal Risk. The perceived personal risk model examined which domains of perceived personal risk predicted general perceived personal risk. Step one did not account for significant variance in general perceived personal risk, F(3,125)=1.00, p =.406, R2=0.02. At step two, the model was significant, F(10, 118) )=2.44, p =.011, R2=0.17, and accounted for a significant increase in R2 , ΔF(7,118)=3.02, p=.006, ΔR2=0.15. Perceived risk of physical problems was significantly, positively associated with general perceived risk (sr2=0.041). Table 3. Hierarchical Regression Results: General Perceived Risk Effect Estimate SE 95% CI p LL UL Model 1: General Perceived Risk to Others Step 1 - - - - - Sex assigned at birth -0.05 0.16 -0.38 0.27 .751 Age 0.03 0.05 -0.08 0.13 .58 Past 3-month use frequency -0.11 0.03 -0.16 -0.06 <.001 Step 2 - - - - - Sex assigned at birth -0.09 0.16 -0.40 0.22 .572 Age 0.03 0.05 -0.07 0.13 .527 Past 3-month use frequency -0.08 0.03 -0.13 -0.03 .002 Less productive 0.01 0.12 -0.22 0.24 .951 Lower energy 0.02 0.11 -0.21 0.25 .869 Cognitive impairment 0.11 0.12 -0.12 0.34 .353 Problems at school or work -0.08 0.11 -0.29 0.13 .455 Physical health problems 0.30 0.09 0.11 0.48 .002 Legal problems -0.02 0.08 -0.17 0.14 .828 Dependence/withdrawal 0.07 0.10 -0.12 0.25 .481 Model 2: General Perceived Personal Risk Step 1 - - - - - Sex assigned at birth -0.10 0.14 -0.37 0.18 .485 Age -0.03 0.04 -0.12 0.06 .498 Past 3-month use frequency 0.03 0.02 -0.01 0.07 .188 Step 2 - - - - - Sex assigned at birth -0.06 0.13 -0.33 0.20 .644 Age -0.07 0.04 -0.15 0.02 .131 Past 3-month use frequency -0.01 0.02 -0.06 0.04 .708 Less productive 0.04 0.12 -0.19 0.27 .709 Lower energy 0.09 0.13 -0.16 0.34 .476 Cognitive impairment 0.14 0.12 -0.09 0.38 .227 Problems at school or work -0.13 0.12 -0.37 0.11 .273 Physical health problems 0.29 0.12 0.05 0.53 .017 Legal problems -0.06 0.08 -0.22 0.10 .427 Dependence/withdrawal -0.01 0.12 -0.24 0.22 .95 Note. CI = confidence interval; LL = lower limit; UL = upper limit, CUD = cannabis use disorder, p <.05 bolded in significant models. Cannabis, A Publication of the Research Society on Marijuana 56 Perceived Risk to Others of Regular Cannabis Use (Table 4) Model 3. Risk to Others Domains Predicting Use Frequency. Step one including covariates only did not account for significant variance in past 3-month use frequency, F(2,123)=0.97, p=.381, R2=0.02. In step two including perceived risk domains, the model was significant, F(9, 116)=2.22, p=.026, R2=0.15, and accounted for significant change in R2 , ΔF(7,116)=2.55, p=.026, ΔR2=0.13. Perceived risk of dependence/withdrawal was significantly, negatively associated with use frequency (sr2=0.034). Model 4. Risk to Others Domains Predicting Negative Consequences. Step one did not account for significant variance in past 3-month negative consequences, F(3,122)=1.97, p=.122, R2=0.05. In step two, the model was significant, F(10,115)=2.48, p .010, R2=0.18, and accounted for significant change in R2 , ΔF(7,115)=2.66, p=.015, ΔR2=0.13. Perceived risk of legal problems was significantly, negatively associated with negative consequences (sr2=0.031). Model 5. Risk to Others Domains Predicting CUD Symptoms. Step one accounted for significant variance in CUD symptoms, F(3,122)=13.26, p <.001, R2=0.50. In step 2, the model remained significant, F(10,115)=5.39, p <.001, R2=0.57, although the change in the model was not, ΔF(7,118)=1.77, p=<.001, ΔR2=0.07. Perceived risk of legal problems (sr2=0.031) and age (sr2=0.030) were significantly, negatively associated with CUD symptoms. However, use frequency (sr2=0.159) was significantly, positively associated with CUD symptoms. Models 6, 7, and 8. Risk to Others Domains Predicting Motivation to Change. The readiness, importance, and confidence to change models were not statistically significant. Table 4. Hierarchical Regression Results: Perceived Risk to Others Domains Effect Estimate SE 95% CI p LL UL Model 3: Past 3-Month Use Step 1 - - - - - Sex assigned at birth -0.71 0.57 -1.84 0.42 .216 Age 0.15 0.19 -0.22 0.51 .431 Step 2 - - - - - Sex assigned at birth -0.32 0.56 -1.43 0.79 .572 Age 0.05 0.18 -0.31 0.41 .793 Less productive 0.78 0.41 -0.03 1.60 .060 Lower energy -0.65 0.41 -1.45 0.15 .112 Cognitive impairment 0.15 0.43 -0.70 0.99 .731 Problems at school or work -0.26 0.39 -1.02 0.51 .506 Physical health problems -0.17 0.33 -0.83 0.49 .604 Legal problems -0.07 0.28 -0.62 0.48 .790 Dependence/withdrawal -0.72 0.33 -1.38 -0.06 .033 Model 4: Past 3-Month Negative Consequences Step 1 - - - - - Sex assigned at birth -2.32 1.07 -4.44 -0.21 .032 Age -0.27 0.35 -0.95 0.42 .442 Past 3-month use frequency -0.05 0.17 -0.38 0.28 .772 Step 2 - - - - - Sex assigned at birth -2.24 1.04 -4.31 -0.18 .033 Age -0.27 0.34 -0.94 0.39 .418 Past 3-month use frequency 0.05 0.17 -0.29 0.39 .767 Less productive 0.73 0.78 -0.81 2.26 .352 Lower energy -0.30 0.76 -1.81 1.20 .692 Cognitive impairment 0.97 0.79 -0.59 2.53 .219 Problems at school or work 0.77 0.71 -0.64 2.19 .282 Physical health problems 0.46 0.62 -0.77 1.68 .460 Legal problems -1.08 0.51 -2.10 -0.06 .038 Dependence/withdrawal 0.42 0.63 -0.83 1.66 .511 Perceived Risk, Cannabis Use, and Outcomes 57 Effect Estimate SE 95% CI p LL UL Model 5: Past-Year CUD Symptoms Step 1 - - - - - Sex assigned at birth -0.69 0.46 -1.60 0.22 .133 Age -0.31 0.15 -0.61 -0.02 .037 Past 3-month use frequency 0.40 0.07 0.26 0.54 <.001 Step 2 - - - - - Sex assigned at birth -0.58 0.46 -1.49 0.33 .208 Age -0.33 0.15 -0.62 -0.04 .027 Past 3-month use frequency 0.39 0.08 0.24 0.54 <.001 Less productive 0.48 0.34 -0.20 1.16 .162 Lower energy -0.30 0.33 -0.96 0.36 .374 Cognitive impairment 0.10 0.35 -0.59 0.78 .780 Problems at school or work 0.31 0.31 -0.32 0.93 .332 Physical health problems 0.39 0.27 -0.15 0.92 .160 Legal problems -0.52 0.23 -0.97 -0.07 .023 Dependence/withdrawal -0.26 0.28 -0.81 0.29 .356 Model 6: Readiness to Change Step 1 - - - - - Sex assigned at birth -1.63 0.63 -2.87 -0.38 .011 Age 0.23 0.20 -0.18 0.63 .270 Past 3-month use frequency -0.21 0.10 -0.40 -0.01 .038 Step 2 - - - - - Sex assigned at birth -1.66 0.64 -2.93 -0.38 .011 Age 0.20 0.21 -0.21 0.61 .344 Past 3-month use frequency -0.15 0.11 -0.36 0.06 .153 Less productive -0.53 0.48 -1.47 0.42 .274 Lower energy 0.28 0.47 -0.65 1.21 0.550 Cognitive impairment 0.96 0.49 -0.01 1.92 .051 Problems at school or work -0.21 0.44 -1.08 0.67 .639 Physical health problems 0.03 0.38 -0.73 0.78 .942 Legal problems -0.21 0.32 -0.83 0.42 .518 Dependence/withdrawal 0.31 0.39 -0.46 1.08 .420 Model 7: Importance to Change Step 1 - - - - - Sex assigned at birth -0.61 0.67 -1.93 0.72 .365 Age -0.09 0.22 -0.52 0.34 .681 Past 3-month use frequency -0.06 0.11 -0.26 0.15 .599 Step 2 - - - - - Sex assigned at birth -0.62 0.68 -1.96 0.72 .360 Age -0.13 0.22 -0.56 0.30 .559 Past 3-month use frequency -0.01 0.11 -0.23 0.21 .927 Less productive -0.38 0.50 -1.38 0.62 .453 Lower energy 0.50 0.49 -0.48 1.47 .316 Cognitive impairment 1.17 0.51 0.16 2.18 .023 Problems at school or work -0.12 0.46 -1.04 0.80 .796 Physical health problems -0.24 0.40 -1.03 0.56 .556 Legal problems -0.42 0.33 -1.08 0.24 .214 Dependence/withdrawal 0.24 0.41 -0.57 1.05 .564 Model 8: Confidence to Change Step 1 - - - - - Sex assigned at birth 0.19 0.51 -0.83 1.20 .361 Age 0.06 0.17 -0.27 0.38 .344 Past 3-month use frequency -0.02 0.08 -0.18 0.14 .278 Step 2 - - - - - Sex assigned at birth 0.17 0.54 -0.89 1.23 .746 Age 0.06 0.17 -0.28 0.40 .734 Past 3-month use frequency -0.03 0.09 -0.21 0.14 .708 Less productive 0.26 0.40 -0.53 1.05 .512 Lower energy -0.08 0.39 -0.85 0.69 .838 Cognitive impairment -0.09 0.40 -0.89 0.72 .832 Problems at school or work -0.29 0.37 -1.02 0.44 .428 Physical health problems 0.04 0.32 -0.59 0.67 .899 Legal problems 0.06 0.26 -0.47 0.58 .832 Dependence/withdrawal 0.11 0.32 -0.53 0.75 .733 Note. CI = confidence interval; LL = lower limit; UL = upper limit, CUD = cannabis use disorder, p <.05 bolded in significant models Cannabis, A Publication of the Research Society on Marijuana 58 Perceived Personal Risk (Table 5) Model 9. Personal Risk Domains Predicting Use Frequency. Step one including only covariates did not account for significant variance in past 3- month use frequency, F(2,12)=0.73, p =.484, R2=0.01. In step two with perceived risk domains, the model was significant, F(9,119)=31.78, p <.001, R2=0.29 and accounted for a significant increase in R2 , ΔF(7,119)=6.70, p=<.001, ΔR2=0.28. Perceived personal risk of cognitive impairment was significantly, positively associated with use frequency (sr2=0.067). Model 10. Personal Risk Domains Predicting Negative Consequences. Step one did not account for significant variance in past 3-month negative consequences, F(3,125)=1.94, p =.127, R2=0.04. Step two was significant, F(10,118)=2.57, p =.007, R2=0.18, and accounted for a significant increase in R2 , ΔF(7,113)=2.76, p=.011, ΔR2=0.14. Perceived personal risk of legal problems was significantly, negatively associated with negative consequences (sr2=0.053). Model 11. Personal Risk Domains Predicting CUD Symptoms. Step one accounted for significant variance in CUD symptoms, F(3,125)=15.04, p =<.001, R2=0.26. Step two also accounted for significant variance, ΔF(7,118)=3.82, p <.001, ΔR2=0.14. Perceived personal risk of legal problems (sr2=0.024) and age (sr2=0.032) were significantly, negatively associated with CUD symptoms. Perceived personal risk to productivity (sr2=0.031) and use frequency (sr2=0.069) were significantly, positively associated with CUD symptoms. Models 12, 13, and 14. Personal Risk Domains Predicting Motivation to Change. Readiness (Model 12) and importance (Model 13) to change models were not statistically significant at step one or two. The confidence to change model (Model 14) was not significant at step one, F(3,125)=0.17, p=.916, R2=0.00 but was significant at step two, F(10,118)=2.14, p=.026, R2=0.15, and accounted for a significant increase in R2 , ΔF(7,118)=3.00, p <.001, ΔR2=0.15. Perceived personal risk of dependence/withdrawal was significantly, negatively associated with confidence to change (sr2=0.045). Lower energy was also negatively associated with confidence to change, although not statistically significant (p=0.50, sr2 =0.028). Table 5. Hierarchical Regression Results: Perceived Personal Risk Domains Effect Estimate SE 95% CI p LL UL Model 9: Past 3-Month Use Step 1 - - - - - Sex assigned at birth 0.14 0.19 -0.23 0.51 .464 Age -0.61 0.58 -1.76 0.55 .299 Step 2 - - - - - Sex assigned at birth 0.00 0.17 -0.34 0.34 .990 Age -0.28 0.53 -1.32 0.76 .598 Less productive -0.22 0.45 -1.12 0.68 .626 Lower energy 0.83 0.49 -0.15 1.81 .094 Cognitive impairment 1.48 0.45 0.60 2.37 .001 Problems at school or work -0.32 0.47 -1.25 0.60 .493 Physical health problems -0.56 0.47 -1.50 0.37 .235 Legal problems 0.16 0.31 -0.46 0.78 .607 Dependence/withdrawal 0.44 0.45 -0.46 1.34 .331 Model 10: Past 3-Month Negative Consequences Step 1 - - - - - Sex assigned at birth -0.17 0.34 -0.85 0.51 .619 Age -2.42 1.06 -4.52 -0.31 .025 Past 3-month use frequency - - - - - Step 2 Sex assigned at birth -0.19 0.34 -0.87 0.49 .578 Age -2.01 1.05 -4.08 0.06 .057 Past 3-month use frequency -0.30 0.18 -0.66 0.06 .105 Less productive 1.26 0.90 -0.53 3.05 .166 Lower energy 0.79 0.99 -1.18 2.76 .429 Cognitive impairment 0.79 0.92 -1.04 2.62 .396 Problems at school or work -0.17 0.93 -2.01 1.67 .856 Physical health problems -0.26 0.94 -2.13 1.61 .786 Legal problems -1.74 0.63 -2.98 -0.50 .006 Dependence/withdrawal 1.32 0.90 -0.47 3.11 .147 Perceived Risk, Cannabis Use, and Outcomes 59 Effect Estimate SE 95% CI p LL UL Model 11: Past-Year CUD Symptoms Step 1 - - - - - Sex assigned at birth -0.68 0.45 -1.58 0.21 .134 Age -0.32 0.15 -0.61 -0.03 .030 Past 3-month use frequency 0.42 0.07 0.28 0.55 <.001 Step 2 - - - - - Sex assigned at birth -0.36 0.43 -1.22 0.50 .415 Age -0.36 0.14 -0.64 -0.08 .013 Past 3-month use frequency 0.28 0.08 0.13 0.43 <.001 Less productive 0.93 0.38 0.18 1.67 .015 Lower energy -0.06 0.41 -0.87 0.76 .890 Cognitive impairment 0.56 0.38 -0.20 1.32 .148 Problems at school or work 0.20 0.39 -0.57 0.96 .609 Physical health problems -0.27 0.39 -1.05 0.51 .494 Legal problems -0.56 0.26 -1.08 -0.05 .033 Dependence/withdrawal 0.38 0.38 -0.36 1.12 .314 Model 12: Readiness to Changea Step 1 - - - - - Sex assigned at birth 0.22 0.20 -0.17 0.62 .271 Age -1.62 0.62 -2.85 -0.39 .011 Past 3-month use frequency -0.18 0.10 -0.37 0.00 .054 Step 2 - - - - - Sex assigned at birth 0.20 0.22 -0.23 0.62 .357 Age -1.57 0.65 -2.86 -0.28 .018 Past 3-month use frequency -0.19 0.11 -0.41 0.04 .104 Less productive 0.02 0.57 -1.10 1.14 .973 Lower energy -0.10 0.62 -1.32 1.13 .878 Cognitive impairment -0.02 0.58 -1.17 1.12 .966 Problems at school or work 0.11 0.58 -1.04 1.26 .848 Physical health problems 0.74 0.59 -0.43 1.91 .210 Legal problems -0.31 0.39 -1.09 0.46 .423 Dependence/withdrawal -0.10 0.56 -1.22 1.02 .861 Model 13: Importance to Changea Step 1 - - - - - Sex assigned at birth -0.57 0.66 -1.88 0.74 .392 Age -0.08 0.21 -0.50 0.34 .704 Past 3-month use frequency -0.02 0.10 -0.22 0.18 .860 Step 2 - - - - - Sex assigned at birth -0.34 0.67 -1.67 0.99 .618 Age -0.12 0.22 -0.56 0.32 .591 Past 3-month use frequency -0.10 0.12 -0.33 0.14 .417 Less productive 1.12 0.58 -0.03 2.27 .057 Lower energy -0.11 0.64 -1.38 1.15 .859 Cognitive impairment 0.57 0.59 -0.61 1.74 .342 Problems at school or work 0.17 0.60 -1.01 1.36 .771 Physical health problems 0.05 0.61 -1.16 1.25 .940 Legal problems -0.83 0.40 -1.63 -0.04 .041 Dependence/withdrawal -0.21 0.58 -1.36 0.94 .721 Model 14: Confidence to Change Step 1 - - - - - Sex assigned at birth 0.11 0.51 -0.89 1.12 .826 Age 0.09 0.16 -0.24 0.41 .593 Past 3-month use frequency -0.03 0.08 -0.18 0.12 .703 Step 2 - - - - - Sex assigned at birth 0.04 0.50 -0.95 1.02 .939 Age 0.06 0.16 -0.26 0.38 .717 Past 3-month use frequency 0.12 0.09 -0.06 0.29 .181 Less productive 0.68 0.43 -0.17 1.53 .115 Lower energy -0.94 0.47 -1.87 0.00 .050 Cognitive impairment -0.21 0.44 -1.08 0.66 .629 Problems at school or work -0.10 0.44 -0.98 0.77 .820 Physical health problems 0.04 0.45 -0.84 0.93 .921 Legal problems 0.55 0.30 -0.04 1.13 .069 Dependence/withdrawal -1.07 0.43 -1.92 -0.22 .014 Note. CI = confidence interval; LL = lower limit; UL = upper limit, CUD = cannabis use disorder, p <.05 bolded in significant models, aoverall model not significant Cannabis, A Publication of the Research Society on Marijuana 60 Differences by Cannabis Use Frequency (Table 6) Compared to individuals who used cannabis infrequently (n=64), individuals who used more frequently (n=65) reported significantly lower general perceived risk to others, but did not significantly differ on any risk to others domains. Individuals who used infrequently did not differ from those who used more frequently on general personal risk but reported significantly higher ratings of personal risk domains (medium to large effects), except for personal physical and legal risk. Table 6. Means, Standard Deviations, and One-Way Analyses of Variance by Use Frequency Measure Infrequent Use (n=64) Frequent Use (n=65) F(4,123) p η2 M SD M SD Perceived Risk to Others Less productive 2.56 0.80 2.49 0.77 0.21 .652 0.002 Lower energy 2.59 0.73 2.29 0.84 4.19 .043 0.033 Cognitive impairment 2.36 0.8 2.14 0.75 2.44 .121 0.019 Problems at school or work 2.53 0.84 2.17 0.80 6.28 .014 0.048 Physical health problems 2.22 0.98 1.88 0.80 4.50 .036 0.035 Legal problems 2.77 0.85 2.51 0.97 2.47 .119 0.019 Dependence/withdrawal 3.06 1.02 2.78 1.01 2.55 .113 0.020 General risk to others 1.98 0.86 1.48 0.69 13.10 <.001 0.094 Perceived Personal Risk Less productive 1.56 0.77 2.12 0.84 16.21 <.001 0.115 Lower energy 1.47 0.67 2.06 0.88 19.37 <.001 0.134 Cognitive impairment 1.44 0.56 1.97 0.83 18.31 <.001 0.128 Problems at school or work 1.39 0.63 1.82 0.81 10.58 .001 0.078 Physical health problems 1.41 0.61 1.68 0.77 5.13 .025 0.039 Legal problems 1.59 0.75 2.05 1.02 8.72 .004 0.065 Dependence/withdrawal 1.34 0.54 1.83 0.89 13.46 <.001 0.097 General personal risk 1.28 0.58 1.45 0.71 2.08 .152 0.016 Note. Sex assigned at birth and age included as covariates (not shown), significant at Bonferroni corrected p <.003 (bolded) DISCUSSION The present study tested how general and domain-specific perceived risks to others and oneself were cross-sectionally associated with cannabis use frequency, negative outcomes, and motivation to change among a sample of undergraduates who use cannabis. Results partially supported hypotheses. For Aim 1, general perceived risk to others was negatively correlated with use frequency, whereas general perceived personal risk was positively associated with consequences/CUD symptoms and some facets of motivation to change (i.e., importance, readiness). For Aim 2, after accounting for shared variance among risk domains, legal and dependence/withdrawal risk were uniquely associated with outcomes. For Aim 3, undergraduates who used cannabis more frequently reported greater perceived personal risk in five of seven risk domains and less perceived general risk to others compared to those Perceived Risk, Cannabis Use, and Outcomes 61 who used less frequently, despite no significant differences in general personal risk or domain- specific risk to others. Findings suggest assessing both general and domain-specific risk, as well as risk to others and oneself, can provide a more nuanced understanding of perceived risk and its association with cannabis outcomes among undergraduates. Perceived personal risk was rated in the no-to- slight risk range, on average, whereas perceived risk to others was rated in the slight-to-moderate risk range, even though average use in the sample was higher than “regular use” as defined in risk to others items. Thus, undergraduates may minimize their personal risk of cannabis use, despite acknowledging a similar level of use poses risk to their peers. Although general perceived risk to others was negatively associated with use frequency, general perceived personal risk was not. Rather, general perceived personal risk was positively related to negative consequences and CUD symptoms. In contrast with prior work finding no association between perceived risk and the experience of negative consequences (i.e., academic, social) among undergraduates who use cannabis (Kilmer et al., 2007), the present findings indicate individuals who experience problems related to their use may perceive greater personal risk, despite viewing their personal use as less risky than a similar level of use for others. Findings also underscore the importance of considering domain-specific perceived risk. Only the physical risk domain predicted general perceived risk to self and others. Undergraduates may be focusing on the “physical risk” portion of the item when rating general perceived risk. Importantly, although only the physical risk domain significantly predicted general risk, students rated physical risk as having the lowest risk of any domain. Physical health problems are rarely reported by young adults who use cannabis (Buckner et al., 2010; Terry-McElrath et al., 2022), and negative physical health effects of cannabis tend to be cumulative (Volkow et al., 2014). The infrequency and lack of immediacy of these consequences may contribute to an inaccurate perception of actual physical risks associated with heavy, frequent cannabis use. Second, given the increase in legalization of medical cannabis use in the past decade and attention toward prescribing cannabis to manage physical concerns (National Academies of Sciences & Medicine, 2017), students may view cannabis as being less risky physically. This is supported in part by decreases in perceived risk post-legalization (Mennis et al., 2023). Several domains of risk emerged as important predictors of outcomes. Consistent with hypotheses, dependence/withdrawal risk was a significant predictor of cannabis outcomes; however, cognitive, productivity, and legal risks also served as unique predictors. Perceived dependence/withdrawal risk to others and perceived personal cognitive risk were negatively associated with use frequency, and perceived personal risk to productivity was positively associated with CUD symptoms. Similar to other work finding greater perceived risk is protective against cannabis use (e.g., D’Silva et al., 2020), present findings indicate some domain-specific perceived risks might mitigate risk associated with certain cannabis use behaviors. Additionally, perceived legal risk to self and others were negatively associated with consequences and CUD symptoms. Knowing perceived legal risk may mitigate harms associated with cannabis use, college campuses may consider maintaining strict cannabis policies with required intervention post- violation as evidence suggests undergraduates decrease their cannabis use post-sanction (Buckner et al., 2018). In addition to decreasing use, strict campus policies and an associated intervention may result in higher-risk students experiencing fewer negative outcomes and maintaining more accurate perception of legal/systemic risk despite changes in state/federal policies. General perceived personal risk (but not general or domain-specific perceived risk to others) was positively associated with importance and readiness to change. Thus, perceiving risk to others may be too distal an association to motivate individuals to change their own cannabis use. Notably, even though the overall sample reported high confidence to change, greater personal risk of dependence/withdrawal was negatively associated with confidence to change. Individuals who perceive themselves at risk of withdrawal/dependence may experience uncertainty about how to change use and/or manage withdrawal/dependence symptoms. As high confidence to change is an especially important predictor of changes in substance use among young adults (Bertholet et al., 2012), it Cannabis, A Publication of the Research Society on Marijuana 62 may be advantageous to emphasize personal risk, rather than risk to others, when working with individuals who use cannabis. Further, as confidence to change increases early on in cannabis-related treatment (Chung & Maisto, 2016), continuing to provide psychoeducation, problem-solving, and skills around managing withdrawal/dependence to increase confidence remains important. Consistent with hypotheses, participants who used frequently reported less general perceived risk to others than those who used infrequently. However, regarding domain-specific perceived personal risk, those who use more frequently perceived five out seven domains as riskier than those who use infrequently, consistent with prior research finding undergraduates who engage in more frequent cannabis use rate their personal risk higher (O'Callaghan et al., 2006). Notably, results could reflect differences in how personal risk vs risk to others items were framed. Specifically, participants were asked to rate risk to others who used once a week or more, but rate personal risk based on one’s current frequency of use. If a participant’s personal use rate was much higher than once per week, they may view their use as inherently riskier given consequences they are currently or have previously experienced. It will be important to test whether results remain consistent if risk to others is assessed more similarly to personal risk (i.e., “if others use marijuana at your current rate of use”). Clinicians may want to focus on increasing perceived personal risk among undergraduates who use cannabis rather than general risk, as perceived personal risk (but not perceived risk to others) was associated with negative consequences and CUD symptoms. Further, the current study suggests it may be particularly useful to emphasize psychoeducation on long- term risks and consequences of cannabis including physical, legal, and dependence risks for undergraduates who use cannabis more broadly. Individuals may be reporting short-term perceived personal risks more accurately as they may have already experienced them; however, there may be a misperception of risks associated with long-term consequences due to their lack of immediacy. Additionally, it may be beneficial to provide psychoeducation on domains individuals are most concerned about (e.g., dependence/withdrawal among individuals who use heavily, legal risk among individuals in states with illegal recreational cannabis). Notably, as perceived risk appears to be a protective factor for undergraduates who do not use or use infrequently (e.g., Hanauer et al., 2021; Kilmer et al., 2007), targeting perceived risk to others may be particularly useful in prevention programs when paired with education on safer use to protect against potential negative consequences. Results should be viewed in light of study limitations. First, data were collected cross- sectionally. Further examination of research questions longitudinally could provide critical information regarding temporal relations among study variables including how individuals’ risk perceptions change over time and how various risk perceptions may protect against or contribute to the onset or maintenance of problematic cannabis use. Second, questions were administered as part of an intervention which may have impacted results. Relatedly, participants rated general and domain-specific risk to others ‘if they use marijuana regularly [once a week or more]” and personal risk “if you use marijuana at your current rate of use.” Future research would benefit from using a consistent use frequency anchor to assess risk to self and others. Third, data was collected in a state where recreational cannabis was illegal which may contribute to findings associated with perceived legal risk. Given previous research has shown a decrease in felony convictions, arrests, and police involvement related to cannabis following legalization in certain states (Maxwell & Mendelson, 2016), it is critical to assess perceived risk of cannabis use across states with varying legal status. Further, the sample was relatively small and predominantly comprised of non- Hispanic/Latin, White, female participants; replication with larger and more diverse samples in terms of race/ethnicity, age, and sex assigned at birth will be important. Notably, legal risk may differ unfairly among racial and ethnic minority groups due to bias-driven racial disparities in cannabis-related arrests and convictions (Bunting et al., 2013). Clinicians working with patients who use cannabis should consider the legal status and related legal risks in their state of practice, and potential impacts of legal disparities for their clients; exploring perceived legal risk and providing psychoeducation if indicated may be Perceived Risk, Cannabis Use, and Outcomes 63 useful for patients expressing ambivalence about changing their use. Fourth, although the present study provides important information on various domains of risk, domains were not comprehensive. Measures of domain-specific perceived risk to self and others were developed specifically as part of the PFI for this study. Although results suggest preliminary construct validity, future work testing other psychometric properties of these measures is necessary. Further, although risk domains were empirically informed, there may be other risk domains (e.g., driving while intoxicated, financial challenges) not captured in the present study. Future work could use qualitative interviews to further identify the most relevant domains to undergraduates and test whether results replicate in a larger, more diverse sample. Taken together, future research should examine relations posed in this study longitudinally, utilize questions with consistent use rates for risk to others and oneself, in states with varying legal status of cannabis, in larger, more diverse samples, outside the context of an intervention study, and include a wider variety of risk domains. Perceived risk of using cannabis is on a consistent decline despite known risks (Lipari & Jean-Francois, 2016; Waddell, 2022). Results of the present study suggest future research could benefit from expanding current conceptualizations of perceived risk to include perceived personal risk and domain-specific risks in addition to general risk to others. College-based psychoeducation and prevention efforts should consider various facets of risk including identifying the most salient areas of perceived risk for different undergraduates based on their personal circumstances. REFERENCES Azofeifa, A., Mattson, M. E., Schauer, G., McAfee, T., Grant, A., & Lyerla, R. (2016). National estimates of marijuana use and related indicators—National Survey on Drug Use and Health, United States, 2002–2014. Morbidity and Mortality Weekly Report: Surveillance Summaries, 65(11), 1-25. https://doi.org/10.15585/mmwr.ss6511a1 Bachman, J. G., Johnson, L. D., & O'Malley, P. M. (1998). Explaining recent increases in students' marijuana use: Impacts of perceived risks and disapproval, 1976 through 1996. American Journal of Public Health, 88(6), 887- 892. https://doi.org/10.2105/AJPH.88.6.887 Bachman, J. G., Johnston, L. D., O'Malley, P. M., & Humphrey, R. H. (1988). Explaining the recent decline in marijuana use: Differentiating the effects of perceived risks, disapproval, and general lifestyle factors. Journal of Health and Social Behavior, 29(1), 92-112. https://doi.org/10.2307/2137183 Bertholet, N., Gaume, J., Faouzi, M., Gmel, G., & Daeppen, J.-B. (2012). Predictive value of readiness, importance, and confidence in ability to change drinking and smoking. BMC Public Health, 12, 708. https://doi.org/10.1186/1471-2458-12-708 Buckner, J. D., Bonn-Miller, M. O., Zvolensky, M. J., & Schmidt, N. B. (2007). Marijuana use motives and social anxiety among marijuana- using young adults. Addictive Behaviors, 32(10), 2238-2252. https://doi.org/10.1016/j.addbeh.2007.04.004 Buckner, J. D., Ecker, A. H., & Cohen, A. S. (2010). Mental health problems and interest in marijuana treatment among marijuana-using undergraduates. Addictive Behaviors, 35(9), 826-833. https://doi.org/10.1016/j.addbeh.2010.04.001 Buckner, J. D., Jeffries, E. R., Terlecki, M. A., & Ecker, A. H. (2016). Distress tolerance among students referred for treatment following violation of campus cannabis use policy: Relations to use, problems, and motivation. Behavior Modification, 40(5), 663-677. https://doi.org/10.1177/0145445515610315 Buckner, J. D., & Schmidt, N. B. (2008). Marijuana effect expectancies: Relations to social anxiety and marijuana use problems. Addictive Behaviors, 33(11), 1477-1483. https://doi.org/10.1016/j.addbeh.2008.06.017 Buckner, J. D., Walukevich, K. A., Lemke, A. W., & Jeffries, E. R. (2018). The impact of university sanctions on cannabis use: Individual difference factors that predict change in cannabis use. Translational Issues in Psychological Science, 4(1), 76-84. https://doi.org/10.1037/tps0000147 Bunting, W. C., Garcia, L., & Edwards, E. (2013). The war on marijuana in black and white. American Civil Liberties Union. https://www.aclu.org/sites/default/files/field_d ocument/1114413-mj-report-rfs-rel1.pdf Cannabis, A Publication of the Research Society on Marijuana 64 Burdzovic Andreas, J., Sivertsen, B., Lønning, K. J., & Skogen, J. C. (2021). Cannabis use among Norwegian university students: Gender differences, legalization support and use intentions, risk perceptions, and use disorder. Addictive Behaviors Reports, 13, 100339. https://doi.org/10.1016/j.abrep.2021.100339 Caldeira, K. M., Arria, A. M., 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 undergraduates. Addictive Behaviors, 33(3), 397-411. https://doi.org/10.1016/j.addbeh.2007.10.001 Chung, T., & Maisto, S. A. (2016). Time-varying associations between confidence and motivation to abstain from marijuana during treatment among adolescents. Addictive Behaviors, 57, 62-68. https://doi.org/10.1016/j.addbeh.2016.02.015 Compton, W. M., Han, B., Jones, C. M., Blanco, C., & Hughes, A. (2016). Marijuana use and use disorders in adults in the USA, 2002–14: Analysis of annual cross-sectional surveys. The Lancet Psychiatry, 3(10), 954-964. https://doi.org/10.1016/S2215-0366(16)30208- 5 D’Silva, J., Boekeloo, B., Seaman, E., & Quinton, S. (2020). Marijuana risk perceptions are associated with past-month marijuana use among black first-year undergraduates at a historically black university. Journal of Ethnicity in Substance Abuse, 19(4), 623-634. https://doi.org/10.1080/15332640.2018.155676 5 Danseco, E. R., Kingery, P. M., & Coggeshall, M. B. (1999). Perceived risk of harm from marijuana use among youth in the USA. School Psychology International, 20, 39-56. https://doi.org/10.1177/0143034399201004 Gates, P. J., Norberg, M. M., Copeland, J., & Digiusto, E. (2012). Randomized controlled trial of a novel cannabis use intervention delivered by telephone. Addiction, 107(12), 2149-2158. https://doi.org/10.1111/j.1360- 0443.2012.03953.x Grevenstein, D., Nagy, E., & Kroeninger- Jungaberle, H. (2015). Development of risk perception and substance use of tobacco, alcohol and cannabis among adolescents and emerging adults: Evidence of directional influences. Substance Use & Misuse, 50(3), 376-386. https://doi.org/10.3109/10826084.2014.984847 Gunn, R. L., Aston, E. R., Sokolovsky, A. W., White, H. R., & Jackson, K. M. (2020). Complex cannabis use patterns: Associations with cannabis consequences and cannabis use disorder symptomatology. Addictive Behaviors, 105, 106329. https://doi.org/10.1016/j.addbeh.2020.106329 Hanauer, M., Walker, M. R., Machledt, K., Ragatz, M., & Macy, J. T. (2021). Association between perceived risk of harm and self- reported binge drinking, cigarette smoking, and marijuana smoking in young adults. Journal of American College Health, 69(4), 345-352. https://doi.org/10.1080/07448481.2019.167675 7 Janz, N. K., & Becker, M. H. (1984). The health belief model: A decade later. Health Education Quarterly, 11(1), 1-47. https://doi.org/10.1177/109019818401100101 Kasten, S., van Osch, L., Candel, M., & de Vries, H. (2019). The influence of pre-motivational factors on behavior via motivational factors: A test of the i-change model. BMC Psychology, 7(1), 7. https://doi.org/10.1186/s40359-019- 0283-2 Keith, D. R., Hart, C. L., McNeil, M. P., Silver, R., & Goodwin, R. D. (2015). Frequent marijuana use, binge drinking and mental health problems among undergraduates. American Journal on Addictions, 24(6), 499-506. https://doi.org/10.1111/ajad.12201 Kilmer, J. R., Hunt, S. B., Lee, C. M., & Neighbors, C. (2007). Marijuana use, risk perception, and consequences: Is perceived risk congruent with reality? Addictive Behaviors, 32(12), 3026-3033. https://doi.org/10.1016/j.addbeh.2007.07.009 Lipari, R., & Jean-Francois, B. (2016). Trends in perception of risk and availability of substance use among full-time undergraduates. Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration. Magnan, R. E., & Ladd, B. O. (2019). “It’s all good”: Perceived benefits but not perceived risks or worries among adult marijuana users. Cannabis, 2(2), 112-119. https://doi.org/10.26828/cannabis.2019.02.001 Perceived Risk, Cannabis Use, and Outcomes 65 Maxwell, J. C., & Mendelson, B. (2016). What do we know about the impact of the laws related to marijuana? Journal of Addiction Medicine, 10(1), 3-12. https://doi.org/10.1097/ADM.00000000000001 88 Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in survey data. Psychological Methods, 17(3), 437-455. https://doi.org/10.1037/a0028085 Mennis, J., McKeon, T. P., & Stahler, G. J. (2023). Recreational cannabis legalization alters associations among cannabis use, perception of risk, and cannabis use disorder treatment for adolescents and young adults. Addictive Behaviors, 138, 107552. https://doi.org/10.1016/j.addbeh.2022.107552 Miller, W. R., & Rollnick, S. (2012). Motivational interviewing: Helping people change. Guilford press. Morris, P. E., & Buckner, J. D. (2023). Cannabis- related problems and social anxiety: The roles of sex and cannabis use motives updated. Addictive Behaviors, 137, 107528. https://doi.org/10.1016/j.addbeh.2022.107528 National Academies of Sciences, Engineering, & Medicine. (2017). The health effects of cannabis and cannabinoids: The current state of evidence and recommendations for research. The National Academies Press. https://doi.org/10.17226/24625 O'Callaghan, F., Reid, A., & Copeland, J. (2006). Risk perception and cannabis use in a sample of young adults. Journal of Substance Use, 11(2), 129-136. https://doi.org/10.1080/14659890500237366 Okaneku, J., Vearrier, D., McKeever, R. G., LaSala, G. S., & Greenberg, M. I. (2015). Change in perceived risk associated with marijuana use in the united states from 2002 to 2012. Clinical Toxicology, 53(3), 151-155. https://doi.org/10.3109/15563650.2015.100458 1 Patrick, M. E., Fairlie, A. M., Cadigan, J. M., Abdallah, D. A., Larimer, M. E., & Lee, C. M. (2019). Daily motives for alcohol and marijuana use as predictors of simultaneous use among young adults. Journal of Studies on Alcohol and Drugs, 80(4), 454-461. https://doi.org/10.15288/jsad.2019.80.454 Sales, B. D., & Folkman, S. E. (2000). Ethics in research with human participants. American Psychological Association. Schulenberg, J. E., Patrick, M. E., Johnston, L. D., O'Malley, P. M., Bachman, J. G., & Miech, R. A. (2021). Monitoring the Future national survey results on drug use, 1975-2020. Volume II, College students & adults ages 19-60. Institute for Social Research. Stephens, R. S., Roffman, R. A., & Curtin, L. (2000). Comparison of extended versus brief treatments for marijuana use. Journal of Consulting and Clinical Psychology, 68(5), 898-908. https://doi.org/10.1037/0022- 006X.68.5.898 Suerken, C. K., Reboussin, B. A., Egan, K. L., Sutfin, E. L., Wagoner, K. G., Spangler, J., & Wolfson, M. (2016). Marijuana use trajectories and academic outcomes among undergraduates. Drug and Alcohol Dependence, 162, 137-145. https://doi.org/10.1016/j.drugalcdep.2016.02.0 41 Terry-McElrath, Y. M., Patrick, M. E., O'Malley, P. M., & Johnston, L. D. (2022). Self-reported perceived negative consequences of marijuana use among US young adult users, 2008–2019. Addictive Behaviors, 124, 107098. https://doi.org/10.1016/j.addbeh.2021.107098 Volkow, N. D., Baler, R. D., Compton, W. M., & Weiss, S. R. (2014). Adverse health effects of marijuana use. New England Journal of Medicine, 370(23), 2219-2227. https://doi.org/10.1056/NEJMra1402309 Waddell, J. T. (2022). Age-varying time trends in cannabis- and alcohol-related risk perceptions 2002–2019. Addictive Behaviors, 124, 107091. https://doi.org/10.1016/j.addbeh.2021.107091 Walukevich-Dienst, K. Targeting perceived risk through an online personalized feedback intervention for cannabis-using college students (2019). Master's theses. Louisiana State University. https://doi.org/10.31390/gradschool_theses.48 52 Walukevich-Dienst, K., Lewis, E. M., Neighbors, C., Green, J. C., & Buckner, J. D. (2021). Online personalized feedback intervention reduces cannabis-related problems among college students with high problem distress. Experimental and Clinical Cannabis, A Publication of the Research Society on Marijuana 66 Psychopharmacology, 29(1), 14-22. https://doi.org/10.1037/pha0000361 Walukevich-Dienst, K., Neighbors, C., & Buckner, J. D. (2019). Online personalized feedback intervention for cannabis-using college students reduces cannabis-related problems among women. Addictive Behaviors, 98, 106040. https://doi.org/10.1016/j.addbeh.2019.106040 Funding and Acknowledgements: This research was supported by grants from the National Institute on Alcohol Abuse and Alcoholism (NIAAA) and National Institute on Drug Abuse (NIDA). Data collection was supported by funding from Louisiana State University's Department of Psychology awarded to Dr. Walukevich-Dienst. Data analysis and manuscript preparation were supported by F31DA057796 (PI: Smith- LeCavalier), T32AA007455 (PI: Larimer), F32AA029589 (PI: Walukevich-Dienst), K23AA031034 (PI: Walukevich-Dienst), R21AA030071 (PI: Buckner), and R21DA056846 (Buckner). The content is solely the responsibility of the authors and does not necessarily represent the official views of NIAAA or NIDA. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Copyright: © 2024 Authors et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction, provided the original author and source are credited, the original sources is not modified, and the source is not used for commercial purposes. https://creativecommons.org/licenses/by/4.0/