Moreover, minimal research has tested the measurement invariance of the CUDIT-R across countries, hindering cross-national comparisons. This is especially relevant among nations with different cannabis consumption policies , as differing use patterns may lead to differences in cannabis use problems experienced. Participants were college students recruited to participate in an online survey from 7 countries between February 2019 and March 2020 . In the present study, participants that reported consuming cannabis at least once in their lifetime and completed the CUDIT-R were included in the psychometric analyses. Analyses examining correlations between the CUDIT-R and non-CUDIT-R measures were limited to 2402 students that reported cannabis consumption during the last 30 days . For the U.S., Canadian, England, and South African sites, students were recruited from psychology department pools and received research participation credits. In Argentina and Uruguay, students were recruited through online social networks, e-mail listings and flyers , and those who completed the survey were entered into a raffle for prizes. In Spain an email was sent to all the students of the university inviting them to participate in the research. The participants received 5 euros for completing the survey. Study procedures were approved by the institutional review boards at the participating universities. Only students who endorsed past 30-day use of marijuana were administered non CUDIT-R measures. Invariance testing of all non-CUDIT-R measures demonstrated metric invariance across the countries, which is necessary when examining associations between study constructs across different groups . Cannabis misuse was assessed using the CUDIT-R . In Argentina, Spain, and Uruguay, we used a version translated into Spanish. The translation was performed by two Spanish-speaking psychologists proficient in English and Spanish and with expertise in test adaptation and addictive behaviors.
Once the first drafts of the scale were performed in Castilian Spanish and Argentinian Spanish; the two versions were compared and only minor differences were kept across both Spanish versions to facilitate the understanding of the scale in each cultural context . For both the Spanish and English versions, items 1 to 7 use a 5-point Likert scale that ranks from 0 to 4, while item 8 uses a 3-point Likert scale . In the present study, a total score was obtained by summing scores on each of the 8 items . Past 30-day typical cannabis use frequency and quantity were assessed using the Marijuana Use Grid . The measure was translated into Spanish for students in Argentina, Spain, and Uruguay. Specifically, each day of the week was broken down into six 4-hour blocks of time , and participants were asked to report at which times they used cannabis grow equipment during a “typical week” in the past 30 days as well as the quantity of grams consumed during that time block. We calculated typical frequency of cannabis use by summing the total number of time blocks for which they reported using during the typical week . We calculated typical quantity of cannabis use by summing the total number of grams consumed across time blocks . To examine the internal structure of the CUDIT-R across sites, we conducted confirmatory factor analyses using a diagonally weighted least squares estimator in Mplus 8.4 among lifetime cannabis users. To examine the internal consistency of the CUDIT-R, we estimated Cronbach’s alphas and ordinal omegas with 95% CIs . To determine the configural, metric and scalar invariance of the CUDIT-R across countries and gender groups, we conducted multi-group confirmatory factor analyses.Evidence of convergent and criterion validity was examined by correlating the total CUDIT-R score with the B-MACQ and cannabis outcomes scores , respectively. To explore the differences in the magnitude of the correlations across countries, the absolute value of the differences in the magnitude of the correlations for pairs of countries were computed in the sample of last 30-day cannabis users. As the statistical tests of these differences can be oversensitive to small differences when including differences in sample sizes across countries, attention was paid to the magnitude of these differences. As done in previous studies , we considered differences < 1 SD small, differences between 1 SD and 2 SD medium, those between 2 SD and 3 SD large, and any over 3 SD were considered substantial. The correlations between the CUDIT-R, the B-MACQ and the criterion variables are presented in Table 3. The correlations between the CUDIT-R with the B-MACQ were large; while correlations with quantity and frequency of cannabis use were small to moderate. Correlations with cannabis motives were from moderate to large across groups .
When partial correlations between the CUDIT-R and cannabis motives were performed the results showed that the CUDIT-R was mainly related to the internal cannabis motives . When the absolute value of the correlations was compared between pairs of groups, the average difference in correlations was 0.077 across 143 possible comparisons. Thus, differences < 0.141 were considered small, differences between 0.141 and 0.205 were considered medium, those between 0.206 and 0.269 were considered large, and any over 0.270 were considered substantial. The correlation differences between the CUDIT-R and the rest of the criterion variables are presented in Table 4. We found large differences in the correlation of the CUDIT-R with cannabis quantity between the Argentinian sample and the sample from South Africa and Spain . The correlations between conformity motives and the CUDIT-R were higher in Spain than in Canada, South Africa and Argentina. Specifically, large and substantial differences were found between the CUDIT-R and conformmotives in South Africa and Argentina compared with Spain respectively. When the effect of the rest of motives was controlled for, large correlation differences in conformity motives between Spain and Canada were found, and substantial correlation differences in Spain compared with South Africa and Argentina were shown. The CUDIT-R is one of the most widely used measures to assess and detect problematic cannabis use . However, limited research has tested the measurement invariance of the questionnaire across different countries. Thus, a main aim of the present research was to test the measurement invariance of the CUDIT-R among college students from seven countries. The results showed that the measure was invariant at different levels among the U.S., Canada, Argentina, Spain, and South African samples suggesting that the CUDIT-R is a suitable measure to compare cannabis-related problems among students from these five countries. Due to their low item endorsement, it was not possible to the test the invariance of the questionnaire in Uruguay and England. Future studies with a higher sample size of students from these two countries are required. In addition, measurement invariance of the questionnaire across males and females was explored. A previous study performed with a large sample of undergraduates from the U.S. found configural and metric invariance of the measure across gender groups . Our results extend previous findings, showing that the structure , the factor loadings and the thresholds were similar across a sample of males and females from different nationalities and languages.
Scalar invariance of the questionnaire across groups is relevant, as differences in the CUDIT-R scores across countries and gender groups can be interpreted as differences in problematic cannabis use, rather than merely artifacts of other processes, such as the interpretation of items . To this end, when mean comparisons of the CUDIT-R total score were examined, as it is usually found, males showed higher problematic cannabis use than females . We also found lower problematic cannabis use in the Spain and Argentina than in U.S., and also in Spain compared with the rest of the countries. These differences could be related to cannabis polices, as lower rates of CUDITR were found in countries in which the access to cannabis is more difficult . The present research also provided reliability evidence of the CUDITR scores, showing that both Cronbach’s alphas and ordinal omegas were higher than the standard cut-off of 0.70, and similar to those found in previous studies that have explored the unidimensional structure of the questionnaire . Convergent validity evidence was also provided in five countries, as the CUDIT-R showed large correlations in magnitude with the B-MACQ. The magnitude of these associations was similar to that found in previous studies with undergraduates from the U.S. , suggesting that the Spanish version provided similar convergent validity evidence with the B-MACQ as the English version. Finally, criterion validity evidence of the CUDIT-R scores was provided using different measures of cannabis consumption and cannabis related motives. As expected, CUDIT-R scores positively correlated to cannabis frequency and quantity across the five countries and gender groups . Although the correlations with quantity were positive and significant across all countries, the magnitude of this correlation was lower in Argentina than in the remaining countries. This result is somewhat different to past research comprising Argentinean college students in which higher correlations between cannabis quantity and cannabis-related problems have been found . Thus, more studies are needed to better clarify the magnitude and differences of these associations. Among motives, all five cannabis motives were associated with the CUDITR scores across groups, as found in recent meta-analysis . In the same vein, when the rest of the motives were controlled for, the CUDIT-R was mainly related to coping motives across countries and gender groups. Lower but significant associations between the CUDIT-R scores and enhancement and expansion motives were also found. Our findings and previous research performed with college students suggest that internal cannabis motives are indicators of risk. The only exception was found in the sample of Spanish youths, in which the correlation between the CUDIT-R scores and conformity motives remained significant even when accounting for other motives.
Consequently, substantial differences in the magnitude of the correlations between the CUDIT-R scores and conformity motives in Spain compared with South Africa and Argentina arose. The higher correlation between the conformity motives and cannabis-related problems in Spain could be due to differences in cannabis policies across sites. For example, in Spain buying, selling,mobile grow system and the use of cannabis in public places is illegal, however, there are “cannabis social clubs ” in which the “private” sale and consumptions is allowed from 18 or 21 years old depending on the club or the Spanish autonomous community . Thus, in this particular context in which the cannabis use is allowed and shared with peers, it is possible that youths have lower self-efficacy to reject consumption which in turn could be related to higher problems with cannabis. Nonetheless, further research is needed to better clarify the association between conformity motives and the CUDIT-R in Spanish youths. The present study has a number of limitations. The relatively low number of participants from Uruguay and England impeded testing the invariance of the CUDIT-R’s scores in these samples. Moreover, our convenience sampling procedures impedes generalizing our findings. Sample size and the gender distribution were different across countries which, alongside the online data collection, may have introduced unwanted bias.Despite these limitations, the present research offers encouraging evidence of the psychometric properties of the CUDIT-R scores in college student cannabis users from five countries and across gender groups. Overall, these results suggest that the CUDIT-R could be efficiently used to identify English or Spanish-speaking college students with problematic cannabis use, who would benefit from interventions aimed at reducing cannabis use and its negative consequences. Dr. Bravo was supported by a training grant from the National Institute on Alcohol Abuse and Alcoholism in the United States during the duration of data collection for this project. Data collection was supported, in part, by grant T32-AA018108. NIAAA had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication. Data collection in Spain was also supported by grants UJIA2019-08 from the Universitat Jaume I and RTI2018-099800-B-I00 from the Spanish Ministry of Science, Innovation and Universities . Data collection in Argentina was also supported by grants from the National Secretary of Science and Technology and by grants from the Secretary of Science and Technology- National University of C´ ordoba . Cannabis is the most widely used substance in the United States after alcohol and tobacco, including during pregnancy .