Reflective and non-reflective influences on cannabis use among undergraduate students: A qualitative study

Abstract

Objective

Around 40% of US university students use cannabis, 25% of whom present with cannabis use disorder, which endangers health. We investigated the concurrent contribution of reflective processes, which generate action via conscious deliberation, and non-reflective processes, which prompt behavior automatically, to undergraduates’ cannabis consumption.

Participants

Eighteen UK undergraduates who regularly consume cannabis (11 female, 7 male; mean age 20 y).

Methods

Semi-structured interviews explored cannabis motives, routines, cues, and decision points. Thematic analysis identified themes, in each of which reflective and non-reflective dimensions were coded.

Results

Four themes were identified: cannabis use for relaxation, social bonding, and symbolic-affective significance, and contexts and triggers. Some influences guided cannabis use reflectively in some settings, and non-reflectively in others. Even when cannabis use was consciously driven, non-reflective processes were deployed to execute subservient acts, such as rolling joints.

Conclusions

Findings highlight specific processes and pathways that might be targeted to reduce cannabis-related harm.

Keywords:

• Automaticity
• cannabis
• dual process
• motivation
• qualitative
• theory

Introduction

College students are more likely to use drugs than others of comparable age.¹ In the UK, for example, cannabis use prevalence among people aged 16–24 years in 2018–2019 was 17%,² but a national survey found that 28% of UK undergraduates use cannabis.³ Of these, half use cannabis at least monthly.³ Similarly, a 2019 US national survey reported that 43% of college students had used cannabis in the past year, with 26% reporting use in the past 30 days.¹ Many college students who use cannabis experience few significant consequences,⁴ but typically around 30% of cannabis users experience problematic patterns of use.⁵ While reliable data are unavailable for UK undergraduates, in one US study, around 25% of past-year college student users reported cannabis use disorder,⁶ whereby cannabis continues to be used despite significant clinical or social impact.⁷ Cannabis use disorder represents a risk factor for mental ill-health and bronchitis,⁸ and heavy cannabis use has been variously linked to long-lasting deficits in concentration, attention and memory,⁹ and poorer educational outcomes.¹⁰ Calls have been made for greater intervention efforts to prevent consumption patterns that can lead to problematic use, and to reduce harms arising from problematic cannabis use among university students.^11,12

Designing effective cannabis harm-reduction interventions requires understanding and targeting modifiable determinants of consumption patterns. Dual process models propose that two information processing systems generate behavior.¹³ The reflective system involves deliberating over situational demands and available responses, and generates behavior relatively effortfully and slowly, via reasoned intentions. By contrast, the non-reflective system generates action rapidly, based on cued activation of non-conscious knowledge, which automatically prompts behaviors associated with such cues.

Psychological factors that sustain cannabis use have been categorized according to whether they operate reflectively, by informing conscious deliberation, or non-reflectively, facilitating automatic action. Research into undergraduate cannabis use has mainly focused on reflective factors. Students variously report using cannabis for stress relief, relaxation, pleasure, heightened sensuality, enhanced creativity, and to fit in with friends.^14–18 Studies undertaken among users with problematic consumption patterns have portrayed cannabis use as the outcome of non-reflective processes. Problematic users show greater attentional bias to marijuana cues,^19,20 with cannabis-dependent heavy users showing greater bias than non-dependent heavy users.²¹ Among problematic users, exposure to associated cues directly activates craving,^22,23 which in turn automatically triggers cannabis-seeking behavior and so consumption, without conscious deliberation.²⁴ People who report using cannabis in the absence of conscious intention, awareness or control tend to have increased consumption.²⁵

Previous research into the reflective and non-reflective processes driving cannabis use among students has been limited in two important ways. First, it has implicitly focused only on factors likely to determine the initiation of a cannabis use episode, but a different set of predictors may be involved in executing the actions that sustain ongoing episodes.²⁶ Second, research has tended to position cannabis use as either directed by reflective or by non-reflective processes. That is, cannabis consumption is commonly portrayed either as deliberative, driven by expected outcomes of cannabis use, or as impulsive, generated automatically by activated habitual or emotional associations.^27,28 Yet, if factors involved in opting to use cannabis can differ from those required to execute use, it is possible that people may make conscious, deliberative decisions to instigate a cannabis use episode, yet perform cannabis use actions automatically, or vice versa.²⁹

Cannabis use may be driven by both reflective and non-reflective processes concurrently.¹³ Hierarchical models propose that actions can be broken down into yet smaller actions³⁰; smoking a joint, for example, involves a sequence of actions involving preparing the joint, lighting it, and repeated inhalations.³¹ These acts may be broken down further; preparing a joint requires placing cannabis into cigarette papers, folding the papers, and so on. Thus, selection of a higher-level action unit (e.g., deciding to smoke a joint) in turn activates a sequence of lower-level units (e.g., prepare joint, light joint, inhale) that serve the higher-level goal.²¹ Behavior may therefore be driven by reflective processes on some levels, and non-reflective processes on others.³² A person could, for example, be automatically triggered to instigate cannabis use, yet smoke mindfully, to regulate intoxication.³³ Understanding the complementary roles played by reflective and non-reflective processes in consumption could aid the development of cannabis harm-reduction efforts. For example, pinpointing when people make conscious decisions to consume cannabis could inform point-of-choice interventions to reduce consumption among problem users,³⁴ whereas identifying automated aspects of consumption could inform the design of in-the-moment strategies to curtail ongoing cannabis use.³⁵

This study explored the potentially concurrent contribution of reflective and non-reflective processes to cannabis consumption among university students. We used qualitative methods to elicit in-depth, participant-led accounts of experiences of initiating and executing cannabis use, drawing on an interview schedule designed to probe reflective and non- reflective factors involved in cannabis consumption. While concern has been raised about whether people have insight into non-reflective influences on their actions,³⁶ people can adequately report action features suggestive of non-reflective influences, such as acting efficiently and effectively while attending to other tasks.^37,38

Materials and method

Participants and procedure

Participants were undergraduate students at universities in London, UK. In the UK, cannabis is an illegal (“class B”) substance, with penalties of up to five years’ imprisonment and an unlimited fine for use. Representative data on the prevalence of cannabis use or cannabis use disorder in UK or London-based undergraduates are not routinely collected.

Participants were recruited, over a predetermined data collection period of four months (November 2019–February 2020), via study advertisements placed in student common areas on one university campus, and social media (Facebook, Instagram). Potential participants who responded by email were sent a URL for a screening questionnaire, to self-declare eligibility (i.e., aged 18 years or over, current UK undergraduate, using cannabis at least 2–3 times per month, and last used in the last month).³⁹ We chose not to assess clinically problematic use (e.g., Cannabis Use Disorder), because this would have necessitated formal referrals to university support services in cases of dependency, which we felt would compromise trust between the researcher and participant.

Those meeting eligibility criteria participated in a face-to-face interview in a private room at a local university campus, or other private location, based on participant preference. On completion, participants were given a £10 Amazon voucher and a URL directing to a free-to-access national drug advice service. Study procedures received ethical approval from the host institution (HR-19/20-14246).

Eighteen participants were recruited, of whom 11 (61%) were female and 7 (39%) male (see Table 1). Mean age was 20 years, though two participants did not report age. Participants were most commonly White British (9 participants; 50%). Due to human error, participants’ year of academic study (e.g., Year 1) was not recorded. Participants typically consumed cannabis at least once daily (10 participants; 56%), and 10 (56%) had consumed cannabis on the day of the interview (but did not report being high during the interview).

Table 1. Participant characteristics.

Characteristic	Available N	Categories	N (%)
Gender	18	Female	11 (61%)
		Male	7 (39%)
Age (years)	16	–	Mean (SD) = 20.38 years Range = 19–22 y
Ethnicity	18	White British	8 (50%)
		White Irish	1 (5%)
		White Other	5 (28%)
		Mixed: White and Black African	1 (5%)
		Mixed: other	1 (5%)
		Any other	2 (11%)
Cannabis use frequency (N, %)	18	More than once a day	8 (44%)
		Once a day	2 (11%)
		5–6 times a week	3 (17%)
		3–4 times a week	2 (11%)
		Twice a week	1 (5%)
		Once a week	0 (0%)
		2–3 times a month	2 (11%)
Cannabis use recency* (N, %)	18	I am currently high	0 (0%)
		Today	10 (56%)
		Yesterday	5 (28%)
		This week	1 (6%)
		Last week	2 (11%)

* Cannabis use frequency and recency were reported using the Daily Sessions, Frequency, Age of Onset, and Quantity of Cannabis Use Inventory (Cuttler & Spradlin, 2017).

Interview schedule

Participants were interviewed by a male undergraduate Psychology student (SP), trained by the senior author (BG). Interviews were semi-structured, with topics capturing reflective (e.g., motives, expectations, experiences) and non-reflective processes (e.g., contextual cues, decision points, automaticity; see Supplementary File). Digital interview recordings were manually transcribed verbatim.

Analysis

Transcripts were analyzed using inductive Thematic Analysis procedures, based on realist assumptions,⁴⁰ to identify discrete clusters of factors supporting cannabis use. Rather than separating data deductively into two preconceived themes (i.e., reflective versus non-reflective factors) divorced of real-world context and meaning, we organized data into inductively derived themes capturing lived experience and identified reflective and non-reflective determinants as relevant within each of these themes.

Both authors independently coded three interview transcripts, assigning conceptual labels to events within the data. Discussion between authors resulted in a preliminary coding framework.⁴¹ Subsequent analysis, undertaken by SP, involved applying and iteratively refining the framework to code remaining transcripts, and extract, review, and name themes. Discussions between coders occurred throughout analysis, to ensure the credibility of emerging themes.

Reflexivity

The study was conceived by the first author (SP) and undertaken as undergraduate Psychology coursework under supervision of the senior author (BG). Both authors have personal experience of using cannabis in adolescence and early adulthood, prior to the study being planned. Neither author had conducted prior research in the cannabis domain. The second author is Senior Tutor, a role which involves coordinating a system of student welfare support, on an undergraduate Psychology programme, so has ground-level experience of student mental health and wellbeing concerns.

Results

Four overarching themes were extracted: cannabis use as a form of relaxation; cannabis use as social bonding; the symbolic-affective significance of cannabis use; and contexts and triggers conducive to cannabis use.

Cannabis use as a form of relaxation

This theme, which incorporated reflective components only, captured participants purposefully using cannabis to achieve a state of relaxation (“it allows my mind to kind of soften, and just relax”; Participant 1 [P1]). A state of cannabis intoxication was sought as a means of disconnecting from external pressures:

Even in downtime, I’m just thinking about work I have to do. But when I’ve smoked [cannabis], it’s a lot easier to… push that work to the side and focus on enjoying myself. (P15)

Some used cannabis as a symbolic signing off from work time, so initiated cannabis consumption only after work demands had been satisfied:

[I] only smok[e] once I’ve got everything done, I’ve got home, and I’ve done all the work that I want to do for the day. (P16)

Participants used the feeling of stress alleviation to focus on meaningful non-work tasks, and several described heightened creativity while intoxicated (“it opens up a creative side to me, definitely”; P14). Others used their relaxed state as a detached mental space for contemplation of life stressors:

[Cannabis] helps me to slow down and think calmly about things and not get too wrapped up in what is crazy third year life. (P3)

Participants also employed cannabis intoxication for physical relaxation, including inducing tiredness and regulating sleep patterns (“it helps me get into a better routine with my sleep”; P12). Some used cannabis to self-medicate:

[Cannabis] enhances so many different parts of my life that I’d otherwise have to rely on medication for. […] I never liked sleeping pills … [and] herbal remedies never worked. So just [smoking] a joint before bed, that works for me. (P13)

Cannabis use as social bonding

This theme, encompassing reflective components only, reflected use of cannabis as a means of connecting with other people (“you can really bond with people over being high, having a joint together”; P10). Consuming cannabis with others was seen by many as “part of the fun” (P18), experiencing a rewarding social event was for some a stronger motive than intoxication (“I treat weed as more of a social experience than like I want to get high”; P5).

Smoking a cannabis joint with others was reportedly motivated not only by a desire for a communal intoxication experience (“you want everybody else to feel how great you’re feeling right now”; P15), but also the symbolic act of sharing a precious resource with others. Using another’s cannabis could create a perception of obligation to share one’s own cannabis, fueling further consumption:

[Passing a joint is a] reach-out to being friendly … if someone’s handing me their weed to use, that’s like … someone offering me a bite of their food. I almost feel … an element of reciprocation. (P14)

Cannabis intoxication was used by many to enhance social experiences, by variously prolonging contact with others, offering shared experiences of altered perception (“the reason why it’s so good for socializing is because it just makes everything hilarious”; P14), or leading to deeper and more meaningful interactions:

Everyone’s really chill and open-minded [during a cannabis use session] … it’s just the bond that makes when you smoke with your friends, and then you [end] up in some really philosophical questions. (P8)

For some, cannabis consumption was central to their identity and affiliation with others and was purposefully employed to reaffirm group membership (“a very important part of my relation towards [my friends] is that we smoke together”; P1).

Several participants described discrete actions involved in cannabis consumption episodes as socially meaningful. While only one person was deemed responsible for “rolling a joint,” others often assisted—by, for example, finding filters—such that joint-rolling was for some an important “communal aspect” (P9) of consumption. Compelling implicit norms surrounding consumption routines were also cited, including that the person who constructed the joint has a “right to light it” (P13); “‘rollers’ rights’ is an unspoken, known agreement” (P11). Other norms dictated the number of inhalations deemed socially acceptable prior to passing a joint to another, and the direction in which a “joint” is passed around.

The symbolic-affective significance of cannabis use

This theme incorporated both reflective and non-reflective components. Although one participant reported that cannabis use could worsen a prior bad mood, cannabis was generally used to attain the goal of positive affective experiences. Some derived pleasure from the experiential satisfaction of smoking a joint itself. Many reported that intoxication afforded “a new perspective on things” (P1), and used cannabis to heighten enjoyment of everyday activities:

I’ll have a joint before I go to the gym sometimes because it makes me get into the groove a bit more. (P14)

Some used cannabis consumption as an incentive for completing work, or to counter-act unpleasant experiences:

If I’m at work and it’s raining outside … [and] I’ve got two hours left of my shift … I always think, I’m going to get home, get in my PJs, whack on some TV, and roll a joint. (P1)

Participants generally sought to regulate their intoxication to maintain an affectively optimal experience. Many preferred smoking over other consumption methods for this reason:

You can put out a joint. You can stop it and carry on. Pick it up and put it down. Whereas you can’t do that [when eating cannabis]. (P14)

Several participants reported enjoying self-constructed joints more than those rolled by others, due to greater control over ingredients, taste, and intoxication:

When [someone else] puts loads of [tobacco in], and crumbles hash on top, for me … smoking it, compared to one of my joints, I can taste a difference. It’s not as nice. (P1)

The act of rolling a joint appeared to have both reflective and non-reflective underpinnings. As a deliberated act, rolling a joint was a particularly valued aspect of the cannabis use routine. Some participants saw rolling a joint itself as relaxing, while some viewed it as self-expression (“it’s an absolute art”; P15). Several participants reported carefully considering which materials to use for the joint, such as filter width, papers, and tobacco (“deciding whether or not to put tobacco in is a very big decision for me”; P7) and felt that such preferences defined users (“all my friends say that I use a lot of tobacco in my joints”; P14). However, while several participants consciously strove to construct high-quality joints in social settings (“you want to be like, ‘look, I’ve rolled this sick [high-quality] joint!’”; P13), they reported automatically executing the rolling of the joint in pursuit of this goal (“[once you] know your shit, you do it without really thinking about it”; P1). Some likened the automaticity of rolling joints to playing an instrument or exercising simple muscle memory (“all I have to do is put my hands there, and my hands just do the work for me”; P15). The finer-grained acts involved in constructing the joint only required conscious cognitive resources when normal progress was disrupted (“if something goes wrong [when rolling the joint], then you have to concentrate”; P13).

Contexts and triggers

This theme, which addressed the settings conducive to cannabis, incorporated both reflective and non-reflective pathways to cannabis use. Participants were selective in the settings in which they consumed cannabis. Contexts in which cannabis use was deemed inappropriate were characterized by goals incongruent with intoxication, such as needing to complete work tasks (“When you’ve got so much stuff to do you just can’t afford to be slow and dopey”; P15).

Participants also regulated their consumption in anticipation of future settings:

If I have a lot of work to do the next day, I would still smoke a joint, but I wouldn’t get really stoned, [I would] only have one. (P13)

One participant reported a distinction between the goals of alcohol and cannabis intoxication, such that cannabis consumption was not appropriate in alcohol-dominated settings:

If we’re in the smoking area with a couple of pints and some cigarettes, I think if I was to pull out a joint then … [people would think] not here, not now… it’s the setting for drinking, not smoking. (P11)

In contexts in which cannabis intoxication was deemed permissible, the urge to consume cannabis could result from deliberation, or as an impulsive, immediately triggered response: If [I’m] alone then [I make the decision about whether to use cannabis] very mindfully … Because normally I wouldn’t consider it. But if someone in the house is smoking [a joint], then I don’t really think before I agree to. (P4)

Some participants reported smoking cannabis as a default, habitual action in regular consumption contexts:

I always have a joint before bed. […] Rather than actually just having a joint because I feel like it, I do it kind of, impulsively. […] It takes more effort for me to be like, “Oh, okay, I’m not going to have a joint [before bed].” (P1)

Notable triggers—such as the sight of cannabis or smoking paraphernalia, or invitations from others to smoke—were those that prompted participants to pass a ‘point of no return’, whereby they fully committed to using cannabis:

When I have the conversation with my friends that we’re going to smoke weed … after that, I’m pretty much guaranteed to do it. There’s no going back after that point. (P7)

Passing the “point of no return” rapidly translated into consumption except where cannabis was unavailable, when it prompted deliberative cannabis-seeking behaviors:

If I’ve convinced myself that I have weed in the house but [I then discover that] someone’s smoked it … when I get home, [then] I’m like ‘it’s fine, we’ll have some weed’, and … I have to get it. (P15)

Discussion

Interviews with undergraduates who regularly use cannabis illustrated how reflective and non-reflective factors operate concurrently to support cannabis use episodes. Modifying non-conscious influences on action requires different techniques to those required to change deliberative influences.³² Findings may enhance the precision of interventions for students with problematic cannabis use patterns, by highlighting specific processes and pathways that might be targeted to reduce cannabis-related harm.

Past research has tended to depict problematic cannabis use as the result of non-reflective factors such as enhanced cue-reactivity,^19,21 and non-problematic use as resulting from conscious decision-making, driven by relaxation, socializing, and pleasure motives.^15,17,18,42 A key assumption underpinning such research has been that cannabis use is generated either via reflective decision-making, based on deliberation over outcome expectancies, or non-reflective processes, which can prompt action without conscious awareness, control, or intention. Yet, our data, collected among frequent but apparently non-problematic users, demonstrated that both reflective and non-reflective processes work complementarily to generate cannabis use, in two ways. First, the contribution of reflective versus non-reflective processes to the instigation of cannabis use episodes varied across occasions. Conceptually similar influences affected behavior deliberatively in some instances, but impulsively in others. For example, participants consciously decided to smoke with friends to maintain social affiliations, but also reported that being offered a joint by friends could automatically cue consumption. Second, on many occasions, reflectively initiated action was enacted non-reflectively: even where users were consciously pursuing cannabis-related goals, non-reflective processes were recruited to execute lower-level actions that serve those goals, such as the hand movements required to roll joints.³¹ This attests to the hierarchical nature of cannabis use, and how it may be supported by multiple processes concurrently at different levels of analysis.¹³ Future research should move away from asking whether cannabis use is reflective or non-reflective, and instead ask what roles are played by reflective and non-reflective processes in sustaining cannabis use.⁴³

Among our participants, reflective processes appeared to more often determine whether or why to use cannabis, rather than how to perform the act of using cannabis. Participants reported, for example, considering whether social settings were conducive to cannabis consumption, and consciously deciding to use cannabis to unwind. Nonetheless, action was non-reflectively selected in some cases. Some participants reported that cannabis use had become automatically triggered by some contexts, and that it would be effortful to inhibit cannabis use urges, both of which are characteristic of habitual behavior.³³ Similarly, some participants reported reflectively executing some aspects of cannabis use routines. For example, while many participants described rolling joints effortlessly and automatically, enactment difficulties prompted deployment of attentional resources to rectify problems.³⁰ We also observed the potential for reflective processes to satisfy non-reflectively generated cannabis consumption urges. Even when the urge was automatically triggered, when cannabis was unavailable participants consciously engaged in behaviors directed toward the goal of obtaining cannabis.

Our results have important implications for understanding cannabis use. First, they demonstrate that both the decision to use cannabis, and the act of using cannabis, can be driven by reflective processes, non-reflective processes, or some combination of both. Second, they illustrate that, in some instances, commitments and urges to use cannabis triggered by non-reflective processes can require activation of reflective processes to generate action directed at satisfying such urges.⁴⁴ Third, and relatedly, findings speak to the importance of identifying key decision points in cannabis consumption, beyond which people are unlikely to disengage from the goal of using cannabis. Motivation-based behavior change interventions are unlikely to be effective when presented after such decision points have been reached.³⁴

Our findings offer recommendations for intervening to reduce cannabis consumption among problem users, or to prevent cannabis consumption patterns that can lead to problematic use. Strategies might focus on reflective or non-reflective routes to behavior change. For example, participants’ prioritization of life demands over cannabis use suggests that persuasive messages emphasizing the adverse impact of cannabis on concentration and attention might influence cannabis decisions. While some participants used cannabis to induce sleep or alleviate stress, long-term use can build tolerance and exacerbate sleep problems and stress.^45,46 University wellbeing teams might provide support for safer alternatives to cannabis for sleep issues. Intervention strategies might also target non-reflective processes that support cannabis use. For example, implementation intentions that specify alternative responses to known triggers might be promoted to prevent students from automatically initiating a cannabis use episode on exposure to other cannabis users^47,48 (but see Ref. 49). Additionally, interventions might disrupt consumption sequences by inserting salient cues to cessation into cannabis contexts. Messages might for example be placed on rolling papers to prompt deliberation over whether to curtail cannabis consumption within an ongoing session.

Study limitations must be acknowledged. The representativeness of findings from a UK undergraduate sample to American college students might be questioned, given legal, regulatory and socio-cultural differences between UK and US cannabis use contexts. However, we limited the focus of our interview schedule to experiences of cannabis consumption rather than procuring cannabis, and participants reported consuming cannabis in private rather than public settings. While legal and regulatory contexts necessarily frame cannabis use, there is no reason to expect that the lived experience of consuming cannabis privately—and the reflective and non-reflective processes underpinning such experiences—should differ between UK and US undergraduate students.

Interview methods may have overstated the involvement of reflective processes in cannabis use. People may misconstrue actions truly directed by non-reflective processes as arising from conscious deliberation.⁵⁰ Alternatively, participants may have deemed automated, mechanistic actions as unworthy of discussion. More sophisticated data collection techniques may be needed to provide a more comprehensive picture of non-reflective factors sustaining cannabis use. While we aimed to generate suggestions for tackling problematic cannabis use patterns, we assessed only cannabis frequency and recency among our sample. While most of our participants used cannabis at least once daily, and some had consumed cannabis on the day of the interview (though none reported being high at interview), we did not assess the patterns of abuse and dependence that characterize problematic use.⁵¹ Additionally, the contribution of reflective and non-reflective processes documented in our sample may not generalize across all users. Actions such as rolling a joint become automatic as people acquire the skills necessary to effortlessly execute them. Inexperienced cannabis users are therefore more likely to act mindfully. Conversely, people with cannabis use disorder may be guided more by habit.⁵² While our results demonstrated the potential of reflective and non-reflective influences to concurrently support cannabis use, the specific contribution of such processes will likely differ across users, behaviors and contexts.³²

Our study is the first to our knowledge to document the concurrent roles of reflective and non-reflective processes in cannabis use behaviors. We showed not only that people may engage in cannabis use via reflective or non-reflective processes in different settings, but also that such processes may operate in concert, such that influences on action selection may differ from those required to execute such actions. These findings point to the complex, hierarchical nature of cannabis consumption, and call for precision in identifying targets for cannabis use interventions and the processes by which such targets may reduce harm.

Conflict of interest disclosure

The authors have no conflicts of interest to report. The authors confirm that the research presented in this article met the ethical guidelines, including adherence to the legal requirements of the United Kingdom and received approval from the King’s College London Research Ethics Office.

Funding

No funding was used to support this research and/or the preparation of the manuscript.