CHAPTER 3 RESEARCH METHODOLOGY
4.6 THE ROLE OF PERCEPTIONS IN PERSONAL DRINKING BEHAVIOUR IN THE CONTEXT OF OTHER FACTORS THE CONTEXT OF OTHER FACTORS
5.3.2 P ERCEIVED D IFFERENCE FROM THE N ORM
In the first instance, the analysis provided evidence of extensive misperceptions of the drinking norm, similar to the findings of other researchers (e.g. Carey et al., 2006; Kypri & Langley, 2003). Across various measures of quantity and frequency of drinking, the majority of respondents consistently perceived the drinking behaviour of various peer reference groups as more permissive than their own. More specifically, for each day of a typical week, respondents perceived their friends, co‐residents and students in general as drinking more than they do. This illustrates the consistency and depth of perceived self‐other differences for peer behaviour on a day‐to‐day basis. Further analysis showed that the total perceived weekly drinking of all reference groups was statistically significantly more than personal drinking, indicating that students perceived their own drinking as significantly more conservative than the norm, similar to other research utilising a similar measure (Lewis & Neighbors, 2004: 336). The perception of self‐other differences was so profound that male students even thought the drinking norm for female students exceeded their own drinking. A study in the United States incorporating 130 higher education institutions (N = 76 145) investigated the number of alcoholic drinks consumed the last time students ‘partied’/socialised. It found that almost three in four students overestimated the drinking norm at their college/university (Perkins et al., 2005: 473). The results of the latter study were reiterated by findings from another large‐scale study conducted in Canada (Perkins, 2007: 2650).
Similar to the descriptive norm, my results revealed that in general students perceived vast differences between their own approval of drinking and the perceived approval of others. The majority think other students approve more of drinking than they personally do. Consistent with other research (Prentice & Miller, 1993: 245, Study 1), Stellenbosch students on average thought that they were more conservative than their peers. In their meta‐analytic integration of 23 studies, Borsari and Carey (2003) found that 91% of hypothesis tests determined that survey respondents perceived other students as drinking more or holding more permissive attitudes towards alcohol consumption than they themselves had.
Consistent with drinking and approval of drinking, there were also significant self‐other differences in terms of academic behaviour. In general, students at Stellenbosch University perceived the general academic norm as worse than their own academic behaviour. The qualitative results of this study confirmed the survey findings in this regard. Similarly, other research looking at academic success found that, compared to themselves, students perceived their peers as less academically motivated (Abhold, Hall & Serini, 1999, as cited in Berkowitz, 2004:
28). The current results more particularly showed that students perceived their close friends, students in their course and students in general as working less and performing worse than they personally do.
Furthermore, consistent with other research (Neighbors et al., 2008: 580; Carey et al., 2006: 6; Perkins, 1997: 189), my research shows that the more socially distant the reference group, the more permissive the perception of the norm. Therefore, although students in general perceive all peer reference groups as consuming more alcohol than they do, approving more of drinking than they do, and academically working less and performing worse than they do, they also perceive more similarity between their own behaviour/approval and that of closer reference groups. This difference in perceptions of similarity also explains the lower correlations between own behaviour and the perceived behaviour of more distant reference groups described in the previous section (see 5.3.1). Consistent with other research, across the descriptive and injunctive norm, students seem to perceive their friends’ behaviour and attitude as more similar to their own, as shown by the stronger correlations and the smaller mean difference, than that of more distant reference groups (Prentice & Miller, 1993: 245, Study 2). For academic behaviour, the same trend was found: students observe larger self‐other differences with more distant reference groups, such as students in their course, and even more so with students in general.
Perkins (1997: 186‐191) suggests that these perceived self‐other differences are partly caused by errors in attribution, and the visibility and overrepresentation of drinking behaviour in students’ public conversation. Errors in attribution largely pertain to the actor‐observer effect, i.e. when differing causes of behaviour are allocated to own behaviour and the behaviour of others (Baron et al., 2006: 102). When students reflect about the causes of their own behaviour, they tend to attribute the causes more to external situations or the social environment. This is known as fundamental attribution error (Baron et al., 2006: 97). However, when assigning causes to others’ behaviour, they tend to attribute the causes more to those individuals’ personal disposition or attitude rather than to situational causes. This is explained by the limited knowledge individuals have of other students, for two reasons. Firstly, the only knowledge individuals have available to attribute causes to student behaviour is the discrete observations of students public behaviour. Secondly, individuals do not have direct access to information about the dispositions of other students; access is indirect, possibly only through the observation of public behaviour. Students may therefore think that the random occurrences of drunkenness they witness are the general behaviour of students and may attribute the causes of that behaviour to the personal attitudes of their peers because they have neither knowledge that could explain these actions, nor knowledge about the continuous behaviour of peers. In contrast, for their own behaviour, individuals have more intimate knowledge of the causes of their own behaviour. Consequently, as results showed, students tend to perceive themselves as more conservative than other students. They therefore assign a larger dispositional cause to other students’ behaviour than to their own behaviour. However, because perceptions of dispositions will partly be based on the perceived behaviour of others, the exaggerated perception of peer behaviour also explains the more permissive perception of other students’ dispositions. Qualitative interviews confirmed the overrepresentation of talk about drinking behaviour in public conversation, which further contributes to the cognitive visibility of drinking and the exaggerated and false perceptions of others’ behaviour. Hence, attribution theory also explains why students have more exaggerated perceptions of more distant reference groups than they have of their close friends. When students reflect about their
friends, they are referring to their perceptions of people they actually know and interact with, so their perceptions are based on more continuous observations and knowledge. However, when reflecting about general students, for example, they rely on a perception based on a subjective idea of an abstract construct, i.e. the student population as a whole. They do not know every individual in this population and therefore cannot have an accurate perception of all students’ behaviour or attitudes because they simply do not have access to this information. In accordance, research by Holtz and Miller (1985: 890) showed that students tend to be less certain about the behaviour of an out‐group than an in‐group. Nevertheless, the perceptions of ‘an amalgamation of others’ will still partly be based on personal observation, because students only have their personal experiences and thus subjective views to rely on in attributing causes to others’ behaviour (Neighbors et al., 2008: 580). Observations of discrete experiences affirm their broader expectations of the general campus environment. Therefore, the actor‐observer effect and fundamental attribution error will be more enhanced for distant others, meaning students will be more inclined to attribute the observed behaviour of general students to dispositional causes than doing so for the behaviour of their close friends.
For academic behaviour, the same social processes come into play as with drinking behaviour except, as expected, peer academic behaviour is underestimated. However, as with drinking, the perceived academic behaviour of peers can still be seen as more permissive, because it is associated with worse performance and less effort. Here the possible low social desirability of academic behaviour, its low visibility and underrepresentation in (desirable) conversation might have the inverse effect than for drinking behaviour, leading to the underestimation of the academic norm. Of course, students do observe peer academic behaviour, because it forms an integral part of the university experience, but it is not as memorable as the perceived drinking‐ related behaviour of students and is therefore not talked about that much or as positively. This effect will again be amplified for more socially distant reference groups due to the lack of knowledge of their actual behaviour and the consequent generalisation of discrete observations.
Additionally, I postulate a relation between the underestimation of academic behaviour and the overestimation of drinking behaviour and vice versa. For students, both these activities potentially consume a significant amount of their free time. However, the perceptual salience of drinking behaviour in social life and the consequent overestimation of the norm might contribute to the underestimation of the academic norm, as students erroneously perceive the majority of other students as heavy drinkers. As a consequence, academic responsibilities may be perceived as taking a backseat to social drinking behaviour. Alternatively, the perception of the dispositional cause of the drinking behaviour of other students may be augmented by the presence of academic behaviour: even though it might be expected to inhibit drinking, heavy drinking by the majority (or the perception thereof) still occurs (Baron et al., 2006: 97).
During interviews, the issue of achieving a balance between drinking behaviour and academic behaviour also surfaced. This perceived balance of peer behaviour gravitates towards social drinking behaviour, with a perceived necessity for performing academic work directly before tests, exams and assignment deadlines. Although this perceived norm might be seen in a positive light because students make time for both social and academic activities, it can also be interpreted as a detrimental norm to subscribe to because it caters for students’ social drinking habits by legitimising the need to study and work only when it is really necessary. This may provide normative support for not focusing continuously on the academic aspects of campus life, such as going to class or reviewing work, but merely working hard on some occasions. This scenario places more pressure on students and has the added disadvantage that they may possibly underestimate
the task at hand and therefore the time needed to adequately prepare. In turn, this may translate into poor performance in tests or the late submission of assignments for some.
Although it is apparent from the above discussion that students perceive other students’ behaviour as more permissive than their own when asked to respond by selecting a singular category in the questionnaire, the qualitative interviews revealed that participants perceive the student population as stratified, as consisting of cohorts with differing drinking habits, levels of alcohol consumption and academic behaviour. Among students, there is variation in the perceived size of these strata. The perception of the student population as stratified in terms of level of alcohol consumption might therefore mitigate the social influence of the erroneously perceived drinking norm. According to Prentice and Miller (1993: 243‐244), the power of a norm to influence attitude or behaviour largely depends on its perceived universality. ‘As consensus (or the appearance of consensus) breaks down, the norm loses its influence’ (Prentice & Miller, 1993: 244). Their research accordingly showed that the average students is not as comfortable with drinking habits on campus as is universally perceived (Prentice & Miller, 1993: 245, Study 1). Consistent with this, my results showed that on average students thought the majority of the student population drank heavily, pointing to a degree of universality in their perceptions of a permissive drinking norm.
While it is clear that students are aware of the different subpopulations on campus, they do not assign enough weight to the moderate drinkers. When asked to provide an indication of either their friendship group or typical students’ drinking behaviour, they tended to select a more permissive behaviour and attitude than their own. As explained earlier, this might be because of the low visibility of lighter consumers. The interviewees were aware of a group of heavy drinkers on campus and mentioned highly visible behaviour associated with drinking. Participants also indicated that they thought the actions witnessed by their peers had an influence on their peers’ perceptions of the student population. As proposed by proponents of the social norms theory (Perkins, 1997: 190), these observations exacerbate students’ perceptions of the drinking norm, or rather, due to the evidence of a stratified perception of the student population, their perception of the size of the heavier drinking group.
Proof of a general perception of a more permissive social norm than existed in reality, prompted an investigation into the relationship between these perceived differences and students’ own behaviour. Such an analysis might posit possible explanations for the perception of self‐other differences, because theory on misperceptions suggests that own behaviour plays a role in how we perceive ourselves in relation to our social environment (Berkowitz, 2004: 7‐8).