Wolfe ONLINE DRINKING: EXPLORATORY STUDY 1
Online Drinking: An Exploratory Study of Alcohol Use and Intoxication during Internet Activity
Wendy L. Wolfe
ArmstrongAtlanticStateUniversity
Despite the commonplace use of the internet for socializing and recreating, little is known about alcohol use during online activity. This study investigates the prevalence of online drinking (drinking/intoxication during internet use) in an American college student sample, differences in internet use associated with online drinking, consequences of online drinking, and the relationship between alcohol problems and internet addiction. Because social anxiety has been found to increase risk for both internet addiction and alcohol problems, this study also examines the relationship between social anxiety and online drinking. Results demonstrate that online drinking is commonplace, and tends to occur in conjunction with entertainment/social networking-based internet activities. For females, a significant positive correlation was found between scores on problem drinking and internet addiction screening measures. Online drinkers also had significantly higher internet addiction scores. Given the apparent lack of a role for social anxiety in explaining online drinking, other directions are proposed to advance this new area of research within the alcohol/addictions field.
Skyping, gaming, friending, tweeting – our society has embraced the internet as an alternate, and sometimes preferable, social universe. This is particularly true for young people. According to the PewResearchCenter’s Internet and American Life Project (2010), 95% of 18 to 29 year-olds use the internet, more than any other age group. Among internet users, teens and young adults are more likely than older adults to use the internet for entertainment purposes, such as gaming, watching videos, and downloading music; and for social purposes, such as using social networking sites, blogs, and instant messaging (Pew Research Center, 2009). While internet use has many beneficial aspects, it can also be misused and overused, with internet scams and cyber bullying as examples of the former and internet addiction an example of the latter. Indeed, internet addiction is under review for inclusion in the DSM-V and has been conceptualized as an impulse control disorder (similar to compulsive gambling) involving symptoms such as preoccupation with going online, spending increasing amounts of time online, having difficulty cutting back on internet use, and continuing online activity in the face of negative consequences (Young, 1996).
Despite the conceptualization of excessive internet use as an addiction, and the increasingly commonplace use of the internet as a vehicle for socializing and recreating, the relationship between internet activity (and addiction) and substance use has garnered little research attention. In one of the few studies in this area, Ko et al. (2008) examined the co-occurrence of problematic alcohol use and internet addiction among high school students in Taiwan and found a significant positive correlation between the two. Further, they found that those with both problematic behaviors were more often males with co-occurring problems in the areas of family conflict, family alcohol use, and deviant behaviors among friends (Ko et al., 2008). Yen, Ko, Yen, Chen, and Chen (2009) also investigated the relationship between problematic alcohol use and internet addiction, but among college students in Taiwan. They also found a significant positive correlation between the two problem areas, and additionally found both to be positively correlated with depression. In the only published investigation of internet and alcohol use using an American sample, Epstein (2011) sought to examine the association between occurrence of drinking (lifetime and past month) and computer use (both time spent on the computer, and self-reported frequency of engaging in various online activities) in a sample of 13-17 year-olds. Participants who reported using alcohol in the past month reported significantly more time spent on the internet engaging in non-academic tasks than participants who denied using alcohol in the past month. Moreover, participants reporting alcohol use at some point in their lifetime reported significantly greater frequency of internet use for social networking and for downloading and listening to music (past month use was also associated with greater frequency of the latter) than participants who denied ever having used alcohol (Epstein, 2011).
These studies suggest a relationship between drinking and internet use, and (in the case of the studies conducted in Taiwan) an association between problematic alcohol use and internet use. However, it is unclear if the correlation between problematic alcohol use and internet addiction would generalize to an American population. While the Epstein study opens the door to investigating this connection in an American population, the investigation focused on occurrence, duration, or frequency of use and did not include measures of problematic use (of either drinking or internet use). In addition, there remains no empirical information about whether alcohol use is combined with online activity and, if this is the case, what typifies online drinking. An article on the Japan Trends website focuses on the rising popularity of online drinking among young people through “net nomikai” in which individuals socialize through webcams while drinking (Andrews, 2010). However, the author notes that there is no research or data to support this “boom.” In addition to gleaning insight into similarities and differences between online drinking and more traditional alcohol use behavior (e.g., drinking during in-person social activities), the relevance of online drinking is clear given the immediacy of online activity and the disinhibiting effect of alcohol. For example, it is quite possible that problematic online behaviors such as cyber bullying or sending/posting sexually explicit photos are more likely when alcohol and internet use are combined. Finally, given that social anxiety has been identified as a risk factor for both internet addiction (e.g., Caplan, 2010;Chak & Leung, 2004; Kraut et al., 1998) and problematic alcohol use (e.g., Crum & Pratt, 2001; Kushner, Sher, & Beitman, 1990; Morris, Stewart, & Ham, 2005; Thomas, Randall, & Carrigan, 2003), the question begs as to whether individuals who combine internet activity (particularly activity that is social in nature) with alcohol use are higher in social anxiety than those who do not.
The current study sought to gather data to begin to better understand the relationship between alcohol use and internet activity. The following questions were addressed:
What percent of college student drinkers consume alcohol during online activities, or use the internet while under the influence of alcohol (hereafter combined and referred to as online drinking, unless discussed separately)?
Are there online activities that are more likely to be engaged in while drinking or under the influence of alcohol (e.g., interactive, social activities such as instant messaging)? Relatedly, do online drinkers demonstrate different internet use patterns in general, as compared to alcohol users who do not engage in online drinking?
What consequences are encountered when people combine drinking with internet use? The well-publicized case of a Morgan Stanley commodities trader who made $10 million in risky online trades while intoxicated (Hosking, 2009), suggests that the immediacy of online activities can lead to unique consequences for those who use the internet while under the disinhibiting effects of alcohol.
Given the association between social anxiety and problematic use of both alcohol and the internet, are individuals who combine alcohol use and internet activity more socially anxious than those who do not? It has been found that socially anxious individuals use alcohol with the expectation that drinking will reduce their anxiety in social situations (Carrigan & Randall, 2003). Are socially anxious individuals more likely to engage in interactive internet activities while drinking or under the influence of alcohol, as compared to their general involvement in interactive online activities?
Can the relationship between problematic use of both alcohol and the internet, as reported by Ko and colleagues, be replicated in an American population?
Finally, given that sex differences have been found for both the relationship between social anxiety and alcohol (e.g., Morris et al., 2005) and between problematic alcohol use and internet addiction (Ko et al., 2008), do males and females differ in terms of the above questions?
Method
Participants
Students (128 males, 169 females) at a regional state university in the southeast United States participated in the IRB-approved study in exchange for class credit, with an average age of 23.65 years (SD = 6.34). Caucasians (68%) and African-Americans (20%) comprised most of the sample. The majority were undergraduates (95%) and were divided equivalently (20-27%) across academic classifications and academic fields of study. Fifty-seven percent of the participants could be characterized as traditional college students (i.e., 17-22 years of age, single, and without children), which is reflective of the university enrollment as a whole (ArmstrongAtlanticStateUniversity, 2010). Participants’ reported internet use was typical, however, for a young adult population in that 89% reported using the internet either daily or several times daily, with an average of 13.18 (SD = 13.19) hours online per week estimated by participants. The majority (81%) reported their favorite location for internet use as their home. Participants reported that 60% of their online time is spent engaged in leisure activities (email, social networking, online research, listening to music, and general web surfing). Because this investigation focused on online drinking, only current drinkers were eligible to participate in the study. Drinking status was verified by participant response to item 1 on the AUDIT (i.e., How often do you have a drink containing alcohol?) and those few non-drinkers (n = 25) who erroneously volunteered for the study were excluded from data analyses.
Measures and Procedure
After providing informed consent via an informed consent document presented to participants at the outset of the survey, the following measures were administered using the questionnaire administration program, Survey Monkey:
Internet Use. A 6-item Internet Use Questionnaire was developed by the author to assess the frequency and locations for internet use. Participants were asked to report the locations in which they use the internet (e.g., home, school/work, internet café or other public location), their preferred location, their frequency of internet use, estimated hours per week of internet use, and their estimated hours per week of academic/work-related internet use versus leisure use. The Internet Activities Questionnaire (IAQ) was also developed by the author to assess self-reported likelihood of engaging in various activities while online. The items were selected based a literature review of research on internet use behavior. Participants used a Likert rating scale from 1 (very unlikely) to 5 (very likely) to indicate how likely it would be that they would engage in a variety of online activities (see Table 1 for the list of internet activities included in both the IAQ and the AIAQ, described below).
TABLE 1 Likelihood of Engaging in Various Internet ActivitiesDuring Online Drinking
Activity / n / M / SD
Social networking sites / 135 / 3.71 / 1.27
Watch/downloading video / 149 / 3.48 / 1.31
Listen to/downloading music / 151 / 3.38 / 1.44
General web surfing / 151 / 3.21 / 1.38
Interactive chat / 99 / 2.75 / 1.51
Reading/Responding to Email / 148 / 2.59 / 1.45
Cybersex websites / 82 / 2.51 / 1.39
Reading/Posting to Blogs / 101 / 2.33 / 1.40
Gaming / 72 / 2.13 / 1.15
Interactive Gaming / 59 / 2.12 / 1.35
Shopping / 116 / 2.05 / 1.28
Research / 143 / 1.95 / 1.19
Interactive Cybersex / 38 / 1.95 / 1.37
Dating Websites / 38 / 1.84 / 1.24
Roleplay Gaming / 46 / 1.78 / 1.26
Gambling / 32 / 1.44 / .91
Note. Only participants who reported using the internet while drinking or intoxicated are reflected in the table. For each activity, ratings are only included for participants who reported ever engaging in that particular online activity. Ratings were assigned using a 1 (Very Unlikely) to 5 (Very Likely) scale in response to the question, “How likely are you to engage in the following internet activities while drinking alcohol or while under the influence of alcohol?”.
Alcohol and Internet Use. Participants were asked two items to assess if they had ever consumed alcohol while using the internet or had ever used the internet while under the influence of alcohol. Those who responded yes to at least one of the items went on to complete the Alcohol and Internet Activities Questionnaire (AIAQ) and the Alcohol and Internet Consequences Questionnaire (AICQ), both developed by the author. The AIAQ asked participants to rate the likelihood of engaging in the same 16 online activities as in the IAQ, during online drinking. By including the same activities on the IAQ and the AIAQ, comparisons could be drawn between the likelihood of engaging in each online activity when drinking or intoxicated, versus general likelihood of engaging in the activity. On this measure, only participants who ever engaged in that particular online activity were asked to rate their likelihood of engaging in the activity while drinking or intoxicated, from 1 (very unlikely) to 5 (very likely). The AICQ was used to assess whether participants had ever experienced 11 possible consequences related to online drinking (see Table 2). Since research has not explored online drinking consequences, items were selected based on a review of measures of general alcohol use consequences (such as the Drinker’s Inventory of Consequences; Miller, Tonigan, & Longabaugh, 1995) and a literature review of problematic internet-related behaviors. Participants indicated “yes” or “no” to each item and total scores reflect the total number of consequences reported by participants.
TABLE 2 Consequences of Combining Alcohol and Online ActivityConsequence / % %Reporting Yes / %Males / % Females
Spent too much time online / 61 / 54 / 67
Neglected something important / 53 / 52 / 54
Found activity more enjoyable / 51 / 61 / 42
Said/wrote something later regretted / 44 / 44 / 44
Felt more comfortable being myself / 41 / 46 / 36
Got into an argument / 38 / 34 / 42
Felt less anxious when online / 30 / 32 / 28
Did something I later regretted / 29 / 27 / 31
Gave too much info about self / 29 / 24 / 33
Spent too much money / 23 / 24 / 22
Drank more than intended / 14 / 18 / 11
Note. Above reflects percentage of online drinkers who reported the consequence has ever occurred to them as a result of their use of the internet while drinking or intoxicated.
Internet Addiction Test. The IAT was developed by Young (1996, 1998) based on DSM-IV criteria for impulse control disorders such as
based on DSM-IVcriteria for impulse control disorders such as compulsive gambling. Participants are asked how often they have experienced 20 situations related to their internet use, capturing the effect of internet use on their relationships, emotions, sleeping, and daily activities, among others. Scores range from 20-100, with scores greater than 40 indicating frequent problems and scores greater than 70 indicating significant problems related to internet use. Factor analysis of the IAT has yielded six subscales, with good internal consistency. Concurrent validity has also been established (Widyanto & McMurran, 2004).
Alcohol Outcome Expectancy Scale. The AOES was administered to assess the outcome expectancies that participants have for their alcohol use. The measure was developed by Leigh and Stacy (1993) and asks participants to rate the likelihood that 34 different outcomes would occur to them if they were to consume alcohol. The outcomes include negative outcome categories (negative social, emotional, physical, and cognitive/performance effects), and positive outcome categories (social facilitation, fun, sex, and tension reduction). The subscales and the positive and negative dimensions have been found to have good internal consistency (Leigh & Stacy, 1993). While both positive and negative expectancies have been found to be associated with drinking behavior, endorsement of stronger positive expectancies has been found to account for greater variability in alcohol use (Leigh & Stacy, 1993). A particular interest in this investigation was the expectation that alcohol will facilitate social interaction and reduce tension.
Social Anxiety and Distress Scale. The SAD is a well-established measure of anxiety in social situations. It consists of 28 true-false items and yields a total score with excellent internal consistency and good one-month test-retest reliability (Watson & Friend, 1969). There is no established cut-off score (higher scores indicate greater social anxiety), although past research has yielded mean scores and standard deviations for college males and females.
Alcohol Use Disorders Identification Test. The AUDIT is an established screening measure for alcohol problems and was developed in collaboration with the World Health Organization to provide a gender and culture-neutral screening measure for the early detection of alcohol problems (Babor, Higgins-Biddle, Saunders, & Monteiro, 2001). It contains ten items that assess both quantity and frequency of alcohol use, as well as dependence symptoms and the occurrence of problems related to alcohol use. The AUDIT yields a total score of up to 40, with a criterion of 8 typically used to identify individuals with potential alcohol problems.
Demographic Questionnaire. A questionnaire was used to gather demographic information on participants, including their age, sex, race, class standing, major program of study, marital status, and residence status.
Preliminary Analyses
As research on alcohol use during online activity is a new area of investigation, several questionnaires were developed by the author due to an absence of established measures of the constructs of interest. Specifically, the IAQ assessed reported likelihood of engaging in a variety of internet activities when online, the AIAQ assessed reported likelihood of engaging in those same activities during online drinking, and the AICQ assessed whether a variety of consequences had ever occurred as a result of online drinking. Activities surveyed by the IAQ and AIAQ are listed in Table 1 and consequences that were included in the AICQ are listed in Table 2. The Cronbach’s Alphas for the IAQ, AIAQ, and AICQ are presented in Table 3, along with preliminary validation in the form of correlations with other measures used in the
TABLE 3 Internal Consistency and Concurrent Validity of NewMeasures
Measures / Cronbach’s Alpha / Correlation with Measures
Internet Use Questionnaire
-online frequency
-weekly hrs online / Not
assessed / IAT (r = .15*)
IAT (r = .30***)
IAQ / .63 / IAT (r = .42***)
AIAQ / .66 / AUDIT (r = .26***),
IAT (r = .31***)
AICQ / .77 / AUDIT (r = .34***),
IAT (r = .41***)
Note: Other than Internet Use Questionnaire, analyses above were based on total scores for the measures. * p < .05, *** p < .001. IAQ = Internet Activities Questionnaire, AIAQ = Alcohol and Internet Activities Questionnaire, AICQ = Alcohol and Internet Consequences Questionnaire, IAT = Internet Addiction Test, AUDIT = Alcohol Use Disorders Identification Test.
study that were expected to yield associations with the constructs of interest. As can be seen in the Table, internal reliability is marginally acceptable for the IAQ and AIAQ, but good for the AICQ. However, with regard to the IAQ and AIAQ, it has been noted that internal consistency in the .6 to .7 range is acceptable for exploratory studies (Garson, 2011), which accurately describes the nature of this investigation. Total scores for the IAQ and AIAQ were not a focus of this investigation and were not used in the principal analyses. However, IAQ and AIAQ total scores were examined relative to other potentially related measures for the purpose of exploring construct validity. Expected correlations between the newly developed measures and more established measures of related constructs were found, providing preliminary support for their validity. For example, greater likelihood of engaging in a variety of online activities on the IAQ was significantly positively associated with scores on the IAT, whereas greater likelihood of engaging in various online activities while drinking and reports of greater consequences of online drinking were significantly positively associated with both IAT scores and AUDIT scores.