Attributes of the drinker prototype among Thai adolescents

Abstract Favorability of and perceived similarity to the attributes of the prototypical teen drinker (a known risk factor for adolescent alcohol use) have not been examined in Thailand. The purpose of this study was to identify the key attributes of the drinker prototype among Thai adolescents. Forty-three adolescents aged 13–15 were recruited from a public middle school in Ubon Ratchathani Province in Thailand’s northeast region. Participants first individually listed characteristics of adolescent Thai drinkers and then did the same in age/gender matched focus groups. Member checking with two to three participants from each focus group was conducted to validate the lists. Content analysis was used to identify the most relevant attributes of the adolescent drinker prototype. Six attributes were identified: sociable, fighter, talkative, cool, mature, and funny. These attributes were used to modify an existing prototype questionnaire so that favorability of and similarity to the drinker prototype can be examined in Thai adolescents.

ABOUT THE AUTHOR Dr. Patcharee Jaigarun completed this study as part of a doctoral dissertation at the University of Illinois at Chicago College of Nursing in the United States. Her research is focused on identifying modifiable cognitive risk factors for alcohol use in Thai teenagers in order to prevent early alcohol use. Dr. Jaigarun is now on faculty at Boromarajonani College of Nursing, Sanpasitthiprasong, Ubon Ratchathani, Thailand. Dr. Corte was Dr. Jaigarun's faculty advisor and dissertation chairperson at the University of Illinois at Chicago College of Nursing. Dr. Corte's program of research is focused on the drinker identity as a cognitive vulnerability for alcohol use in adolescents. The other coauthors were members of Dr. Jaigarun's PhD dissertation committee. Dr. McCreary has expertise in development and adaptation of cross-cultural measures. Dr. Hughes is an alcohol researcher. Dr. Finnegan is a methods expert. Dr. Thorkildsen is an educational psychologist with expertise in adolescent development.

PUBLIC INTEREST STATEMENT
Prototypes are images of a "typical" person who engages in a specific type of risk behavior, such as alcohol use. The prototype of the teenage drinker is that they are funny, cool, popular, confident, etc. Researchers have found that teenagers who view these traits as positive and teenagers who believe they are similar to these traits are at risk for high levels of alcohol use. The traits of the teenage drinker prototype differ for different age groups and people of different cultures. In Thailand, the traits of the teenage drinker prototype are unknown. The purpose of this study was to identify the key traits of the Thai teenage drinker prototype. We conducted group-based interviews with 43 adolescents aged 13-15 from a school in northeast Thailand. The interviews revealed six key traits of Thai adolescent drinkers: sociable, fighter, talkative, cool, mature, and funny. These traits will be used to modify an existing drinker prototype questionnaire to make it appropriate for use with Thai adolescents.
Perceived favorability of and similarity to the drinker prototype-a vivid image of a [same-age] person who drinks alcohol-is a known risk factor for alcohol use in adolescents (de Leeuw, Blom, & Engels, 2014;Litt, Stock, & Gibbons, 2015;Teunissen et al., 2014) and young adults (Norman, Armitage, & Quigley, 2007). Many studies have shown that adolescents have very clear images of the type of person their age who drinks alcohol (Armenta, Hautala, & Whitbeck, 2015;Gerrard, Gibbons, Houlihan, Stock, & Pomery, 2008). These images are characterized by defining positive attributes, e.g., cool, funny, popular; to ensure lack of response bias, negative attributes are also included, dull, boring.
The attributes that characterize the drinker prototype may differ depending on culture and age. For example, among Dutch adolescents, Spijkerman, Van den Eijnden, Vitale, and Engels (2004) found 22 attributes that characterized the drinker prototype (e.g., cool, looking tough, and being sociable). Among US adolescents, Gibbons and Gerrard (1995) found that 12 attributes characterized the college drinker prototype (smart, confused, popular, immature, cool, self-confident, independent, careless, unattractive, dull, considerate, and self-centered). Gerrard et al.'s (2006) study of African-American adolescents aged 10-12 showed that six of these attributes (popular, selfish, smart, cool, unattractive, and dull) also characterized the preadolescent drinker prototype. Among Native American-Indian adolescents 10-12 years of age, Armenta et al. (2015) found that nine attributes characterized the adolescent drinker prototype (popular, smart, cool, tough, good-looking, mature, dull or boring, independent, and self-confident).
The attributes that have been used to measure the drinker prototype have all been drawn from populations outside of Thailand, and thus, they may not be culturally relevant in Thailand. We don't know whether the attributes used in other parts of the world reflect the Thai adolescent drinker prototype. The purpose of this research is to identify and validate the attributes of the Thai adolescent drinker prototype so that these attributes can be used to culturally adapt and test a drinker prototype measure for use with Thai adolescents.

Setting and participants
This study was conducted at a public middle school randomly selected from 81 public middle schools in Ubon Ratchathani Province in the northeastern region of Thailand (Secondary Educational Service Area Office #29, 2014). Thai adolescents (N = 43) aged 13-15 years who were able to read, write, and speak Thai language were recruited from the school. Of the 43 adolescents who participated, 54% were girls. The average age was 13.93 (SD = .83). Most of the youth (65%) lived in an urban area.

Procedures
The study was approved by The Institutional Review Board of the University of Illinois at Chicago and the Boromarajonani College of Nursing, Sanpasitthiprasong, Ubon Ratchathani, Thailand. After receiving permission from a public middle school to collect data, the researcher met the parents of the adolescents at a school-parent meeting to describe the study.
We also described the study to those adolescents who received parental permission and were interested in participating. In a private room during their self-study hour, the researcher first asked each participant to independently list characteristics of Thai adolescents who drink alcohol on a card ("free list"). Approximately 1-2 weeks later, six age-and gender-matched focus groups were conducted to identify and come to consensus about the attributes of the typical Thai adolescent who drinks alcohol. Before conducting the focus groups, adolescents provided a fictitious name. Each focus group lasted 30-45 min. After completing the focus groups, the investigator gave each adolescent US$3.20 (100 Baht) to provide assistance with their lunch or school supplies. A subgroup from each focus group validated the attributes derived in the focus groups.

Free listing
Free listing participants were asked to "think for a minute about the type of person [girl/boy] your age [13, 14, or 15] who drinks (alcohol)" and to list on a card all the adjectives they could think of that described this type of adolescent.

Focus groups
A total of six focus groups were conducted in a private room during self-study time. The focus group technique can provide greater depth and richness of data, comparing with the nominal group technique bringing together ideas of individual participants and Delphi technique relying on the collective opinion of an expert panel (Morgan, 1996). The focus groups were gender-and agematched to avoid gender-or age-based dominance. For example, we wanted to be sure that both boys and girls could speak openly and honestly without concern about how their responses would be perceived by the "other" gender, and we wanted to eliminate any concern related to youth trying to "impress" opposite sex youth. In addition, because in Thai culture, people respect anyone who is older than them, we were concerned that in a mixed age focus group, younger adolescents may defer to older adolescents, and thus, we would not be able to obtain the perspective of younger adolescents. Finally, we made sure that all youth had an opportunity to share and that no one person dominated the focus group discussion.
Each focus group was audio-taped. Following the semi-structured interview format "Please describe in your own words, your thoughts and attitudes about the type of teenagers who drink alcohol," the adolescents in each focus group were asked to describe the characteristics of someone their age who drinks alcohol. These characteristics were written on flip charts, and the researcher took field notes during the focus groups. All characteristics generated during the free listing that did not appear on the flip charts were discussed and a determination was made about their inclusion or exclusion as a key characteristic. Then, adolescents were asked to arrive at a group consensus about the key defining characteristics of an adolescent who drinks alcohol. Three of the adolescents from each focus group volunteered (total of 18 volunteers) to help confirm the agreement of the information through member checking within 2-4 weeks after the focus group sessions. Member checking-taking specific descriptions or themes back to a subgroup of participants to determine whether they feel that they are accurate-is a common strategy used to enhance the accuracy of the findings in a qualitative study (Creswell, 2014).
The questions used in the focus groups were developed based on a review of the existing literature (Armenta et al., 2015;Gerrard et al., 2008) on the drinker prototype in English. Then, the researcher translated the guidelines using a committee approach (Furukawa, Driessnack, & Colclough, 2014) for the interviews about the attributes of drinker prototypes from English to Thai. Next, a PhD prepared Thai nursing instructor who had been educated in the United States confirmed the accuracy of the Thai translation. Lastly, a Thai teacher who teaches Thai language courses for 13-15 year adolescents confirmed the Thai language attributes as appropriate for the target population. The audio-taped focus group interviews were transcribed into the central-Thai language dialect and the English language. The researcher (first author) listened to the audio-taped interviews and compared them to the Thai dialect transcripts and then reread the Thai dialect transcripts and compared them to English translations to check accuracy. Codes were then applied to the attributes. To confirm the accuracy of the codes, an expert in alcohol-related cognitions (second author) and an expert in instrument development and focus groups (third author) provided feedback on the codes derived from the focus groups. Then, the researcher and a Thai nursing instructor who is experienced with conducting focus groups, transcription, and coding independently coded the transcribed interviews to provide evidence of consistency/ reliability.

Data analysis
Content analysis was used to identify the most relevant drinker prototype attributes identified in the focus groups. As part of the process, the researcher identified the most frequently listed attributes. Frequencies were used to describe the sample characteristics and the attributes of the drinker prototypes. Percent agreement that can be expected by chance and Scott's pi were used to determine inter-coder reliability (Krippendorff, 2004). We used chance percent agreement because the adjectives used to describe the attributes may occur multiple times in the sentences within the transcripts with coders potentially agreeing about the attribute in one instance but not in another. Neuendorf (2002) suggests that Scott's pi and Cohen's Kappa are appropriate to use for two coders with nominal data. Overall, the percent agreement between the researcher and the nursing instructor for the coded transcriptions was 88%, indicating an acceptable level of consistency (Frey, Botan, & Kreps, 2000). Also, Scott's pi and Cohen's Kappa were .85 and .96, respectively, which are acceptable (Neuendorf, 2002).

Drinker prototype attributes listed in individual sessions
All verbatim responses to the free listing are shown in Table 1. Across all youths, the most frequently listed attributes were "fighter," "sociable," "the life of a party," and "funny," respectively.

Focus group discussions
The characteristics generated during the free listing that did not appear on the flip charts during the focus groups were discussed, and a determination was made about their inclusion or exclusion as key defining characteristics of Thai adolescents who drink alcohol. After discussion of the new characteristics in each focus group, the adolescents voted on their inclusion or exclusion as key defining characteristics. The final characteristics that adolescents selected as most representative of Thai adolescent drinkers are shown in Table 2. These attributes were validated by a subset of the participants, two to three members of each focus group, approximately 2-4 weeks after completion of the focus groups. Names associated with the quotes are pseudonyms chosen by the participants.
(1) Sociable One of the attributes was "sociable." Participants emphasized the "sociable" image of adolescents who drink alcohol, which often involved commemorating special events like a birthday, graduation, or "turning point," as a few participants stated: Like when we want to celebrate something with friends, we have alcohol . . . as a way to socialize. (Peggy-M14 in group 3) But sometimes, it is a way to "party". . .. Like on a friend's birthday, on the final day of class, or after the exam day, drinking can be a way to party. Just like today, some will go out for a party. (Nong Thoek-F15 in group 6) (2) Fighter Another attribute identified was "fighter." It should be noted that most votes for this attribute came from boys. Participants talked about the image of adolescents who drink alcohol as characterized by fighting or aggressive behavior.  (3) Talkative Another attribute identified was "talkative." Participants referred to the disinhibiting effects of alcohol, i.e., that adolescents feel free to say whatever comes to their mind, without filters or judgment of what is appropriate to say. Thus, they may talk about things that are usually kept private. As some participants said: When our friends . . . are drinking whiskey, they always tell everything they have hidden . . . (4) Cool Another attribute identified was "cool." Participants emphasized that the image of adolescents who drink alcohol reflects enhanced social status or prestige and that they project "coolness." In some sense, it is cool . . . . Like . . . they sit in a group and drink together. That makes them look cool. (Idea -F14 in group 3) Of concern was that some participants talked about "competitive drinking" or enhancing their social standing through excessive drinking. As one participant said: . . .it might look cool. . . . When they drink whiskey and compete over who can hold their liquor best. . . . meaning who is the last to get drunk. It looks cool. (Nadet-M15 in group 5) (5) Mature Another attribute identified was "mature." The adolescents explained that adolescents who drink alcohol are viewed as "mature" because they are "role playing" adult behaviors. It should be noted that most votes for this attribute came from girls.
They act like an adult. Like, drinking like adults, imitating the way grownups drink, and speaking loudly like adults do when they drink. (Bones-M15 in group 5) This image also involved expression of masculinity. As one participant stated: . . .It seems like when you are drinking you can express your manhood. (Peng-M13 in group 1) (6) Funny The last attribute was "funny." Participants indicated that adolescents who drink alcohol are funny and that drinking made social events more enjoyable, pleasurable, or light-hearted, especially when music or dancing occurred. It should be noted that all votes for this attribute were from girls.
. . .when they drink, it looks very enjoyable, just for the sake of their fun. . . . That's because they also dance with the music in a good mood. (Stitch-F14 in group 3) Alcohol was considered a "social lubricant." One participant said: . . .Most people who drink usually laugh and enjoy their talk. (Moo Noi-F15 in group 6)

Discussion
In this study, through free listing followed by focus groups and member checking to confirm the accuracy of our findings, we found that six attributes characterized the adolescent drinker prototype in Thailand-sociable, fighter, talkative, cool, mature, and funny. In general, these attributes are consistent with prior literature on the attributes of a drinker prototype among North American and European youths. Our data support the idea that adolescents possess a clear image of someone their age who engages in drinking behavior (Gerrard et al., 2008). These attributes can be used to culturally adapt and psychometrically test a drinker prototype measure.
Three of the attributes-sociable, funny, and talkative-are traits on the extraversion dimension of personality (Goldberg, 1993;Roberts & DelVecchio, 2000). Many studies have shown that high extraversion is associated with alcohol use in adolescents (Norman et al., 2007;Spijkerman, Van den Eijnden, Overbeek, & Engels, 2007). Investigators have found that being sociable is a characteristic of British youth aged 11-17 who drink (Davies, Martin, & Foxcroft, 2013), Dutch adolescents age 12-16 who drink (Spijkerman et al., 2007), and Dutch adolescents and young adults aged 18-25 who drink (van Lettow, Vermunt, de Vries, Burdorf, & van Empelen, 2013). This suggests that "sociable" is a rather ubiquitous characteristic that defines drinkers from adolescence into adulthood. Given that membership in peer groups is very important to youth (Brown, 1990), being sociable may be a mechanism for inclusion in social groups. Another extraversion trait-funny-was also identified as characteristic of Dutch drinkers aged 18-25 years of age (van Lettow et al., 2013). Although being talkative was not identified as a defining attribute of adolescent drinkers in the other studies of adolescents referenced above, an exaggerated version (being loud) has been identified as a key defining attribute of adult drinkers (Norman et al., 2007;van Lettow et al., 2013). In our study, Thai drinkers were viewed as talking more than usual and even divulging secrets which likely speaks to the disinhibiting effects of alcohol. Given that in Thai culture, people tend to express their thoughts or feelings less freely than in Western culture (Von Glinow, Shapiro, & Brett, 2004), they may become more talkative, but not loud, as a result of the disinhibiting effect of alcohol.
Thai adolescents' views of adolescent drinkers as "cool" are consistent with several previous studies of adolescents of various ages in many parts of the world. Being "cool" has been identified as a key defining attribute of drinkers in a variety of adolescent samples including US populations overall, aged 10-12 years (Stock et al., 2013) and 10-14 years (Andrews & Peterson, 2006;Dal Cin et al., 2009;Gibbons et al., 2010), and in subpopulations of African-Americans aged 10-12  and Native American Indians aged 10-12 years (Armenta et al., 2015). Cool is also a defining attribute of adolescent drinkers in Dutch adolescents aged 12-16 (Spijkerman et al., 2004), British aged 11-17 (Davies et al., 2013), and Europeans aged 14-19 (Kalebić Maglica, 2011). Lapyai (2008) further suggested that Thai adolescents may try alcohol because it gives the image of high status in society.
In our study, Thai adolescents also identified "mature" as a key defining attribute of adolescent drinkers. This characteristic was also identified by Litt et al. (2015) in a study of US adolescents overall, aged 13-15, and Native American adolescents aged 10-12 (Armenta et al., 2015). It is likely that adolescents consider drinking alcohol to be part of the adult role.
The last attribute-"fighter"-is noteworthy because it is a more negative characteristic than the others. However, this attribute is consistent with findings from other studies. In US adolescents overall (average age 15.6), Chassin, Tetzloff, and Hershey (1985) found "fighting" was one of the images that the adolescents had of adolescent alcohol drinkers. Spijkerman et al. (2007) found "looking tough" to be part of the prototypical image of Dutch adolescent drinkers aged 12-16. Armenta et al. (2015) also found that being "tough" was part of the prototypical image of Native American-Indian adolescent drinkers aged 10-12. In a study of older Dutch adolescents and young adults aged 18-25, van Lettow et al. (2013) found one of the attributes that characterize the drinker prototype was "volatile." This is the first study to identify the attributes of the adolescent drinker prototype in Thailand. We feel confident that the six attributes characterizing the Thai adolescent drinker identified in this study are accurate given our methodology and our adequate inter-coder reliability. The full utility of these attributes will be determined when used to modify and culturally adapt a drinker prototype measure. Our ultimate goal is to develop interventions that target the drinker prototype to prevent early alcohol use in Thai adolescents. The present study to identify the attributes of the drinker prototype in Thai adolescents is the first step in that work. The next step will be to psychometrically test a measure based on these attributes in samples of Thai adolescents. Given previous success of other interventions aimed at modifying perceived favorability of or perceived similarity to the drinker prototype (Davies et al., 2013;, we are optimistic about this work. The findings of our study should be considered in light of limitations. Our nonrepresentative sample of Thai adolescents was drawn from a single public middle school in the northeast region of Thailand. Therefore, the results may not generalize to Thai teens from other types of schools or other regions of Thailand. In addition, the drinker prototype attributes may differ for children and for older adolescents. Although focus groups are an appropriate method for exploring a specific topic (Morgan, 1996), some adolescents may have been too uncomfortable to openly express their ideas, especially about a sensitive topic such as adolescent alcohol use. Further studies are needed to determine whether the findings can be replicated in other age groups and other regions of Thailand.
In sum, we determined that there are six attributes of the Thai adolescent drinker prototype: sociable, fighter, talkative, mature, cool, and funny. We will use these attributes to culturally adapt a measure of the drinker prototype for use in Thailand. This work holds promise given that prototypes are modifiable and may be a viable intervention target. & Theresa Thorkildsen, Cogent Psychology (2018), 5: 1502501.