Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Nov 13.
Published in final edited form as: J Adolesc. 2013 Feb 22;36(3):435–446. doi: 10.1016/j.adolescence.2013.01.006

Adolescents’ Civic Engagement and Alcohol Use: Longitudinal Evidence for Patterns of Engagement and Use in the Adult Lives of a British cohort

Andrea K Finlay 1,1,a, Constance Flanagan 2,b
PMCID: PMC4643693  NIHMSID: NIHMS735375  PMID: 23462198

Abstract

Participation in discretionary activities during adolescence may facilitate the development of social networks that recruit youth into adult civic life or provide risky contexts that promote alcohol problems. Using data from the 1970 British Cohort Study, latent class analysis was used to identify adolescents’ patterns of civic engagement, alcohol use, and other out-of-school activities at age 16, and test longitudinal links with adult civic engagement and alcohol use at ages 26, 30, and 34. Three classes were identified for both genders. The latent class characterized by involvement in more activities was more likely to be civically engaged in adulthood. The class characterized by the most alcohol use in adolescence had the highest likelihood of adult alcohol use and problems. Results are discussed in light of the health risks associated with each latent class and potential interventions that could be tailored to adolescents based on their patterns of activities.

Keywords: adolescence, discretionary activities, civic engagement, alcohol use, latent class analysis, British Cohort Study

The framework of healthy youth development posits that youth are resources to be developed through opportunities, relationships, experiences, and support that aid in their growth as they become competent, functioning adults (Bernat & Resnick, 2006). Discretionary activities, defined here as activities outside of academics or work, are one facet of youth development that provide opportunities for growth in youths’ lives by allowing adolescents to explore their values and worldviews as well as gain and practice civic skills and build social connections with broader society (Fredricks & Eccles, 2006; Larson, 2001). However, adolescents may also engage in activities during their free time, such as alcohol use, that can be risky for immediate and long-term health (e.g., Brown et al., 2008). These protective and risky activities play into healthy youth development by increasing or decreasing the likelihood of positive and negative outcomes. For example, different patterns of activity involvement are related to youth thriving (e.g., Linver, Roth, & Brooks-Gunn, 2009; Metzger, Crean, & Forbes-Jones, 2009), but alcohol use, despite its prevalence among high school students, is considered as a separate outcome from these identified patterns. We extend previous research by examining a variety of potentially protective or risky discretionary activities and include alcohol use as an activity itself, one that adolescents consider a viable use for their free time with youth potentially meeting up for the purpose of drinking.

Understanding how adolescents use their discretionary time is important insofar as many choices made during this period impact long-term pathways (Feinstein, Bynner, & Duckworth, 2006; Hardy, Pratt, Pancer, Olsen, & Lawford, 2011; Salmela-Aro & Schoon, 2009; Schoon & Cheng, 2011), which may have consequences for individuals as well as for society. Adult heavy drinkers vote less often and are less connected to their communities (Denny & Doyle, 2007; Lindstrom, 2005), patterns of behavior that may be apparent before adulthood. Yet, the long-term associations between adolescent patterns of activities and adult behaviors and the mechanisms explaining them are understudied. Youth involved in a variety of activities may be more likely to develop social networks that carry into adulthood. Alcohol use may be one of the many ways that they use their free time and we know that such use in adolescence is a risk factor for problems with alcohol in adulthood. On the other hand, an absence of activities or social networks in adolescence can lead to social disconnection in adulthood. Such disconnection may or may not be associated with alcohol problems. The current study seeks to elucidate links between patterns of adolescent activities and adult social connection through civic engagement and adult alcohol use and problems.

We examined the various ways British adolescents allocate their discretionary time (in civic activities, alcohol use, and other out-of-school activities) that are commonly available in adolescence. We then tested whether the ways in which adolescents combined these allocations of their free time predicted adult outcomes of civic engagement and alcohol use. Identifying precursory patterns in adolescence is important insofar as both civic engagement and alcohol use are associated with individual and societal well-being (Banaji, 2008; Miller, Levy, Spicer, & Taylor, 2006; Musick, Herzog, & House, 1999; NIAAA, 2003).

Activity Experiences During Adolescence

Adolescence is a crucial time period when youth build skills and connect with peers and mentors who can shape pathways to and outcomes in adulthood across a variety of cognitive, emotional, social, and civic domains. Civic engagement can serve as a protective experience that cultivates social networks that continue to recruit young people into civic life in adulthood (Flanagan, 2004; Kay & Bradbury, 2009; Read, 2010). For example, youth in the U.S. and Canada who volunteered or were involved in clubs while in secondary school had higher rates of political and community participation as adults, including higher rates of valuing of one’s community, social responsibility, and political and community participation (Gallant, Smale, & Arai, 2010; Jennings & Stoker, 2004; McFarland & Thomas, 2006; Metz & Youniss, 2005).

Other discretionary activities can also contribute to healthy youth development, though some negative associations have been observed. Adolescent participants in sports, volunteering, religious services, or school-related activities (e.g., school government, pep club) completed more years of education six years after high school, had lower levels of social isolation, and had higher self-esteem as compared to nonparticipants, but sports participants also had higher rates of marijuana use (Barber, Eccles, & Stone, 2001; Eccles, Barber, Stone, & Hunt, 2003). Some activities, like alcohol use, are generally associated with negative outcomes such as academic problems, drunk driving accidents, and suicide ideation and attempts (Centers for Disease Control, 2003; NIAAA, 2003; Windle, Miller-Tutzauer, & Domenico, 1992). At the same time, alcohol use can be a means whereby college students explore identity (Dworkin, 2005) and experimentation with alcohol is common among adolescents in the U.S. with 70% reporting that they tried alcohol in the past year and 40% in the past 30 days (Johnston et al., 2012). In the current study, we examine activities that are common among adolescents and have been linked with both positive and negative adult outcomes. Insofar as moderate versus heavy alcohol use may be associated with different outcomes, we also include measures of monthly alcohol use, which can be considered more normative use, and heavy drinking, which can be more risky and potentially have more immediate and long-term negative consequences.

Combining Person-Centered and Variable-Centered Analyses

Prior research has clustered adolescents based on civic engagement or out-of-school activities and examined short-term personal and social outcomes. For example, Canadian adolescents involved in their communities and in politics in multiple ways (i.e., activists, helpers, responders) had higher rates of identity achievement, self-esteem, and optimism than the uninvolved (Pancer, Pratt, Hunsberger, & Alisat, 2007). U.S. adolescents who were highly involved in volunteer and other out-of-school activities (e.g., after-school clubs, arts and hobbies) or dedicated their time to a single activity (sports, religious groups, school groups) had higher scores on measures of healthy youth development (e.g., social well-being, school connectedness) and contribution to families and communities (e.g., leadership, helping) than the uninvolved profile; however, the highly involved were also more likely to drink than the uninvolved or single activity profiles (Linver et al., 2009; Zarrett, Fay, Li, Phelps, & Lerner, 2009). A study investigating engagement in leisure activities among South African adolescents (Caldwell et al., 2004) found that alcohol and tobacco use was higher among youth who were engaged in activities but were unmotivated and bored as compared to the youth who did not participate in activities or who participated in activities and were motivated and interested (Tibbits, 2009). Certain activity patterns were consistently identified across these studies: an uninvolved group who participated very little in activities, a sports-only group (when sports were included among the measured activities), and a highly involved group who had high rates of participation in many different activities. In addition, with only one exception (which examined motivations in addition to behaviors; Tibbits, 2009), more highly involved profiles tended to have better personal and social outcomes but also to have more substance use (Linver et al., 2009; Pancer et al., 2007; Zarrett et al., 2009).

The Current Study

The current study extended prior work by explicitly focusing on adolescents’ simultaneous participation in a multitude of discretionary activities. By examining patterns of discretionary activities rather than single activities, we may see a richer picture of the ways that adolescents spend their free time. Using such patterns to predict adult behaviors should deepen our understanding of the developmental trajectories associated with adolescents’ choices.

Hypotheses

There were two aims for the current study. First, drawing from a national longitudinal sample of British adolescents, we described patterns in which adolescents allocate their free time by combining civic engagement, alcohol use, and other out-of-school activities. Based on prior research, we hypothesized that three classes would be identified – uninvolved, sports-only, and highly-involved. Second, we tested whether the activity patterns longitudinally predicted adult civic engagement and alcohol use and problems. We hypothesized that youth classified as highly-involved would have higher rates of civic engagement as well as higher rates of alcohol use in adulthood compared to youth classified as uninvolved or involved only in sports. We expected that, once involved in civic affairs as an adolescent, they would maintain that community involvement into adulthood. Based on other work showing that alcohol use is common among adolescents who are involved in many different activities, we reasoned that the availability of alcohol at social events would make its use a norm that the highly involved group would carry into adulthood. We hypothesized that the sports-only group would have higher rates of alcohol problems in adulthood compared to the highly involved due to the link between sports and drinking (e.g., Eccles & Barber, 1999), but we have no a priori hypotheses about their adult civic engagement. Those classified as uninvolved were hypothesized to have low rates of adult civic engagement, but we had no a priori hypotheses about their alcohol use and problems. They may be less likely to drink from lack of opportunity or exposure or more likely to have alcohol problems if their low involvement is a sign of maladaptive behavior. Figure 1 displays the hypothesized adolescent patterns of discretionary activity and the directional associations with adult civic engagement and alcohol use.

Figure 1.

Figure 1

Open in a new tab

Conceptual model of adolescents’ patterns of discretionary activities predicting adult civic engagement and alcohol use.

Method

Participants

The 1970 British Cohort Study (BCS70) is an ongoing national longitudinal study following a cohort of people from their births to the present (98% of those born during one week in 1970, N = 17 287; Chamberlain & Chamberlain, 2008; Ferri, Bynner, & Wadsworth, 2003; Plewis, Calderwood, Hawkes, & Nathan, 2004). Follow-up data were collected from respondents from England, Scotland and Wales at ages 5, 10, 16, 26, 30, 34, and 38 (Butler & Bynner, 2008a, 2008b; Butler, Dowling, & Osborn, 2008; Bynner, 2008; University of London, 2008a, 2008b, 2008c). The current analytic sample is limited to participants who provided data for at least one wave at ages 16, 26, 30, or 34. A total of 7 002 females (50%) and 7 023 males were included.

Measures

Civic engagement, age 16

To assess civic engagement, adolescents were asked how often they did the following in their spare time: (a) attend a meeting or political club, and (b) volunteer or community work (Feinstein, Bynner, & Duckworth, 2005), with response options coded as 1 = no or 2 = yes for each item.

Alcohol use, age 16

Monthly alcohol use measured adolescents’ frequency of alcohol use in the past year, coded as 1 = no (never or only on special occasions) and 2 = yes (once a month or more). Heavy drinking measured whether the adolescent had consumed 4 or more drinks in a row in the previous two weeks: 1 = no, 2 = yes.

Other out-of-school activities, age 16

Other out-of-school activities assessed how often adolescents did various activities in their spare time (Sacker & Cable, 2005), grouped into categories by activity content. Socializing (e.g., Have friends around to my house; Go to a friend’s house), sports (e.g., Play sports; Go to dance classes), and hobbies and crafts (e.g., Draw or paint for fun; Play a musical instrument) were coded as 1 = low participation (once a week or less) and 2 = high participation (more than once a week). Religious services attendance was measured by one item, coded to 1 = low attendance (once or twice a month or less) and 2 = high attendance (once a week or more).

Civic engagement, adult waves

At ages 26, 30, and 34, participants’ political interest was assessed (Deary, Batty, & Gale, 2008; Denny & Doyle, 2008), coded as 0 = low political interest (little or no interest) and 1 = high political interest (moderate to high interest). Voting in the 1997 and 2001 general elections was assessed at ages 30 (in 2000) and 34 (in 2004), respectively; coded 0 = no and 1 = yes (Deary et al., 2008; Denny & Doyle, 2007).

At age 30, organizational membership reflected if a person was ever part of any organization such as political parties, environmental charities, and voluntary groups (Denny & Doyle, 2008; Feinstein et al., 2006; Paterson, 2008). Union membership reflected current membership in a Trade Union or Staff Association (Arulampalam & Booth, 2000; Paterson, 2008). Both were coded as 0 = no and 1 = yes.

At age 34, civic action was measured by four items asking whether participants had ever (a) attended a public meeting or rally, (b) taken part in a public demonstration or protest, (c) signed a petition, or (d) contacted a government official (Deary et al., 2008), coded as 0 = no (no to all four items) and 1 = yes (yes to any of the four items). Social trust measured how much participants trusted people in their local area, coded 0 = low social trust (little or no trust) and 1 = high social trust (moderate to high trust).

Alcohol use, adult waves

At ages 26, 30, and 34, participants reported on the number of units of alcohol they had consumed in the past week. At ages 26 and 34, participants were measured on their monthly alcohol use, that is, how often they had an alcohol drink of any kind, coded as 0 = no (never or only on special occasions) and 1 = yes (once a month or more). At age 34, problem drinking was assessed using the CAGE (Mayfield, McLeod, & Hall, 1974; Smart, Adlaf, & Knoke, 1991) for lifetime and annual problems due to alcohol use (e.g., ever felt bad or guilty about your drinking). Although it does not provide a clinical diagnosis, the CAGE is a widely used screening tool for alcohol dependence (Bisson, Nadeau, & Demers, 1999). Participants who answered yes to 1 or more of 4 CAGE questions were coded 1 (Problem drinking) with all other participants coded as 0 (No problem drinking).

Plan of Analysis

Multiple imputation using the EM algorithm (Graham, 2009) was conducted prior to analyses because of missing data on distal outcomes (Tan, Lanza, & Wagner, 2011), which would otherwise exclude cases from the dataset. Data were assumed to be missing at random and multiple imputation was conducted using SAS PROC MI. Multiple imputation was used rather than complete case analysis (i.e., listwise deletion) to (a) preserve the population parameters and distributions of the whole dataset and (b) prevent the loss of statistical power (Graham, Cumsille, & Elek-Fisk, 2003). Analyses were conducted on 40 imputed datasets and results were averaged across datasets. Due to gender differences in volunteering, sports, and alcohol use (Caspersen, Pereira, & Curran, 2000; Johnston, O’Malley, Bachman, & Schulenberg, 2012; Lopez et al., 2006), latent class and distal outcome analyses were conducted separately by gender.

Latent Class Analysis (LCA, Collins & Lanza, 2010), a technique that describes differences between classes of people within a population based on individual characteristics, was utilized to examine adolescents’ patterns of discretionary activities. LCA models with two, three, four, and five latent classes were compared for each gender. The best fitting model was determined through examination of a variety of fit statistics and judgments about parsimony (a smaller number of classes), interpretability (the item-response probabilities clearly distinguished one group from another), and stability (the number of models that converged on the same solution using 100 random starting values with at least 80% of models converging). Better fits were indicated by a lower likelihood-ratio G2 statistic, Akaike’s information criterion (AIC, Akaike, 1974), and Bayesian information criterion (BIC, Schwarz, 1978). When these indicators were contraindicated, the model with the smallest BIC value was selected. Although not recommended as a tool for model selection, we also examined entropy which summarizes the certainty of classification across individuals. Values closer to one indicate higher certainty and values closer to zero indicate lower certainty (Collins & Lanza, 2010). Models were estimated using SAS PROC LCA.

Two sets of parameters were estimated from the best fitting models and were used to describe the patterns of discretionary activities identified: class membership probabilities and item-response probabilities (Collins & Lanza, 2010). Class membership probability is the probability that a person is a member of a specific discrete class with a higher likelihood indicating that she or he is more likely to be in that class. Participants were classified into the group in which they had the highest probability of being a member. Item-response probability is the probability of endorsing a specific item taking class membership into account. Item-response probabilities range from 0 to 1, with higher than .5 representing a greater likelihood, and lower than .5 representing a lower likelihood of responding affirmatively to the item. Contrasts between the groups based on their item-response probabilities (e.g., did one group have a stronger likelihood of reporting a specific activity versus another group that was unlikely to report that activity) were used to name the groups.

Distal outcomes (i.e., civic engagement and alcohol use and problems in adulthood) were predicted by including each outcome measure as a covariate in the LCA models and taking uncertainty related to class membership into account (Lanza, Tan, & Bray, 2011). This technique has been shown to produce less biased estimates than techniques in which participants are assigned to a latent class with outcome analyses conducted in a separate step. SAS PROC LCA and a corresponding macro were utilized in analyses and distal outcomes were included as covariates in the LCA models using logistic regression for binary outcomes and ordinary least squares regression for continuous outcomes. Known marginal distributions were used to calculate the probabilities for each binary outcome and the mean for each continuous outcome.

Results

Descriptive Statistics

Descriptive statistics of adolescent civic engagement, alcohol use, and other out-of-school activities are displayed in Table 1. Political club involvement was the least common activity with engagement by less than one-tenth of females and males. Volunteering was more common but still low with over one-fifth of females and almost one-fifth of males participating. Two-thirds of females reported monthly alcohol use and almost a third engaged in heavy drinking. Among males, almost three-fourths used alcohol monthly and over one-third reported heavy drinking. More than two-thirds of females and males engaged in socializing, and two-fifths of females and two-thirds of males participated in sports. Adolescent females and males were significantly different in their endorsement of all activities except for socializing (t values ranged from ±2.41 to 16.39, p < .05).

Table 1.

Descriptive Statistics of Civic Engagement, Alcohol Use, and Other Out-of-School Activities at Age 16 Among British Females and Males

Activity Label Females (n = 7 002) Males (n = 7 023)
Frequency (%)a Frequency (%)a
Civic engagement
 Political club Yes 436 (6.2) 563 (8.0)
 Volunteer Yes 1592 (22.7) 1330 (18.9)
Alcohol use
 Monthly alcohol use Yes 4603 (65.7) 5123 (73.0)
 Heavy drinking Yes 2226 (31.8) 2580 (36.7)
Other out-of-school
 Socializing High 5034 (71.9) 4834 (68.8)
 Sports High 3022 (43.2) 4592 (65.4)
 Hobbies High 2673 (38.1) 2305 (32.8)
 Religious services High 2247 (32.1) 1823 (26.0)
Open in a new tab

Note. Participants were surveyed in 1986 at age 16.

a

Number (%) of participants who indicated yes or high involvement for each activity.

Adult civic engagement and alcohol use measures are shown in Table 2. Political interest was moderately endorsed by about one-quarter of females and two-fifth of males at ages 26 and 30 and even higher rates of endorsement were observed at age 34. Almost two-thirds of females and males endorsed voting at ages 30 and 34. Social trust at age 34 was the most highly endorsed civic indicator with almost three-fourths of females and males reporting high social trust. Organizational and union memberships were least common with less than one-fourth of people participating at age 30. Alcohol use was common in adulthood with more than three-fourths of females and males reporting monthly alcohol use at ages 26 and 34. Adult females and males were significantly different in their endorsement of all outcomes except for voting at age 34 (t values ranged from ±3.91 to 43.52, p < .001).

Table 2.

Descriptive Statistics of Civic Engagement and Alcohol Use Outcomes at Ages 26, 30, and 34 Among British Females and Males

Activity Label Females (n = 7 002) Males (n = 7 023)
Frequency (%)a Frequency (%)a
Age 26 (1996)
 Political interest High 1885 (26.9) 2875 (40.9)
 Monthly alcohol use Yes 5782 (82.6) 6367 (90.7)
 Alcohol units M (SD) 5.8 (9.7) 17.6 (16.0)
Age 30 (2000)
 Political interest High 1891 (27.0) 2929 (41.7)
 Voting Yes 4481 (64.0) 4137 (58.9)
 Organizational membership Yes 1505 (21.5) 1108 (15.8)
 Union membership Yes 1176 (16.8) 1393 (19.8)
 Alcohol units M (SD) 7.7 (15.3) 26.0 (26.4)
Age 34 (2004)
 Political interest High 2351 (33.6) 3455 (49.2)
 Voting Yes 4339 (62.0) 4214 (60.0)
 Civic action Yes 2008 (28.7) 1649 (23.5)
 Social trust High 5099 (72.8) 4850 (69.1)
 Monthly alcohol use Yes 5278 (75.4) 6163 (87.8)
 Lifetime CAGE Yes 2164 (30.9) 3296 (46.9)
 Annual CAGE Yes 1572 (22.5) 2454 (34.9)
 Alcohol units M (SD) 5.9 (9.8) 15.8 (15.8)
Open in a new tab
a

Number (%) of participants who indicated yes or high involvement in each activity.

Model Selection for Latent Class Analysis

The LCA models for each number of classes, averaged across the 40 imputed datasets, are shown in Table 3. The three-class model (in bold) was selected as the best fitting model for both genders because it had the smallest BIC value and was more parsimonious than the four-class model. Incremental decreases in the AIC and BIC up to the three-class model suggested improvement in fit, whereas the BIC increased between the three- and four-class solutions. Moreover, the three-class model was clearly distinguishable and interpretable. The entropy values indicated that the certainty in classification was similar across different numbers of latent classes. Thus, we relied on the fit statistics, which combined with the interpretability and the stability across the 100 random starting values, provided consistent evidence that the three-class model represented the best overall fit. The models were fit without covariates to accommodate each distal outcome but additional tests with covariates of family social class, family structure, academic test performance, and delinquency indicated that the three-class solution is the best model fit for both females and males (results not shown).

Table 3.

Model Comparisons of Latent Class Analysis at Age 16 Among British Females and Males

Number of classes G2 df AIC BIC Entropy
Females
2 422 238 456 573 .37
3 310 229 361 540 .36
4 258 220 328 568 .39
5 Did not converge
Males
2 410 238 444 558 .31
3 307 229 359 537 .31
4 241 220 308 539 .35
5 Did not converge
Open in a new tab

Note. Boldface font indicates the selected model. G2 = likelihood ratio statistic;AIC = Akaike’s Information Criterion; BIC = Bayesian Information Criterion.

Latent Classes for Females and Males

The latent classes of females and males correspond to groups of people characterized by their patterns of discretionary activities. The affirmative item-response probabilities for each class, displayed in Table 4 and plotted in Figures 2 and 3, provide information on how to interpret and label each class. The first class, for both females and males, was labeled Inactive because they are characterized by a low probability of engagement in all activities, aside from socializing and sports, and they had the lowest rates of socializing and sports relative to the other classes. The second latent class for females was comprised of those who were likely to report monthly alcohol use (.93), heavy drinking (.60), and socializing (.81), but unlikely to report the civic engagement or remaining out-of-school activities. This class was labeled Drinking-Social-Active. The second latent class of males was also comprised of people who were likely to report monthly alcohol use (.94), heavy drinking (.61), and socializing (.79), as well as sports (.61), but unlikely to report the civic engagement or remaining out-of-school activities. This class was labeled Drinking-Social-Sports-Active. For both females and males, the third latent class was labeled Multi-Active and was characterized by the highest probability of reporting engagement in the civic activities, sports, hobbies, and religious services relative to the other classes as well as high probabilities of reporting monthly alcohol use and socializing.

Table 4.

Prevalence of Latent Classes and Item-Response Probabilities at age 16 Among British Females and Males

Latent classes for females
Latent classes for males
Inactive Drinking-Social- Active Multi- Active Inactive Drinking-Social- Sports-Active Multi- Active
Prevalence of latent class .37 .40 .23 .40 .36 .22
Item-response probabilities
 Political club .03 .07 .14 .04 .05 .19
 Volunteer .15 .19 .44 .12 .12 .47
 Monthly alcohol use .47 .93 .53 .53 .94 .74
 Heavy drinking .13 .60 .18 .14 .61 .37
 Socializing .61 .81 .70 .61 .79 .68
 Sports .41 .44 .59 .59 .61 .61
 Hobbies .30 .34 .57 .30 .31 .46
 Religious services .28 .26 .51 .27 .22 .38
Open in a new tab

Figure 2.

Figure 2

Open in a new tab

Item-response probabilities of discretionary activity involvement at age 16 among British females.

Figure 3.

Figure 3

Open in a new tab

Item-response probabilities of discretionary activity involvement at age 16 among British males.

The class membership probabilities, which show the distribution of people across the classes, are shown in Table 4. Thirty-seven percent of females and 40% of males were in the Inactive class. Among females, 40% were in the Drinking-Social Active class and among males, 36% were in the Drinking-Social-Sports Active class. The remaining 23% of females and 22% of males were classified in the Multi-Active class.

Predicting Civic Engagement and Alcohol Use in Adulthood

Sixteen models each for females and males were tested to calculate the estimated probabilities of participants engaging in each adult civic engagement and alcohol use outcome as a function of their adolescent discretionary activity patterns (Table 5). Beta coefficient estimates from the LCA models were used to calculate the estimated probabilities of engaging in each behavior for the binary outcomes and the estimated mean for each continuous outcome. Overall differences in classes were indicated by the change in twice the log-likelihood value.

Table 5.

Probabilities of Civic Engagement and Alcohol Use Outcomes Among British 26-, 30-, and 34-Year-Old Females and Males

Probabilities Females
Males
Inactive Drinking-Social Active Multi- Active Δ 2* LLa Inactive Sports-Drinking- Social Active Multi- Active Δ 2* LLa
Age 26 (1996)
 Political interest .25 .25 .41 20.60** .38 .37 .51 15.78*
 Monthly alcohol use .76 .89 .83 70.24*** .83 .95 .91 76.22***
 Alcohol unitsb 4.08 8.68 5.89 112.95*** 11.23 20.52 15.57 205.71***
Age 30 (2000)
 Political interest .25 .25 .40 19.08* .37 .39 .52 17.64*
 Voting .65 .59 .72 30.14*** .60 .55 .65 9.52
 Org. membership .18 .18 .37 40.85*** .13 .13 .28 32.86***
 Union membership .16 .17 .19 2.76 .18 .20 .21 5.13
 Alcohol unitsb 5.87 11.88 8.76 67.31*** 18.13 28.58 24.71 123.31***
Age 34 (2004)
 Political interest .33 .32 .46 18.81** .46 .46 .57 12.00
 Voting .60 .59 .73 19.90** .61 .58 .65 6.01
 Civic action .25 .28 .38 14.14* .23 .24 .26 3.43
 Social trust .71 .72 .78 7.51 .68 .71 .73 5.07
 Monthly alcohol use .69 .81 .76 58.52*** .81 .93 .88 78.07***
 Alcohol unitsb 4.62 8.28 5.04 85.61*** 9.68 18.22 14.09 151.05***
 Lifetime CAGE .24 .38 .33 49.43*** .39 .54 .48 50.14***
 Annual CAGE .18 .29 .20 48.65*** .31 .42 .28 49.16***
Open in a new tab
a

Δ 2*LL = Change in twice the log-likelihood.

b

Mean number of alcohol units consumed.

*

p < .05,

**

p < .01,

***

p < .001.

Females from the Multi-Active class had higher probabilities of endorsing political interest, being organizational members, voting, and engaging in civic action compared to their Drinking-Social Active and Inactive peers. The Inactive and Multi-Active classes consumed fewer units of alcohol and had lower probabilities of alcohol problems in adulthood than the Drinking-Social Active class. The Inactive class had lower probabilities of engaging in monthly alcohol use at ages 26 and 34 compared to the Drinking-Social-Active and Multi-Active classes.

Males from the Multi-Active class had higher probabilities of endorsing political interest at ages 26 and 30 and belonging to organizations at age 30 than the Inactive and Drinking- Social-Sports-Active classes. The Inactive and Multi-Active classes consumed fewer units of alcohol, and had lower probabilities of alcohol problems in adulthood than the Drinking-Social-Sports-Active class. The Inactive class had lower probabilities of engaging in monthly alcohol use at ages 26 and 34 compared to the Drinking-Social-Sports-Active class.

Discussion

Latent Classes of Discretionary Activity Involvement

Three classes of adolescents were found for both females and males: the Inactive classes who were characterized by a low probability of engagement in all activities, aside from socializing and sports and who had the lowest rates of socializing and sports relative to the other classes; the Drinking-Social-Active (females) and Drinking-Social-Sports-Active (males) classes who were characterized by a high likelihood of involvement in alcohol use, socializing, and, in the case of boys, sports, but who had low involvement in the other activities; and the Multi-Active classes who had the highest probabilities of engaging in civic activities, sports, hobbies, and religious services relative to the other classes as well as high probabilities of reporting monthly alcohol use and socializing. The Inactive and Multi-Active classes identified in the current study are similar to those identified in previous person-centered work (Linver et al., 2009; Pancer et al., 2007; Zarrett et al., 2009). The Inactive classes were relatively uninvolved in discretionary activities and Multi-Active classes engaged in a wide range of activities, though the levels varied between females and males. Despite previous research on the protective benefits of civic engagement in predicting alcohol use (e.g., Barber et al., 2001), there was no group that reported both high involvement in civic engagement and no involvement in alcohol use. Perhaps our study did not find a high civic-low drinkers group because our data utilized British participants and previous studies were conducted in the U.S. where the drinking age is older than that in Britain.

Gender Differences

We did not directly test gender differences because previous work and preliminary analyses indicated differences between females and males on all discretionary activities except socializing. There were notable differences, however, in the activity patterns. Consistent with other work, females were more involved than their male peers in volunteer work (Feinstein et al., 2005). However, our study also revealed that there also was a group of males who engaged in volunteer work along with other activities, a finding that could only be uncovered by person-centered analyses examining patterns of activities rather than levels of single activities. Perhaps boys who volunteered were more likely to be involved in activities in general, which would be consistent with a genetic basis for involvement (Fowler, Baker, & Dawes, 2008). Girls were less active in sports and less likely to use alcohol than boys and these activities figured less prominently into their classes. Girls may have had fewer opportunities to engage in these pursuits or may have attached less value to sports or to alcohol use. Despite the differences between females and males, the pattern of latent classes predicting adult outcomes was fairly similar between the two genders.

Predicting Civic Engagement and Alcohol Use in Adulthood

Longitudinal analyses predicting distal outcomes showed that adolescents who were Multi-Active, compared to the other classes, had higher probabilities of being civically engaged as adults, including joining organizations, voting and taking other civic actions, and being interested in political issues. These findings have added a more diverse array of civic outcomes to earlier studies and confirmed that pathways to adult civic engagement begin with civic experiences in adolescence (Feinstein et al., 2006; Hanks & Eckland, 1978; Jennings & Stoker, 2004). Notably, only the Multi-Active classes had exhibited even moderate levels of civic engagement in adolescence. This is important in that it suggests that it is not any discretionary activities during adolescence (e.g., sports, socializing) but rather specific experiences in the civic domain that lay the foundation for long-term civic patterns. Perhaps the social networks established in these early civic experiences increased the likelihood of recruitment into later civic life (Verba, Schlozman, & Brady, 1995) or perhaps youth learn civic skills and competencies that carry into adulthood (Flanagan, 2003). Selection effects, for example due to sociability or enduring genetic differences between people (Son & Wilson, 2010), may account for some observed links between adolescent free time activity and subsequent civic involvement, but quasi-experimental studies of mandated volunteering also show that an ethic of service is gained from involvement, even among initially less motivated youth (Metz & Youniss, 2005). Further investigation of the underlying mechanisms is needed, especially in light of the lower mortality and depression rates and better cognitive functioning associated with adult civic engagement (Harris & Thoresen, 2005; Morrow-Howell, Hinterlong, Rozario, & Tang, 2003; Musick et al., 1999).

In addition to the health benefits of civic engagement, those in the Multi-Active classes consumed fewer alcohol units, and had lower probabilities of alcohol problems than their Drinking-Social-Active and Drinking-Social-Sports-Active counterparts. Multi-Active people may consume alcohol in social contexts where moderate levels of consumption are the norm. This behavioral pattern may have been established earlier in life: Even in adolescence, the Multi-Active group had lower rates of monthly alcohol use and heavy drinking. The protective role of civic engagement in relation to alcohol use has been documented in previous research (e.g., Barber et al., 2001). The current study suggests that for highly involved youth the protective aspect of civic engagement comes from moderation of alcohol use, rather than abstinence. Adolescents who are engaged in a variety of activities may be more experimental in their approach; thus, moderate alcohol use may fit with their general behavioral pattern. Further reducing adolescent drinking among these youth may be effectively communicated through harm reduction strategies that focus on what impact their drinking may have on the community or on their involvement in other activities.

The Drinking-Social-Active and Drinking-Social-Sports-Active classes may be at highest risk for negative health and well-being in adulthood given their higher probabilities of alcohol problems and lower probabilities of civic engagement. As these classes had almost universal engagement in monthly alcohol use and nearly two-thirds engaged in heavy drinking in adolescence, our results suggest that a lack or a perception of a lack of other activity opportunities or encouragement to explore various activities may already exist at age 16. Further work towards understanding why these adolescents are not involved in many other activities would inform the development of intervention resources to reduce risky heavy drinking among these youth. Furthermore, the socializing and sports these youth were involved in did not translate into active community engagement in adulthood as measured by our study. Social networks that carry into other adult behaviors would be another important area of study to provide a clearer picture of this potentially at-risk group. Although altering socializing may be difficult because it is an unstructured activity, sports activities that include an emphasis on community or giving adolescents more autonomy and responsibility may help develop civic skills that will benefit these youth as well as their society. Countering the heavy drinking culture that often accompanies sports is another potential avenue for reducing alcohol problems. This strategy would likely reduce the substantial costs associated with such behavior. A dual-pronged approach of reducing risky behaviors and promoting protective factors, which is considered the most effective strategy (Bernat & Resnick, 2006), may help correct some of the long-term negative health outcomes observed among the Drinking-Social-Active and Drinking-Social-Sports-Active groups.

Despite the lower probabilities of alcohol use and alcohol problems among the Inactive classes, these adolescents are still at risk for poor health in adulthood given their limited civic engagement. Low civic engagement in adolescence and a lower probability of engagement in adulthood suggests that these classes are on a trajectory for limited civic participation through their lives. Given the importance of adult civic involvement for better cognitive function and lower mortality in adulthood (Harris & Thoresen, 2005; Morrow-Howell et al., 2003; Musick et al., 1999), people with low civic trajectories may be at risk for negative health into their later adult years. In addition, the health of a society may be at risk without the participation of its citizens (Banaji, 2008; Smith, 1999). Future work uncovering why these people limit their engagement in both adolescence and adulthood as well as ways to create activities that attract less engaged youth and promote their long-term civic development is essential for individual and societal well-being.

Historical Context

Although our study can generalize somewhat to contemporary adolescents given the stability in activities offered in schools, there are a few notable historical changes. Civic engagement rates in Britain have declined over time. The 1983 voting rate in Britain for 18-to-34-year-olds was 76% compared to recent general elections of 55% (Office for National Statistics, 2011). Furthermore, declining rates of membership in voluntary and community organizations in recent decades have been noted (Bynner, 2005). However, recent volunteer rates in Britain are higher than in previous decades. Compared to 23.8% of females and 18.5% of males observed in our sample, a 2005 sample revealed that 63% of 16- to 19-year-olds volunteered at least once in the past year (Rochester, 2006). Recent policy in Britain has focused on increasing youth volunteering including service learning in secondary schools and the creation of volunteer programs (e.g., Millennium Volunteers; Stanley, 2004). Similarly, volunteering rates increased throughout the 1990s and 2000s in the U.S. (Lopez et al., 2006), which has been attributed to the introduction of mandatory volunteer service requirements for high school graduation in some cities and states and the importance of volunteer work for higher education applications (Andolina et al., 2002; Marcelo, 2007). While concerns that there would be negative effects of mandatory service have not born out (Henderson et al., 2007) and positive outcomes such as increased intentions to be civically engaged and higher rates of voting in young adulthood have been observed (Hart et al., 2007; Metz & Youniss, 2005), more longitudinal work is needed before speculating on what implications mandatory service might have for other forms of adult civic engagement.

In addition to increased volunteering, rates of heavy drinking have increased over time. In 1986, 32% of girls and 37% of boys in our sample participated in heavy drinking, whereas 55% of girls and 52% of boys in Britain reported an episode of heavy drinking in 2007 (Hibell et al., 2009). Monthly alcohol use appears to have remained stable over time, though the measures are somewhat different. Despite these historical changes, other research using cohorts from 1986 to contemporary time indicate that patterns of adolescent discretionary activities remain stable over time (Finlay & Flanagan, 2012).

Limitations and Future Directions

As with any study there were limitations to the current analyses. First, although we cast a wide net to capture diverse discretionary activities, many other discretionary activities that we were unable to include, such as academic clubs and performing arts, may be important. In addition, activities may take place during school time as well as during non-school time (e.g., a school-based service learning project versus afterschool community volunteering) and future work teasing apart whether these diverse settings affect patterns of adolescent activity engagement is warranted. Second, work and work-like activities such as school and paid employment were not studied; these provide similar developmental opportunities such as socializing, developing skills, or building social networks. Understanding the skills adolescents learn in various activity and work settings as well as the transference of these skills from one setting to another would add to our understanding of adolescent development (see Zimmer-Gembeck & Mortimer, 2006 for a review). Third, the motivations behind participating in discretionary activities were not examined. Finally, we were not able to capture all types of adult civic engagement (e.g., volunteering) and some people may have been restricted from certain types of engagement (e.g., not all occupations have unions). Future work examining additional civic attitudes and behaviors may help address these limitations.

Conclusions

The strengths of our study include (a) a focus on diverse discretionary activities that have rarely been examined together or explored through person-centered analyses, (b) the use of a nationally representative sample of 16-year-old British adolescents, and (c) a longitudinal examination of combinations of discretionary activities predicting adult civic engagement and alcohol use into their 30s. Our longitudinal results suggest that the Drinking-Social-Active and Drinking-Social-Sports-Active classes may be at risk for negative health outcomes because of their higher likelihood of alcohol problems and lower likelihood of civic engagement. Underlying processes such as personality differences, the development of different social networks, and diverse adolescent experiences that influence individuals’ developmental trajectories are speculated to explain the results of the current study. Whether these classes of youth differ on other health, work, or relationship outcomes is a potential avenue for future research.

Contributor Information

Andrea K. Finlay, The Pennsylvania State University.

Constance Flanagan, University of Wisconsin-Madison.

References

  1. Akaike H. A new look at the statistical model identification. IEEE Transactions on Automatic Control. 1974;19:716–723. doi:710.1109/TAC.1974.1100705. [Google Scholar]
  2. Andolina MW, Jenkins K, Ketter S, Zukin C. Searching for the meaning of youth civic engagement: Notes from the field. Applied Developmental Science. 2002;6:189–195. [Google Scholar]
  3. Arulampalam W, Booth AL. Union status of young men in Britain: A decade of change. Journal of Applied Econometrics. 2000;15:289–310. doi:210.1002/1099-1255(20005/20006)20015:20003<20289::AID-JAE20549>20003.20000.CO;20002-P. [Google Scholar]
  4. Banaji S. The trouble with civic: A snapshot of young people’s civic and political engagements in twenty-first-century democracies. Journal of Youth Studies. 2008;11:543–560. doi:510.1080/13676260802283008. [Google Scholar]
  5. Barber BL, Eccles JS, Stone MR. Whatever happened to the Jock, the Brain, and the Princess? Young adult pathways linked to adolescent activity involvement and social identity. Journal of Adolescent Research. 2001;16:429–455. doi:410.1177/0743558401165002. [Google Scholar]
  6. Bernat D, Resnick MD. Healthy youth development: Science and strategies. Journal of Public Health Management and Practice. 2006 Nov;(Suppl):S10–S16. doi: 10.1097/00124784-200611001-00004. [DOI] [PubMed] [Google Scholar]
  7. Bisson J, Nadeau L, Demers A. The validity of the CAGE scale to screen for heavy drinking and drinking problems in a general population survey. Addiction. 1999;94:715–722. doi: 10.1046/j.1360-0443.1999.9457159.x. [DOI] [PubMed] [Google Scholar]
  8. Brown SA, McGue M, Maggs JL, Schulenberg JE, Hingson RW, Swartzwelder S, et al. A developmental perspective on alcohol and youths 16 to 20 years of age. Pediatrics. 2008;121(Suppl 4):S290–S310. doi: 10.1542/peds.2007-2243D. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Butler N, Bynner JM. 1970 British Cohort Study: Ten-Year Follow-up, 1980 [computer file] 3. Colchester, Essex: UK Data Archive [distributor]; 2008a. Oct, SN: 3723 http://dx.doi.org/10.5255/UKDA-SN-3723-1. [Google Scholar]
  10. Butler N, Bynner JM. 1970 British Cohort Study: Sixteen-Year Follow- up, 1986 [computer file] 4. Colchester, Essex: UK Data Archive [distributor]; 2008b. Oct, SN: 3535 http://dx.doi.org/10.5255/UKDA-SN-3535-1. [Google Scholar]
  11. Butler N, Dowling S, Osborn A. 1970 British Cohort Study: Five-Year Follow-up, 1975 [computer file] 2. Colchester, Essex: UK Data Archive [distributor]; 2008. Aug, SN: 2699 http://dx.doi.org/10.5255/UKDA-SN-2699-1. [Google Scholar]
  12. Bynner JM. Rethinking the youth phase of the life-course: The case for emerging adulthood? Journal of Youth Studies. 2005;8:367–384. doi:310.1080/13676260500431628. [Google Scholar]
  13. Bynner JM. 1970 British Cohort Study: Twenty-Six-Year Follow-up, 1996 [computer file] 3. Colchester, Essex: UK Data Archive [distributor]; 2008. Aug, August 2008. SN: 3833 http://dx.doi.org/10.5255/UKDA-SN-3833-1. [Google Scholar]
  14. Caldwell LL, Smith EA, Flisher AJ, Wegner L, Vergnani T, Mathews C, et al. HealthWise South Africa: Development of a life skills curriculum for young adults. World Leisure Journal. 2004;3:4–17. doi: 10.1080/04419057.04412004.09674362. [DOI] [Google Scholar]
  15. Caspersen CJ, Pereira MA, Curran KM. Changes in physical activity patterns in the United States, by sex and cross-sectional age. Medicine & Science in Sports & Exercise. 2000;32:1601–1609. doi: 10.1097/00005768-200009000-00013. [DOI] [PubMed] [Google Scholar]
  16. Centers for Disease Control. Drunk driving. 2003 Retrieved February 25, 2008, from http://www.childtrendsdatabank.org/indicators/41DrunkDriving.cfm.
  17. Chamberlain R, Chamberlain G. 1970 British Cohort Study: Birth and 22- Month Subsample, 1970–1972 [computer file] 2. Colchester, Essex: UK Data Archive [distributor]; 2008. Aug, SN: 2666 http://dx.doi.org/10.5255/UKDA-SN-2666-1. [Google Scholar]
  18. Collins LM, Lanza ST. Latent class and latent transition analysis: With applications in the social, behavioral, and health sciences. New York, NY: Wiley; 2010. [DOI] [Google Scholar]
  19. Deary IJ, Batty GD, Gale CR. Childhood intelligence predicts voter turnout, voting preferences, and political involvement in adulthood: The 1970 British Cohort Study. Intelligence. 2008;36:548–555. doi:510.1016/j.intell.2008.1009.1001. [Google Scholar]
  20. Denny KJ, Doyle OM. “…Take up thy bed, and vote” Measuring the relationship between voting behaviour and indicators of health. European Journal of Public Health. 2007;17:400–401. doi: 10.1093/eurpub/ckm002. doi:410.1093/eurpub/ckm1002. [DOI] [PubMed] [Google Scholar]
  21. Denny KJ, Doyle OM. Political interest, cognitive ability and personality: Determinants of voter turnout in Britain. British Journal of Political Science. 2008;38:291–310. doi:210.1017/S000712340800015X. [Google Scholar]
  22. Dworkin J. Risk taking as developmentally appropriate experimentation for college students. Journal of Adolescent Research. 2005;20:219–241. doi:210.1177/0743558404273073. [Google Scholar]
  23. Eccles JS, Barber BL. Student council, volunteering, basketball, or marching band: What kind of extracurricular involvement matters? Journal of Adolescent Research. 1999;14:10–43. [Google Scholar]
  24. Eccles JS, Barber BL, Stone M, Hunt J. Extracurricular activities and adolescent development. Journal of Social Issues. 2003;59:865–889. doi:810.1046/j.0022-4537.2003.00095.x. [Google Scholar]
  25. Feinstein L, Bynner J, Duckworth K. Leisure contexts in adolescence and their effects on adult outcomes. London, UK: Centre for Research on the Wider Benefits of Learning; 2005. [Google Scholar]
  26. Feinstein L, Bynner J, Duckworth K. Young people’s leisure contexts and their relation to adult outcomes. Journal of Youth Studies. 2006;9:305–327. doi:310.1080/13676260600805663. [Google Scholar]
  27. Ferri E, Bynner J, Wadsworth M. Changing Britain, changing lives. London, UK: Institute of Education; 2003. [Google Scholar]
  28. Finlay AK, Flanagan C. U.S. adolescents’ discretionary activities across time: Combinations of civic engagement, free time, and alcohol use. 2012 Manuscript in preparation. [Google Scholar]
  29. Flanagan C. Developmental roots of political engagement. PS: Political Science and Politics. 2003;36:257–261. doi:210.1017/S104909650300218X. [Google Scholar]
  30. Flanagan C. Volunteerism, leadership, political socialization, and civic engagement. In: Lerner RM, Steinberg L, editors. Handbook of adolescent psychology. 2. Hoboken, NJ: John Wiley & Sons; 2004. pp. 721–745. [Google Scholar]
  31. Fowler JH, Baker LA, Dawes CT. Genetic variation in political participation. American Political Science Review. 2008;102:233–248. doi:210.1017/S0003055408080209. [Google Scholar]
  32. Fredricks JA, Eccles JS. Is extracurricular participation associated with beneficial outcomes? Concurrent and longitudinal relations. Developmental Psychology. 2006;42:698–713. doi: 10.1037/0012-1649.42.4.698. doi:610.1037/0012-1649.1042.1034.1698. [DOI] [PubMed] [Google Scholar]
  33. Gallant K, Smale B, Arai S. Civic engagement through mandatory community service: Implications of serious leisure. Journal of Leisure Research. 2010;42:181–201. [Google Scholar]
  34. Graham JW. Missing data analysis: Making it work in the real world. Annual Review of Psychology. 2009;60:549–576. doi: 10.1146/annurev.psych.58.110405.085530. doi:510.1146/annurev.psych.1158.110405.085530. [DOI] [PubMed] [Google Scholar]
  35. Graham JW, Cumsille PE, Elek-Fisk E. Methods for handling missing data. In: Schinka JA, Velicer WF, editors. Research methods in psychology. New York: Wiley; 2003. pp. 87–114. [Google Scholar]
  36. Hanks M, Eckland BK. Adult voluntary associations and adolescents socialization. The Sociological Quarterly. 1978;19:481–490. doi:410.1111/j.1533-8525.1978.tb01191.x. [Google Scholar]
  37. Hardy SA, Pratt MW, Pancer SM, Olsen JA, Lawford HL. Community and religious involvement as contexts of identity change across late adolescence and emerging adulthood. International Journal of Behavioral Development. 2011;35:125–135. doi:110.1177/0165025410375920. [Google Scholar]
  38. Harris AHS, Thoresen CE. Volunteering is associated with delayed mortality in older people: Analysis of the Longitudinal Study of Aging. Journal of Health Psychology. 2005;10:739–752. doi: 10.1177/1359105305057310. doi:710.1177/1359105305057310. [DOI] [PubMed] [Google Scholar]
  39. Hart D, Donnelly TM, Youniss J, Atkins R. High school community services as a predictor of adult voting and volunteering. American Educational Research Journal. 2007;44:197–219. doi: 10.3102/0002831206298173. [DOI] [Google Scholar]
  40. Henderson A, Brown SD, Pancer SM, Ellis-Hale K. Mandated community service in high school and subsequent civic engagement: The case of the “double cohort” in Ontario, Canada. Journal of Youth and Adolescence. 2007;36:849–860. doi: 10.1007/s10964-007-9207-1. [DOI] [Google Scholar]
  41. Hibell B, Guttormsson U, Ahlstrom S, Balakireva O, Bjarnason T, Kokkevi A, Kraus L. The 2007 ESPAD report: Substance use among students in 35 European countries. Stockholm, Sweden: The Swedish Council for Information on Alcohol and Other Drugs; 2009. [Google Scholar]
  42. Jennings KM, Stoker L. Social trust and civic engagement across time and generations. Acta Politica. 2004;39:342–379. doi:310.1057/palgrave.ap.5500077. [Google Scholar]
  43. Johnston LD, O’Malley P, Bachman J, Schulenberg JE. Monitoring the Future national survey results on drug use, 1975–2011. Volume I: Secondary school students. Ann Arbor, MI: Institute for Social Research, The University of Michigan; 2012. [Google Scholar]
  44. Kay T, Bradbury S. Youth sport volunteering: Developing social capital? Sport, Education and Society. 2009;14:121–140. doi:110.1080/13573320802615288. [Google Scholar]
  45. Lanza ST, Tan X, Bray BC. Technical Report No. 11–116. University Park, PA: The Pennsylvania State University, The Methodology Center; 2011. A technical introduction: A model-based approach to latent class analysis with distal outcomes. [Google Scholar]
  46. Larson RW. How U.S. children and adolescents spend time: What it does (and doesn’t) tell us about their development. Current Directions in Psychological Science. 2001;10:160–164. doi:110.1111/1467-8721.00139. [Google Scholar]
  47. Lindstrom M. Social capital, the miniaturization of community and high alcohol consumption: A population-based study. Alcohol & Alcoholism. 2005;40:556–562. doi: 10.1093/alcalc/agh190. [DOI] [PubMed] [Google Scholar]
  48. Linver MR, Roth J, Brooks-Gunn J. Patterns of adolescents’ participation in organized activities: Are sports best when combined with other activities? Developmental Psychology. 2009;45:354–367. doi: 10.1037/a0014133. doi:310.1037/a0014133. [DOI] [PubMed] [Google Scholar]
  49. Lopez MH, Levine P, Both D, Kiesa A, Kirby E, Marcelo K. The 2006 civic and political health of the nation: A detailed look at how youth participate in politics and communities. College Park, MD: CIRCLE; 2006. [Google Scholar]
  50. Marcelo K. Fact Sheet: Volunteering among high school students. Medford, MA: CIRCLE; 2007. [Google Scholar]
  51. Mayfield D, McLeod G, Hall P. The CAGE questionnaire: Validation of a new alcoholism screening instrument. American Journal of Psychiatry. 1974;131:1121–1123. doi: 10.1176/ajp.131.10.1121. [DOI] [PubMed] [Google Scholar]
  52. McFarland DA, Thomas RJ. Bowling young: How youth voluntary associations influence adult political participation. American Sociological Review. 2006;71:401–425. doi:410.1177/000312240607100303. [Google Scholar]
  53. Metz E, Youniss J. Longitudinal gains in civic development through school-based required service. Political Psychology. 2005;26:413–437. [Google Scholar]
  54. Metzger A, Crean HF, Forbes-Jones EL. Patterns of organized activity participation in urban, early adolescents: Associations with academic achievement, problem behaviors, and perceived adult support. The Journal of Early Adolescence. 2009;29:426–442. doi:410.1177/0272431608322949. [Google Scholar]
  55. Miller TR, Levy DT, Spicer RS, Taylor DM. Societal costs of underage drinking. Journal of Studies on Alcohol. 2006;67:519–528. doi: 10.15288/jsa.2006.67.519. [DOI] [PubMed] [Google Scholar]
  56. Morrow-Howell N, Hinterlong J, Rozario PA, Tang F. Effects of volunteering on the well-being of older adults. The Journals of Gerontology. 2003;58B:S137–S145. doi: 10.1093/geronb/58.3.s137. doi:110.1093/geronb/1058.1093.S1137. [DOI] [PubMed] [Google Scholar]
  57. Musick MA, Herzog AR, House JS. Volunteering and mortality among older adults: Findings from a national sample. The Journals of Gerontology. 1999;54B:S173–S180. doi: 10.1093/geronb/54b.3.s173. doi:110.1093/geronb/1054B.1093.S1173. [DOI] [PubMed] [Google Scholar]
  58. NIAAA. Alcohol Alert. Rockville, MD: Author; 2003. Underage drinking: A major public health challenge. [Google Scholar]
  59. Office for National Statistics. Voting turnout: By age and gender: Social Trends 32. Newport, South Wales: Author; 2011. [Google Scholar]
  60. Pancer SM, Pratt M, Hunsberger B, Alisat S. Community and political involvement in adolescence: What distinguishes the activists from the uninvolved? Journal of Community Psychology. 2007;35:1–19. doi: 10.1002/jcop.20176. [DOI] [Google Scholar]
  61. Paterson L. Political attitudes, social participation and social mobility: A longitudinal analysis. The British Journal of Sociology. 2008;59:413–434. doi: 10.1111/j.1468-4446.2008.00201.x. doi:410.1111/j.1468-4446.2008.00201.x. [DOI] [PubMed] [Google Scholar]
  62. Plewis I, Calderwood L, Hawkes D, Nathan G. National Child Development Study and 1970 British Cohort Study technical report: Changes in the NCDS and BCS70 populations and samples over time. London, UK: Centre for Longitudinal Studies; 2004. [Google Scholar]
  63. Read R. Creating reflexive volunteers? Young people’s participation in Czech hospital volunteer programmes. Journal of Youth Studies. 2010;13:549–563. doi:510.1080/13676261003801812. [Google Scholar]
  64. Rochester C. Making sense of volunteering: A literature review. London, UK: Volunteering England; 2006. [Google Scholar]
  65. Sacker A, Cable N. Do adolescent leisure-time physical activities foster health and well-being in adulthood? Evidence from two British birth cohorts. European Journal of Public Health. 2005;16:331–335. doi: 10.1093/eurpub/cki189. doi:310.1093/eurpub/cki1189. [DOI] [PubMed] [Google Scholar]
  66. Salmela-Aro K, Schoon I. Youth development in Europe: Transitions and identities. European Psychologist. 2009;14:372–375. doi:310.1027/1016-9040.1014.1024.1372. [Google Scholar]
  67. Schoon I, Cheng H. Determinants of political trust: A lifetime learning model. Developmental Psychology. 2011;47:619–631. doi: 10.1037/a0021817. doi:610.1037/a0021817. [DOI] [PubMed] [Google Scholar]
  68. Schwarz G. Estimating the dimension of a model. Annals of Statistics. 1978;6:461–464. doi:410.1214/aos/1176344136. [Google Scholar]
  69. Smart RG, Adlaf EM, Knoke D. Use of the CAGE scale in a population survey of drinking. Journal of Studies on Alcohol. 1991;52:593–596. doi: 10.15288/jsa.1991.52.593. [DOI] [PubMed] [Google Scholar]
  70. Smith ES. Effects of investment in the social capital of youth on political and civic behavior in young adulthood: A longitudinal analysis. Political Psychology. 1999;20:553–580. doi:510.1111/0162-1895X.00156. [Google Scholar]
  71. Son J, Wilson J. Genetic variation in volunteerism. The Sociological Quarterly. 2010;51:46–64. doi: 10.1111/j.1533-8525.2009.01167.x. [DOI] [Google Scholar]
  72. Stanley K. Something for something: A national youth action programme. London, UK: Institute for Public Policy Research; 2004. [Google Scholar]
  73. Tan X, Lanza ST, Wagner AT. LCA distal SAS macro users’ guide (Version 1.1.0) University Park, PA: The Methodology Center, Penn State; 2011. Retrieved from http://methodology.psu.edu. [Google Scholar]
  74. Tibbits MK. PhD. The Pennsylvania State University; University Park, PA: 2009. Engagement within and across leisure activities in adolescence: Implications for substance use and sexual activity. [Google Scholar]
  75. University of London, Institute of Education. Centre for Longitudinal Studies. 1970 British Cohort Study: Sixteen-Year Head Teacher Questionnaire, 1986 [computer file] 3. Colchester, Essex: UK Data Archive [distributor]; 2008a. Aug, SN: 5225, http://dx.doi.org/10.5255/UKDA-SN-5225-1. [Google Scholar]
  76. University of London. Institute of Education. Centre for Longitudinal Studies. 1970 British Cohort Study: Twenty-Nine-Year Follow-up, 1999–2000 [computer file] 2. Joint Centre for Longitudinal Research, [original data producer(s)]; Colchester, Essex: UK Data Archive [distributor]; 2008b. Aug, SN: 5558 http://dx.doi.org/10.5255/UKDA-SN-5558-1. [Google Scholar]
  77. University of London. Institute of Education. Centre for Longitudinal Studies. 1970 British Cohort Study: Thirty-Four-Year Follow-up, 2004–2005 [computer file] 2. Colchester, Essex: UK Data Archive [distributor]; 2008c. Aug, SN: 5585 http://dx.doi.org/10.5255/UKDA-SN-5585-1. [Google Scholar]
  78. Verba S, Schlozman KL, Brady HE. Voice and equality: Civic voluntarism and American politics. Cambridge, MA: Harvard University Press; 1995. [Google Scholar]
  79. Windle M, Miller-Tutzauer C, Domenico D. Alcohol use, suicidal behavior, and risky activities among adolescents. Journal of Research on Adolescence. 1992;2:317–330. doi:310.1207/s15327795jra15320204_15327792. [Google Scholar]
  80. Zarrett N, Fay K, Li Y, Carrano J, Phelps E, Lerner RM. More than child’s play: Variable- and pattern-centered approaches for examining effects of sports participation on youth development. Developmental Psychology. 2009;45:368–382. doi: 10.1037/a0014577. doi:310.1037/a0014577. [DOI] [PubMed] [Google Scholar]
  81. Zimmer-Gembeck MJ, Mortimer JT. Adolescent work, vocational development, and education. Review of Educational Research. 2006;76:537–566. doi: 10.3102/00346543076004537. doi:510.3102/00346543076004537. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES