Real Groups: The Design and Immediate Effects of a Prevention Intervention for Latino Children

Abstract

This article describes the development and immediate effects of a small-group intervention designed to complement a school-based prevention program for children and youth. The REAL Groups intervention is the result of a partnership with predominately Mexican American schools located in the central city neighborhoods of a southwestern U.S. metropolitan area. The group members (N = 115) were fifth graders from six central city schools. Group members were identified and referred by their teachers as in need of additional support beyond the keepin'it REAL classroom-based substance abuse prevention intervention, or they were invited by the referred students. The REAL Groups followed a mutual aid approach, and Masters in Social Work student interns trained in the REAL Groups intervention served as the group facilitators. This article describes the small-group intervention and provides an initial report on the results by comparing the small-group members (n = 115) with Mexican-heritage classmates (n = 306) who only received the classroom-based keepin' it REAL prevention intervention. This is a feasibility study in preparation for the follow-up study with seventh graders. As expected due to the low drug-use rates reported by fifth-grade participants, the effectiveness results were inconclusive. The immediate findings, however, provide important information about the design and evaluation of culturally specific group interventions with acculturating children. The article provides important methodological and practice implications for small-group school-based interventions as well as recommendations for future research.

KEYWORDS:

• mutual aid
• small groups
• substance use
• prevention
• at risk
• youth
• children

INTRODUCTION

The intervention described in this article evolved as part of ongoing efforts to respond to unacceptably high substance-use rates among adolescents (Johnston, O'Malley, Bachman, & Schulenberg, 2007). Recent research shows that use rates among younger children are also increasing, and their rates and prodrug attitudes are the precursor of future use (Donovan, 2007). These trends are found across all ethnic groups, but the drug abuse literature tends to present Latino immigrant children across the age spectrum as protected from substance use (de la Rosa, 2002; Warner et al., 2006). Acculturation to mainstream American culture has been linked to prodrug norms and attitudes of immigrant Latino children, leading to higher rates of substance use (Kulis, Yabiku, Marsiglia, Nieri, & Crossman, 2007; Marsiglia & Waller, 2002). On the other hand, greater identification with culture of origin has been shown to be protective against substance use (Holley, Kulis, Marsiglia, & Keith, 2006; Marsiglia, Kulis, Hecht, & Sills, 2004). The dislocation produced by migration and the subsequent acculturation process appear to play an important role in the alcohol and other drug-use trajectory of adolescents, but less is known about the experience of younger children and when is the best time to intervene.

Despite a growing recognition of the risk effects of acculturation and the protective elements within cultures of origin, most prevention programs do not clearly integrate culture in their interventions (Gosin, Marsiglia, & Hecht, 2003). One notable exception is keepin' it REAL (Hecht et al., 2003; Marsiglia et al., 2005), a culturally grounded school-based prevention program that is recognized by the U.S. Substance Abuse and Mental Health Services Administration (SAMHSA) as a model program. Keepin' it REAL is a 10-lesson intervention targeting preadolescents, implemented by trained teachers, and accompanied by instructional videos, scripted and filmed by youth. The program aims at preventing substance use by developing children's capacity to resist drug offers with the REAL (refuse, explain, avoid, and leave) resistance strategies. The main premise of the intervention is that many children initiate substance use not because they desire to use drugs but, rather, because they lack the necessary social skills to successfully resist drug offers (Gosin et al., 2003).

As a universal prevention program (Substance Abuse and Mental Health Services Administration, 2003), keepin' it REAL takes place in regular classrooms within schools across the full spectrum of substance-use risk of the students. The program developers designed REAL Groups as a companion of the larger intervention to address the variation in risk among individual students and to target specifically children that appear to be more vulnerable to use drugs before entering adolescence. The research team, in partnership with the schools, designed and field tested the REAL Groups intervention guided by the following two exploratory hypotheses:

• Hypothesis 1: Students in the REAL Groups will report greater cultural pride, higher self-esteem, and a stronger sense of mutual aid at the completion of the group sessions, relative to baseline. Even though they were at higher risk for substance use, by the time REAL Groups' members reach middle school they will be expected to report similar use rates of alcohol, tobacco, and other drugs than students who only received the classroom-based intervention.

• Hypothesis 2: Students identified by their teachers as being at risk and who participate in the classroom-based and the companion small-group interventions will report similar substance use outcomes to those students who received only the classroom-based intervention.

THE MUTUAL AID APPROACH

The REAL Groups intervention follows a mutual aid approach to social work with groups. Mutual aid draws on resilience research, which highlights the value of social support networks and reciprocity in protecting children from negative outcomes and in facilitating their successful development (Bernard, 2004; Lee, 1986; Werner, 1989). Mutual aid is a mechanism for deriving effective support from the group members and for facilitating the creation of support networks (Shulman, 1984). Connections, relationships, and social networks provide the social capital needed to support children through their school adjustment process, and in the case of immigrant children through their acculturation process (Stanton-Salazar, 2001; Stanton-Salazar & Spina, 2003, 2005). In the mutual aid approach, group participants learn and receive support mostly from the other group members; the group facilitator's role is to support the emergence of the group process (Gitterman, 2005). A positive group process provides the stage for a fluid exchange of thoughts and experiences. Group members encourage and challenge each other through mutual aid, resulting in a collective approach to helping.

Mutual aid strengthens children's interpersonal relationship skills, further develops their personal identity, and prepares them for adolescence's decision-making situations (Bernard, 2004; Bogenschneider, 1996; Hair, Jager, & Garret, 2002; Malekoff, 2007). Mutual aid groups encourage children to connect with peers, express their personal power, and practice “equity and inclusion” (Bernard, 2004, p. 126). Mutual aid groups meet children's developmental needs and assist them to acquire critical-thinking skills, to strengthen their interpersonal relationship skills, and to develop a democratic orientation.

Mutual aid is the appropriate approach to apply with immigrant children coming from communities experiencing dislocation and stressful transitions (Steinberg, 2002). Participating in mutual aid groups enhances the ability of immigrant children to connect with peers going through similar processes (Marsiglia, 2002). Group members learn to identify shared values connected to their culture of origin; and at the same time, they can share with each other possible contradictions they experience between home, school, and peers' expectations. This approach allows group participants to contextualize risky situations by identifying challenges, protective factors, and in the case of the REAL Groups learn and rehearse specific drug resistance strategies within a cultural context.

The small-group component follows a culturally grounded orientation—that is, the lessons taught are rooted in the cultural values and norms of the community of origin (Marsiglia & Kulis, 2009). The children learn how to integrate and discuss norms and values of their culture of origin—in this case Mexican/Mexican American culture—as a resource or strength protecting them from drug use. In keeping with the developmental needs and assets of the target age group, the REAL Groups address peer relationships and interactions, prosocial behaviors, school and neighborhood adjustment, and group membership issues (Masten & Coatsworth, 1998; Phinney, Baumann, & Blanton, 2001; Phinney, Horenczyk, Liebkind, & Vedder, 2001; Phinney, Romero, Nava, & Huang, 2001).

REAL Groups: The Design and Implementation of the Small-Group Component

The REAL Groups intervention applies a variety of strategies to incorporate the mutual aid approach. Structured activities offer opportunities to generate relevant group thinking, whereas the group process facilitates reciprocity and authentic dialogue between the group members. Facilitators support group members' active participation through brainstorming, listening, evaluating options, planning, rehearsing, role-playing, applying information, and reflecting on life experiences and life choices (Gitterman, 2004; Hart, 1990). The group process promotes reciprocity by emphasizing the common needs of the members and by facilitating the development of multiple helping relationships as members give and receive support from their fellow group members.

The Role of the Group Facilitator

In the REAL Groups, group authority is decentralized and members support each other by sharing their skills and strengths (Steinberg, 1999). The facilitator engages group members as trustworthy experts on the acculturation process, school, and home experiences. The group facilitator supports members to make their voices heard and to exercise their power and potential within the safety of the group (Freire, 1998). Facilitators encourage ownership of the group by posing questions to engage students in the teaching–learning process, and by avoiding lecturing to allow students' active engagement in setting the direction of the group. Passivity, or the traditional classroom role of spectator, is consistently discouraged; instead, group members are encouraged to engage in transformative discourse and to question the master narrative's message that drug use is normative (Macedo & Freire, 2003).

The group facilitator makes members accountable for their participation. Active engagement supports the group members' ability to resist the negation of culture or origin and expands their prosocial behaviors (Freire, 1998). The dialogical method (i.e., discussion and critical thinking) teaches group members to rely on others when making decisions and allows students to connect their individual decision-making processes with their families, peers, and communities. For example, through discussion students in the small-group component learn that the belief that using drugs brings shame to the family might not be unique to their family, but instead the antidrug value may be common among most families.

REAL Groups participants received the standard teacher-taught classroom-based keepin' it REAL intervention and, in addition, took part in the 8-week psychosocial group, comprising approximately 10 children and meeting during school hours. The facilitators of the REAL Groups were masters-level social work graduate students who received intensive training in the manualized curriculum and the mutual aid approach to group work by the developers of the small-group intervention. A senior MSW group worker provided the facilitator with ongoing supervision.

Manual Content

The REAL Groups manual provides eight detailed group sessions and general instructions on how to engage participants in discussing their experiences related to the process of acculturation at their appropriate developmental level. The manual helps the facilitators connect the learned strategies with the participants' experiences and their daily lives (Fedele, 2004a, b; Masten & Coatsworth, 1998). The key topics in the eight sessions are

1. When you do not know – fostering mutuality in relationships

2. What is in a name? – recognizing and asserting personal needs linked to culture of origin

3. Let's make room for everyone – balancing uniqueness with inclusion

4. Where are you from? – valuing the self and the history of migration

5. My neighborhood – valuing the self as a resource to others

6. Dream and act – maintaining a vision of the future and acting to realize that vision

7. My family and friends – cultivating a sense of belonging

8. You can count on me – connecting with support networks.

The group sessions help students discuss, rehearse, and apply the REAL resistance strategies to real-life situations connected with aspects of their culture of origin that protect them from risk, such as culturally supported antidrug norms. Students learn together, support each other, author their stories, and rehearse options and choices consistent with cultural norms learned at home and in their communities (Arrington & Wilson, 2000). The REAL Groups intervention aims at making explicit deeply held cultural values and potential value conflicts (Bogenschneider, 1996; Castro & Alarcón, 2002; Greene & Lee, 2003; Hair et al., 2002; Masten & Coatsworth, 1998). The group sessions provide members with opportunities to discuss, address, clarify, and redefine misconceptions and stereotypes about them and their communities of origin. This article reports on the design of the group intervention and its immediate evaluation of effectiveness. A more in-depth follow-up assessment of the same children will follow as group members' progress into middle school and their alcohol, tobacco, and other drug-use rates become more serious.

METHOD

Data for this study came from the Drug Resistance Strategies Project, a longitudinal randomized control trial of keepin' it REAL involving 30 center-city public schools and funded by the National Institute on Drug Abuse. The initial phase of the study included every fifth-grade classroom in the participating schools. The team secured active parental consent and student assent (in accordance with university and school district Institutional Review Board policies) from approximately 82% of the eligible students. University-trained survey proctors administered a 1-hour, written questionnaire for the evaluation of the classroom-based universal prevention program. Similarly, proctors administered a 30-minute written questionnaire in the REAL Groups for the evaluation of the indicated small-group intervention. The pretests were administered in Fall 2004, and the posttests were administered in Spring 2005, after the classroom-based intervention and REAL Groups were implemented. Students had a choice to complete the surveys in either English or Spanish. Survey administrators informed students that the surveys were part of a research project, their participation was voluntary, and their answers to the questions were confidential. More than 96% of the students with parental consent completed the surveys.

Sample Design

The subsample analyzed here included 421 consented students who were in the six schools that implemented the REAL Groups in addition to the classroom-based program. The sample was gender balanced (50% female). Ninety-five percent were aged 10 or 11 years. The same percent of participants (95%) was from low-income families, as indicated by their participation in the federal free or reduced-price lunch program at school. The majority (61%) self-identified as Mexican American or Chicano, whereas 37% self-identified as Mexican, and 2% self-identified as Mexican and some other group. More than one half (54%) were native-born children of immigrants, 26% were foreign born, and 19% were native-born children of native parents.

Teachers completed a 10-item referral form, modeled after the Search Institute's (2010) list of developmental assets for middle childhood. Teachers referred students with a more diffused self-image, lower social skills, known or suspected substance use, and/or exposure to substance abuse through family members. The referred students were encouraged to invite other classmates as a means of preventing the onset of iatrogenic effects common in homogenous groups (Dodge, Dishion, & Landsford, 2007). If all group members share an identified need or risk, it is possible that they could reinforce the shared risk factor, and thus the participants' behavior will not change in the desired direction (Felps, Mitchell, & Byengton, 2006).

Of the 421 students in the subsample, 115 participated in the REAL Groups, forming 12 small groups (two in each randomly selected school); the rest of the students (n = 360) became the default comparison group. The demographics of the group members resembled those of the total sample, with one exception. REAL Groups participants comprised slightly more males (53%) than females. The noted slight gender unbalance is not surprising given that teachers referred students, among other things, based on their substance use risk, and boys tend to be at higher risk for substance use (Dakof, 2000). Analyses related to Hypothesis 1 included only group members (n = 115), whereas analyses related to Hypothesis 2 included all the students in the schools where REAL Groups were implemented (N = 421) divided into intervention and comparison groups.

REAL Groups Outcomes

Analyses related to Hypothesis 1 examined three sets of REAL Groups outcomes: (1) cultural pride, (2) self-esteem, and (3) mutual aid. All items had five response options: strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree. Table 1 provides a detailed listing of the three sets of items.

TABLE 1 Means and Standard Deviations of REAL Groups Outcomes (N = 115)

	Pretest M (SD)	Posttest M (SD)
Cultural pride
There is more than one right way to see life.	4.15(.89)	4.23(.84)
Not everyone's opinion matters. ^a	3.02(1.37)	2.98(1.36)
Each culture (race, ethnicity, nationality) in the U.S. adds something valuable to society.	3.79(.97)	3.83(1.03)
Self-esteem
I have a number of good qualities.	4.10(.92)	4.02(.91)
I don't have much to be proud of. ^a	3.58(1.31)	3.63(1.30)
I have a positive attitude toward myself.	3.86(1.18)	3.66(1.07)
Mutual aid
When I am afraid, I can admit it to other people.	3.32(1.35)	3.36(1.24)
If someone mispronounced my name, I would tell them the right way to say it.	4.17(1.07)	4.28(.97)
I talk about my ideas with other people, even if I think they might not agree.	3.69(1.17)	3.63(.99)
I don't really listen when people talk because I already know what they're going to say. ^a	3.47(1.32)	3.44(1.28)
I make better decisions after I hear other people's ideas.	3.62(1.20)	3.75(1.14)
Some people are scared for no reason. ^a	2.80(1.34)	2.79(1.13)
I don't have to like somebody to understand how they feel.	3.37(1.34)	3.35(1.24)
Note. ^aHigher posttest means were desired except for these items.

Substance Use Outcomes

For Hypothesis 2, the analysis examined 13 substance use attitudes and behaviors as outcomes of the classroom-based intervention. Posttest assessments took place after the implementation of the classroom-based intervention; high values represented undesirable outcomes, that is, indicating higher levels of substance use or stronger prodrug attitudes.

The surveys assessed the frequency of students' substance use in the last 30 days and the self-reported amount consumed of each substance. Participants will complete the same questionnaire in subsequent waves of data collection. The items are developmentally appropriate for middle childhood through preadolescence (Hecht et al., 2003; Kandel & Wu, 1995). Students indicated the frequency of recent substance use by responding to separate questions addressing the number of times they used alcohol, cigarettes, marijuana, and inhalants (from 1 = 0 times, to 6 = 40 or more times). Students indicated the amount consumed by reporting how many drinks of alcohol (from 1 = none, to 7 = more than 30), cigarettes (from 1 = none, to 7 = more than 20), and hits of marijuana they had consumed (from 1 = none, to 7 = more than 40). The analyses treat answers regarding each substance separately. A baseline measure of the substance use outcome, captured at the pretest, was included as a control variable in the regression analyses.

In addition to measures of actual substance use, we examined an array of drug-use attitudes that are precursors of substance use (Elek, Miller-Day, & Hecht, 2006): prodrug personal norms, drug offer vulnerability, intentions to use substances, positive substance use expectancies, prodrug parental injunctive norms, and prodrug friends' injunctive norms. The six scales had good to excellent internal consistency (Cronbach's alpha coefficients of 0.83 to 0.97). The mean of three personal prodrug norms captured the students' opinions on whether use of alcohol, cigarettes, and marijuana was “OK” for someone their age. Responses ranged from 1 (definitely not OK), to 4 (definitely OK).

The self-efficacy and drug offer vulnerability scale (Kasen, Vaughan, & Walter, 1992) used the mean of three items that captured students' confidence in their ability to refuse drug offers. Students indicated the extent to which they were sure they would say no to an offer of alcohol from a “family member,” an offer of a cigarette from “a kid at school,” and an offer of marijuana from a “close friend.” Responses ranged from 1 (very sure), to 4 (not at all sure).

The average of three items captured substance use intentions by whether the students thought they would use alcohol, cigarettes, and marijuana in the coming weekend if they had the chance. Reponses ranged from 1 (definitely no), to 4 (definitely yes).

An average of three items captured the respondents' perceived benefits of substance use (positive substance use expectancies). Items included “Drinking alcohol makes parties more fun,” “Smoking cigarettes makes people less nervous,” and “Smoking marijuana makes it easier to be part of a group.” Responses ranged from 1 (strongly disagree), to 4 (strongly agree) (Hansen & Graham, 1991).

Two items measured injunctive norms, which entail expected negative reactions from significant others to the student's use of substances (Hansen, Johnson, Flay, Graham, & Sobel, 1988). The average of three items captured students' perceptions of how angry their parents would be if they found out that they had drunk alcohol, smoked cigarettes, or smoked marijuana. Responses ranged from 1 (very angry), to 4 (not at all angry). An average of three items captured students' report of how their best friends would react (friends' injunctive norms) if they got drunk, smoked cigarettes, or smoked marijuana. Responses ranged from 1 (very negatively), to 3 (no reaction), to 5 (very positively).

Covariates

A dichotomous variable captured whether the student participated in the REAL Groups (1 = yes, 0 = no). Age in years was an ordinal measure with responses ranging from 1 (7 years), to 9 (15 years or older). Gender was dichotomous with 1 (female) and 0 (male). Participation in the school's federal free or reduced-price lunch program served as a measure of socioeconomic status: 0 (no participation), 1 (free or reduced-price lunch). Usual grades received in school, a common predictor of substance use (Wright & Pemberton, 2004), was captured by an ordinal variable with responses ranging from 1 (mostly Fs), to 9 (mostly As).

The analysis followed a three-step strategy. First, we conducted paired-samples t tests to assess pretest-to-posttest changes in the REAL Groups outcomes (cultural pride, self-esteem, and mutual aid) among the 115 participants who completed the pre- and post-REAL Groups evaluations. Second, we conducted independent samples t tests to assess for differences in means in the pretest substance use outcomes between participants in the classroom program and the REAL Groups and participants in the classroom program only. Third, we conducted linear regression analyses to assess whether participation in the classroom program and the REAL Groups predicted more or less increases in posttest substance use than participation in the classroom program only.

RESULTS

Table 2 shows the pretest and posttest means and standard deviations of the REAL Group proximate outcomes of cultural pride, self-esteem, and mutual aid. Differences were in the desired direction, but the paired-samples t tests yielded no statistically significant changes in means on these variables from pretest to posttest.

TABLE 2 Pretest Means and Standard Deviations, Classroom-Only and REAL Groups Participants (N = 421)

	Participants in Classroom Program Only (n = 306)	Participants in Classroom Program and REAL Groups (n = 115)
Amount consumed in last 30 days
Alcohol	1.15(.51)	1.15(.50)
Cigarettes	1.05(.33)	1.04(.24)
Marijuana	1.04(.40)	1.03(.37)
Frequency of consumption in last 30 days
Alcohol	1.11(.38)	1.15(.54)
Cigarettes	1.02(.15)	1.04(.31)
Marijuana	1.02(.24)	1.03(.37)
Inhalants	1.05(.27)	1.03(.18)
Drug offer vulnerability	1.84(1.24)	2.02(1.29)
Positive drug expectancies	1.39(.71)	1.35(.54)
Parents' prodrug norms	1.31(.78)	1.26(.79)
Friends' prodrug norms	1.96(1.41)	1.79(1.36)
Personal prodrug norms	1.17(.38)	1.19(.38)
Intentions to use substances	1.20(.46)	1.22(.43)

Before testing Hypothesis 2, we tested the assumption that REAL Groups participants were in fact at higher risk for substance use at baseline. Table 2 contains descriptive statistics of the substance use attitudes and behaviors of the REAL Groups participants and the classroom-program-only participants. There were no statistically significant differences between the two groups, indicating that the REAL Groups participants were not significantly different from nonparticipants in their vulnerability levels. This lack of difference remained, even after we excluded invited (or nonreferred) REAL Groups participants.

The fact that there were no differences in risk between the two groups (i.e., nothing to be offset) renders void the expectation of similar posttest outcomes because of the students' participation in the REAL Groups. It is possible that the REAL Groups intervention would have a booster or extra dosage effect, yielding superior outcomes for participants who received the classroom program and the REAL Groups, relative to classroom-program-only participants. We ran regression analyses to test this alternate hypothesis, as well as our original Hypothesis 2.

Table 3 contains ordinary least squares regression results assessing the effect of REAL Groups participation on last-30-day substance use amount and frequency, when controlling for gender (female), the usual grades received in school, current age, school lunch program enrollment, and baseline substance use. We found that REAL Groups participation was not associated with any significant differences in outcomes related to alcohol, cigarette, or inhalant use. However, REAL Groups participation positively predicted greater increases in marijuana use at the posttest, compared to participants in the classroom program only.

TABLE 3 Unstandardized Linear Regression Estimates and Standard Errors, Predicting Posttest Substance Use (N = 407)

	Amount			Frequency
	Alcohol	Cigarettes	Marijuana	Alcohol	Cigarettes	Marijuana	Inhalants
REAL Groups participation	−.019(.077)	.021(.038)	.074∗(.037)	−.013(.069)	.032(.036)	.097∗(.043)	−.037(.064)
Female	−.042(.070)	−.026(.035)	.003(.034)	.042(.062)	−.037(.033)	.037(.039)	.095(.058)
Usual grades	−.012(.020)	−.017+(.010)	−.013(.010)	−.017(.018)	−.021∗(.009)	−.013(.011)	−.020(.017)
Age	.141∗(.062)	.038(.031)	.048(.030)	.082(.055)	.040(.029)	.109∗∗(.034)	.099+(.051)
School lunch participation	.021(.065)	−.018(.032)	−.012(.031)	−.047(.059)	−.020(.031)	−.032(.036)	−.061(.054)
Pretest substance use	.298∗∗∗(.070)	−.034(.055)	.161∗∗∗(.043)	.186∗(.072)	.091(.077)	.313∗∗∗(.066)	.243∗(.113)
Intercept	.365∗∗∗(.320)	1.069∗∗∗(.165)	.755∗∗∗(.153)	.782∗∗∗(.286)	.963∗∗∗(.168)	.347+(.180)	.621∗(.287)
Adj. r ^2	.047	.002	.050	.015	.016	.089	.019
N	401	401	407	401	404	406	407
+p < .10, ∗p< .05, ∗∗p< .01, ∗∗∗p< .001

An examination of the covariates models presented in Table 3 reveals that baseline substance use positively predicted—for most substances—posttest substance use. Better grades in school were associated with lower posttest substance use, but the effect was only statistically significant in the case of the frequency of cigarette use. Older students were associated with a greater amount of alcohol use and a greater frequency of marijuana use at posttest.

Another set of regression models examined the effect of participation in the REAL Groups on six substance use attitudes: drug offer vulnerability, use intentions, parents' prodrug norms, friends' prodrug norms, personal prodrug norms, and positive drug expectancies (results not shown in tables). Participating in the REAL Groups did not yield statistically significant effects for these outcomes. These results do not support the alternate hypothesis that the REAL Groups had a booster effect, improving outcomes for youth who received the REAL Groups and the classroom program.

DISCUSSION

This article describes the design and reports on the immediate evaluation of a small-group component created to supplement a school-based prevention program. The REAL Groups intervention provides additional support (extra dosage) to fifth-grade students perceived to be at higher risk for substance use and to increase their responsiveness to the classroom-based prevention program. Although program results were mixed and inconclusive, the current study is instructive in highlighting methodological challenges associated with the implementation and evaluation of a small-group intervention as a companion to a classroom-based intervention.

One key challenge encountered by the current study was that the referral procedures did not yield a high-risk participant group as intended. It is possible that the concepts of high-risk and substance use vulnerability do not apply to this very young age group with a high representation of low-acculturated Latino children. Some teachers may have referred students with discipline problems, rather than those at risk for substance use. However, were this the case, we might have seen baseline differences between REAL Groups participants and the sample at large, because discipline problems are commonly associated with higher substance use or prodrug attitudes (Drapela, 2005; Wright & Pemberton, 2004). It appears to be very difficult for teachers to identify higher risk fifth-grade students. It is possible that the items on the referral form, which were intended to help teachers identify the risk level of the students, were developmentally not well adapted, or may not have matched well the social setting or the cultural background of the participants.

The pattern of effects for many of the outcomes was in the desired direction, but the evaluation of the behaviors specifically targeted by the REAL Groups interventions (cultural pride, self-esteem, and mutual aid) did not yield conclusive evidence of significant behavioral change. The posttest was administered immediately after the small-group intervention, and the lack of follow-up measures later in time prevents us from assessing the REAL Groups' longer term impact. The small sample size of 115 cases with complete data undermined the power of the statistical tests used to identify significant differences. It is also possible that the outcome measures lacked the sensitivity to capture the small-group component's effects. On the other hand, the REAL Groups participants could have been in an ideal situation to benefit fully from a primary prevention intervention. The effects of the small-group intervention will be captured long term when risk behaviors intensify in middle school and high school due to developmental and contextual factors.

REAL Groups participants did not differ from their classroom-program-only peers in their alcohol, cigarette, and inhalant use rates and prodrug attitudes. We cannot conclude that the small-group component compensated for differential substance use risk, because we did not find baseline differences between the two groups. The increase in self-reported marijuana use among a small number of REAL Groups participants was unexpected. The between-groups differences may be due to chance; but alternatively, it may be an unintended iatrogenic effect of the REAL Groups. The teacher referral process may have somehow caused REAL Groups participants to feel singled out and to engage in behaviors they felt were expected from them—that is, to be referred to the group became a self-fulfilling prophecy. Additional data on students' perceptions of their participation in the groups could be useful in testing this hypothesis. Because the invited group members were very similar to the referred members in all key outcomes, there is a possibility that the inclusion of invited friends by the referred students may have had the unintended effect of augmenting the iatrogenic effect.

Children that are more vulnerable to use drugs tend to move faster through the drug-use continuum in comparison to lower risk groups. Thus, the group intervention might have had an effect in slowing down the small-group participants' progression into drug use, though we do note the surprising increase of marijuana use for some of them. Perhaps the indicated intervention should have focused on a smaller number of students with the greatest need in each classroom; by engaging so many students, the intervention may have not achieved its intended population.

When students in the experimental and comparison groups reach the seventh grade, the longer term prevention effects of the small-group intervention will be assessed more clearly. In the meantime, this article provides a detailed description of the approach and the possible methodological challenges practitioners and researchers might face while assessing the effects of a companion small-group intervention with acculturating children and youth.

Recommendations for Future Practice Research

Project teams need to conduct an in-depth examination of the measures and need to work very closely with teachers to explain the referral criteria and their rationale. Once teachers make referrals, research team members should review and discuss the referral list with teachers to ensure that the referred students meet the membership criteria for the intervention. Conducting a separate analysis of the referral forms could be useful in investigating and documenting an early detection of a lack of distinction between groups and in implementing a corrective action.

Practitioners should consider alternative research designs when researching the benefits of mutual aid groups. For an indicated prevention intervention, it is important to collect evidence that the targeted group is at higher risk for adverse outcomes than the nontargeted group. An alternative design could include using the pretest data to distinguish who was at a higher risk for drug use. It might also be beneficial to assign half the at-risk group to receive the indicated intervention and to continue with the classroom-only (universal) intervention with the other half, creating a more identifiable control group.

The psychometrics and cultural relevancy of small-group interventions need to be further developed, and it could be beneficial to include additional measures to capture participants' feedback and to document the group process more in-depth. Some of the questions in need of further exploration are: (1) Did participants feel part of the group? (2) Did they receive the aid of others? (3) Did they provide aid to others? (4) If so, how was that experience? (5) To what extent was the group transformative for the participants? Measures of students' personal perceptions of the experience might help to better capture its effects. It would also be helpful to integrate into the model the fidelity data collected through interviews and observations of the facilitators. Additional data about the facilitators such as gender, ethnicity, language abilities, and cultural competency will allow for a closer examination of the social worker's role.

Additional questions researchers/practitioners need to ask themselves are: (1) How does one select an at-risk group without creating stigma? (2) Are the resources available for implementing the intervention? (3) What are the benefits relative to the costs? (4) What elements of the intervention are particularly important for producing positive change? (5) Which elements should one strengthen? (6) How generalizable is the intervention?

The examination of the seventh-grader results—a developmental period characterized by spikes in substance use onset (Hornick, 2003)—will allow for a more comprehensive assessment of developmental differences and may offer findings that are more conclusive. Seventh graders will also allow the research team to better test the extra-dosage hypothesis.

The current study provided a snapshot of the design and performance of a small-group component based on the mutual aid approach and designed to supplement the primary prevention program keepin' it REAL. Given the implementation issues associated with referral of group participants, our assessment of the REAL Groups' effectiveness was inconclusive. Further research will assess how this innovative intervention addresses the needs of higher risk students who participate in the school-based universal prevention program. The current study also documents the complexities that arise when designing and evaluating interventions with younger children prior to significant engagement in measurable antisocial behaviors. Conducting prevention with younger children in small groups is necessary even when the effectiveness of such programs is not easily assessed through conventional methods. The challenges identified by the current study call for innovation and more tailored research models.

ACKNOWLEDGMENTS

The National Institute on Drug Abuse (R01 DA05629) supported this study, and the manuscript was developed at the Southwest Interdisciplinary Research Center with support from the National Center on Minority Health and Health Disparities (award P20 MD002316--01, F. Marsiglia, P. I.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Drug Abuse, the National Center on Minority Health and Health Disparities, or the National Institutes of Health.