Positive Individual and Social Behavior Among Gang and Nongang African American Male Adolescents To explore potential bases of positive development among gang youth, attributes of posi-tive individual and social behavior were assessed in individual interviews with 45 Afri-can American adolescent male members of inner-city Detroit gangs and 50 AfricanAmerican adolescent males from the same communities but involved in community-based organizations aimed at promoting positive youth development. As anticipated, thegroups differed in regard to the majority of interviewquestions and to positive attributescores pertaining to parents/family, peer relations, school/education, drug use, sexualactivity, religious activities/religiosity, racial/ethnic identity, role models/confidants,and neighborhood/safety. The correlations of attributes scores were more often signifi-cant (i.e., coupled) for the gang than for the nongang youth. Consistent with the ideas thatall young people have resources pertinent to positive development and that, therefore,gang and nongang youth would have some resource comparability, across the nineattributes, about one quarter of the gang youth had total positive attribute scores thatwere above the average total positive attribute score for the nongang youth. Implicationsof these findings for both research and applications to programs seeking to promote posi-tive youth development among diverse youth are discussed.Keywords: positive youth development; developmental systems theory; African American; adolescent males; gang youth; community-based organizations
Resilience, coping, or the adaptive regulation of behavior across ontogeny
have been foci of developmental research aimed at understanding how ado-lescents, in the midst of the internal and contextual changes prototypic of this
period of life, maintain positive, healthy trajectories (Adams & Marshall,1996; Benson, 1997; Grotevant, 1998; Luthar, Cicchetti, & Becker, 2000). Current theoretical models of adolescent development (e.g., Adams, 1997;Adams, Day, Dyk, Frede, & Rogers, 1992; Benson, 1997; Damon, 1997;Lerner, 1996, 1998, 2002; Muuss, 1996) stress that individual differences inthe positive or negative outcomes of adolescents’ regulatory behaviors areproduced by their history of person-context relations. These relations involveindividual and contextual variables seen as fused within a dynamic develop-mental system (Lerner, Freund, DeStefanis, & Habermas, 2001).
Recently, such developmental systems models have been developed to
understand the characteristics of the development of youth of color (e.g.,Fisher, Jackson, & Villarruel, 1998; Lerner, Sparks, & McCubbin, 1999;McAdoo, 1995, 1998, 1999; Spencer, 1990, 1995, 1999; Spencer, Dupree, &Hartmann, 1997) and, specifically, the resiliency or adaptive modes (e.g.,McAdoo, 1995; Spencer, 1990) used by adolescents of color and their familyand community contexts to promote positive development, especially in eco-nomically poorcommunities. This emphasis on positive youth developmentamong pooryouth of colorconstitutes a significant innovation in develop-mental research (e.g., Benson, 1997; Leffert et al., 1998; Scales, Benson,Leffert, & Blyth, 2000), particularly within the adolescent developmentalperiod where youth are engaging in negative, unhealthy behaviors at histori-cally unprecedented levels (e.g., Dryfoos, 1990; Hamburg, 1992; Lerner,1995; Lerner & Galambos, 1998). Race is the single best predictor of povertyand its behavioral sequelae (Huston, 1991; McLoyd, 1998). Key instances ofthese problem behaviors are delinquency, violence, substance use, andunsafe sex (Dryfoos, 1990; Huston, 1991).
In the midst of such trends, there are also numerous success stories of pos-
itive developmental outcomes among pooryouth of color(Allison, 1993;McAdoo, 1995, 1998, 1999). There are many instances of positive, healthydevelopment of African American male adolescents—arguably the groupwith the highest probability of experiencing the problematic behaviors asso-ciated with race and poverty in America (Mincy, 1994). There is evidencethat the bases of the positive development of these youth and theirability toovercome the odds (Werner & Smith, 1992) against their positive develop-ment lie in combinations of individual and ecological characteristics.
This research was supported in part by a grant from the W. T. Grant Foundation. The authors are grateful for theuseful theoretical and methodological comments of two anonymous reviews. Address correspondence to ei-ther Carl S. Taylor, Institute for Children, Youth, and Families, 27 Kellogg Center, Michigan State University,East Lansing, MI 48824, or to Richard M. Lerner, Eliot-Pearson Department of Child Development, 105 Col-lege Avenue, Tufts University, Medford, MA 02155.
JOURNAL OF ADOLESCENT RESEARCH / November 2003
Luster and McAdoo (1994) sought to identify the factors that contribute to
individual differences in the cognitive competence of African American chil-dren in the early elementary grades. They found that favorable outcomes incognitive and socioemotional development (operationalized as attainmentswithin the top quartile of the sample they studied) were associated with highscores on an advantage index. This index was formed by scoring children onthe basis of the absence of risk factors (e.g., pertaining to poverty or problemsin the quality of the home environment) and the presence of more favorablecircumstances in their lives.
Luster and McAdoo (1994) reported that only 4% of the children in their
sample who scored low on the advantage index had high scores on a measureof vocabulary. In contrast, 44% of the children who had high scores on themeasure of developing under favorable circumstances also had high vocabu-lary scores. Similar differences between low and high scorers on the advan-tage index were found in regard to measures of math achievement (14% vs. 37%, respectively), word recognition (0% vs. 35%, respectively), and wordmeaning (7% vs. 46%, respectively).
Findings reported by Luster and McAdoo in 1996 extended those that they
reported in 1994. Seeking to identify the factors that contribute to individualdifferences in the educational attainment of African American young adultsfrom low socioeconomic status, Luster and McAdoo (1996) found that assetsassociated with the individual (cognitive competence, academic motivation,and personal adjustment in kindergarten) and the context (parental involve-ment in schools) were associated longitudinally with academic achievementand educational attainment.
Given these data about the linkage between individual and ecological
variables and positive development among African American youth, thisreport brings data to bear on the adaptive modes (Spencer, 1990)—the indi-vidual and ecological strengths—present among youth embedded in poor,inner-city neighborhoods and either involved in neighborhood gangs or incommunity-based organizations (CBOs) intended to promote positive devel-opment. Our research is consistent with the interests pursued by Werner andSmith (1977, 1982, 1992) in theirKauai Longitudinal Study and with thecalls forqualitative studies of the pathways to positive development amongall African American youth (Fisher et al., 1998; Jarrett, 1998; McAdoo,1995; McLoyd, 1990; Spencer, 1990). In addition, our research is consistentwith both theory (Joe, 1993; Neely, 1997) and qualitative research (Patton,1998) about African American adolescent male gang members. This litera-ture indicates not only the unsurprising differences between gang andnongang adolescents but, as well, the comparability in family, teacher, and
peer resources for positive development between at least some gang mem-bers (e.g., about 20%; Patton, 1998) and their nongang counterparts.
We have launched a study wherein we attempt to identify the individual
and contextual conditions that may provide comparable resources for posi-tive development among both gang and nongang youth. These resources mayprotect diverse African American male adolescents from the actualization ofrisk (i.e., that allow them to overcome the odds) and, in turn, may promotepositive, healthy, successful development. The qualitative interview methodsdeveloped by Taylor (1990, 1993), procedures that are consistent with otherethnographic approaches to studying gang youth (Patton, 1998; Valdez &Kaplan, 1999), were used in the present research.
The present report is derived from the first wave of a longitudinal study of
African American male adolescents who are either (a) involved in gangsengaged in criminal behavior (e.g., drug use, violence) or (b) participating innongang-related community groups (e.g., 4-H clubs, church-related organi-zations). The goal of this report is to present the methods of our research andthe baseline, descriptive, individual, and ecological characteristics associ-ated with the gang and the nongang youth involved in the project. Sample Characteristics
Participants in this project, which we have labeled Overcoming the Odds
(OTO), were recruited from among the adolescents involved in Taylor’s(1990, 1993, 1996, 2000) Michigan Gang Research Project (MGRP)—anongoing ethnographic field research project with male and female adolescentgang members in Detroit. In the context of providing youth services/programs to the adolescents and families in the poor, African American sec-tions of the city of Detroit, the MGRP is an omnibus research project fromwhich all of Taylor’s (1996, 2000) specific research projects derive. TheMGRP is both a research program and a means to return to the communitysomething of value (support services) for its youth.
The gang members (N = 45) who are participating in OTO have been
involved in their respective gangs for between 5 to 10 years. These youthswere in their gangs throughout their participation in the MGRP; this involve-ment ranged from 5 to 12 years. All gang members volunteered to take part inthe OTO study. The size of this sample is the result of considerations of time,money, and statistical power.
JOURNAL OF ADOLESCENT RESEARCH / November 2003
Recruitment of the nongang members (N = 50) occurred through contacts
with community members participating in the MGRP. All nongang membersvolunteered, as well. The size of this sample represents the same feasibilityissues as noted above. The slightly larger number of nongang youth resultedfrom the efforts of community contacts and the decision not to decline theparticipation of any youth who volunteered to take part. The youth in thenongang group had never been part of a gang, and youth in this group havebeen involved in the MGRP forthe same range of time as have the gangyouth. The nongang youth participated in at least one of several CBOsdesigned to promote positive youth development (e.g., church groups, 4-Hclubs, orMGRP activities conducted by Taylor; Taylor, 1993, 1996). Assuch, we will refer, hereafter, to the youth in the nongang group as the CBOyouth.
In both the gang and the CBO groups, most youth were between 14 and 18
years of age at their most recent birthdays (mean ages = 15.82 years and16.31 years, respectively). The age distribution among the adolescents in thegang group was somewhat younger than that in the CBO group, χ2(8) = 18.4,p < .05. In both groups, the majority of youth were born in Detroit; however,significantly more CBO youth lived with their parents than was the case withthe gang youth, χ2(5) = 16.9, p < .01. In addition, the groups differed signifi-cantly in regard to parental education, χ2(5) = 21.9, p < .001. The CBO youthwere more likely to have parents who completed high school or some othertype of school, whereas the gang youth were more likely to report that theirparents did not complete any school.
Both the gang and the CBO youth were either students in or, had they
attended school, would have been students in one of 20 schools locatedwithin the boundaries of the city of Detroit. It is important to note, however,that forgang youth, school attendance is poor, at best. Forthese youth, theschool is used as a means to identify the area where they live. The preponder-ant majority of the citizens in the economically depressed neighborhoodswithin which these schools are located are people of color. Sampling Issues
Self-selection was the basis forentry into the study forboth groups of
males. Both sets of males also self-selected—albeit through substantially dif-ferent, longer term, and less well-understood processes—into their respec-tive gang orCBO groups. We believe that self-selection is an unavoidablemethodological feature of gang research (Taylor, 1990, 1993), and its pres-ence necessitates caution in formulating generalizations.
However, Valdez and Kaplan (1999) noted that bias in the selection pro-
cess involving gang members and others in their communities is minimizedwhen use is made of ethnographic procedures such as those used in thisresearch. That is, as discussed below, these procedures include extensiveimmersion into the social world of the gang members, avoiding institutionalreferences and relying on personal relationships, maintaining high visibilityin the community, making and maintaining social contacts throughout thecommunity, and using community gatekeepers, such as gang leaders, toobtain access to gang members. As recommended by Valdez and Kaplan inregard to preferred ethnographic methodology for research with adolescentgangs, both gang leaders and consultants from the communities (teachers,counselors, parents, religious leaders, and former gang members) assistedthe primary data collector, Carl S. Taylor, in gaining interviews with both thegang and CBO participants. Interview Method
Consistent with the methodological recommendations of Joe (1993) and
of Valdez and Kaplan (1999), Taylor’s (1990, 1993, 1996, 2000) methodsrely centrally on being part of and serving the community. Such immersionelicits trust that the promise of anonymity of responses will be kept, that therewill be an openness about the study’s purposes, and that there will be readyaccess to the investigators. Moreover, the community understands that inter-view items will be presented in a manner compatible with the literacy needs,the language style, and the social settings of the participants (Taylor, 1990).
As underscored by Joe (1993) and Valdez and Kaplan (1999), the ethno-
graphic interview methods used by Taylor (1990, 1993, 1996, 2000) areneeded because many of the participants, regardless of the nature of theirgang membership or affiliation with another community group, will notrespond in writing to a questionnaire. Participants often display a high degreeof illiteracy (requiring that items have a closed-ended format and are read tothe youth), and gang members, in particular, become very uncooperativewith any mention of names or personal references that might jeopardize themas informants or leave them vulnerable to law enforcement authorities.
To maximize their willingness to respond to the interview questions,
youth have the gatekeeperoption noted above. That is, they may participatein group interviews (e.g., with their gang leaders present). In the first wave ofdata in OTO, 100% of the CBO youth participated in individual interviews,and 75% of the gang youth participated in group interviews.
Taylor (1990) demonstrated that the potential methodological problems
of interviewing youth in a group (e.g., response contamination) represent a
JOURNAL OF ADOLESCENT RESEARCH / November 2003
trade-off necessary to make if one is to collect truthful information about thetopics of concern in this research. The rationale for using group interviews isthat it allows the youth, and particularly gang members, to relax, to see othergang members discuss questions, and to introduce our method of reading thequestions and writing answers down. When group sessions occur, they arefreely and spontaneously conducted. They have a feeling of closeness, muchlike that of a locker room or bull session. The Interview Protocol: Content and Coding
Through a series of ratings by experts of pools of items associated in the
adolescent development literature with measures of risk and resiliency fol-lowed by focus groups about item phrasing held with African Americanyouth, Taylor, Villarruel, & Lerner (1998) developed the interview employedwith the sample. The protocol for this interview is presented in the appendix. As shown in the appendix, the protocol contains 5 items pertaining to demo-graphic information (an unnumbered item about age, Item 1 about peoplewith whom the participant lives, Item 2 about parents’ birthplaces, Item 3about the participant’s place of birth, and Item 17 about parents’ educationallevels) and 39 substantive items. As explained in Tayloret al., the substantiveitems are associated with nine categories. There are 9 parent/family items(Items 10, 11, 12, 13, 14, 15, 16, 18, and 19), 2 peerrelations items (7 and 8),5 school/education items (9, 24, 30, 32, and 33), 10 drug-use items (4, 20, 21,21a, 22, 23, 23a, 27, 28, and 29), 3 sexual activity items (34, 35, and 36), 2religious activities/religiosity items (5 and 6), 2 racial/ethnic identity items(25 and 26), 3 role models/confidants items (Items 37, 38, and 39), and 3neighborhood/safety items (31, 40, and 41).
In addition to conducting item-by-item analyses of the responses to the 39
substantive items, an attempt was made to place the response alternativesassociated with each of the items into eithera positive ora not positive cate-gory or, if necessary, a not applicable (to a positive/not positive classifica-tion) category. This coding was undertaken to begin to identify individualand social characteristics potentially reflective of the adaptive modes associ-ated with positive development among youth of color—even those involvedin such high-risk activities as urban gangs. In all cases, the positive coding(given the score of 1) was reflective of an asset (Benson, 1997; Scales &Leffert, 1999). The not positive categorizations were given the score of 0. The not applicable categorizations were given a score of 2.
To illustrate, for the drug-use item (Item 4, “Do your friends get high?”),
the response “no” (scored as 1) was dichotomized as positive versus “yes”(scored as 0), which was dichotomized as not positive. Similarly, responses
of “excellent” or “fair” in regard to Item 12 (“How would you rate your rela-tionship with your parents?”) could be categorized as positive (and scored as1), whereas the response “poor” to this question could be categorized as notpositive (and scored as 0). The response “talk it out” to Item 7 (“How do yousettle disputes between your friends?”) could be rated as positive (and scoredas 1), whereas the responses “fight with fists” or “guns” could be seen as notpositive (and scored as 0).
Across the 39 substantive questions and the 5 demographic questions,
response alternatives provided to youth ranged from 2 to 8 for a total of 132response possibilities. The raters agreed on 122 of these responses (92%) asbelonging to eitherthe positive, the not positive, orthe not applicable catego-rizations resulting in a Cohen’s kappa of .872 (p < .001). Disagreements wereresolved by consensus. Because the raters agreed that 5 of the 39 substantivequestions (Items 21a, 22, 23a, 28, 29) were not applicable to a positive/notpositive dichotomization, the final scoring included 34 questions. Of these 34items, positive scores ranged from 0 to 9 depending on the number of itemsassociated with each category. Procedure
Over the course of April and May 1999 the interview was administered to
both groups of participants. Each interview was conducted in a settingselected by the youth (e.g., a gang/CBO meeting area; a public location, suchas a restaurant; or the home of the participant or a friend/relative). Individualinterviews usually (70% of the time) were completed in one session thatlasted for an average of 45 minutes. The remaining interviews (30%) weresplit into two sessions at the request of the youth. However, in such cases, thetotal time of the interview was no different from the time taken for interviewprotocols completed in one session. Group interviews typically took signifi-cantly longer and lasted an average of 2.5 hours.
The following analyses were conducted: (a) a comparison of gang and
CBO youth on the items associated with each substantive category, (b) ascer-tainment of the variables that do and do not discriminate between the twogroups, and (c) a discussion of the intercorrelations of items within andacross groups. The goal of these analyses was not just to document areas ofdifference between the groups but, of greatest interest to us, was to determine
JOURNAL OF ADOLESCENT RESEARCH / November 2003
if there are comparable features of positive functioning within the gang andthe CBO youth. Item Comparisons of the Gang and the CBO Youth
The responses of the gang and the CBO youth were compared for each of
the 39 substantive questions in the interview. For each item, the frequencieswith which each of these two groups endorsed each response alternative werecompared using Pearson χ2 measures.1 Significant differences (p < .05)between the distributions of responses for the gang and the CBO youthoccurred for 30 (77%) of the 39 substantive analyses. To understand the pat-tern of similarities and differences in responses across the two groups, it isuseful to consider the responses that are pertinent to each of the nine substan-tive categories.
The groups differed in their response distributions to eight of the nine
parent/family items. As compared to gang youth, CBO youth were morelikely to describe their parents as having rules for the home; to see these rulesdistributed about equally across the categories of “strict,” “moderate,” and“easy”; to rate their relationships with their parents as “excellent” or “fair”; tobelieve that their parents would support them regardless of what they did inlife; to report that family disagreements are settled peacefully; to report thattheirfamilies tend to do things together; to believe that theirfamilies careabout their education; and to have parents who said they were proud of them. The groups did not differ in regard to the distribution of their respectivereports on the extent to which their parents were critical of them.
The groups differed in their response distributions to 7 of the 10 drug-use
items. CBO youth were as likely to have friends who get high from drugs asto have friends who do not get high; however, gang youth were significantlymore likely to have friends who get high. However, although CBO youthwere less likely to report that their parents used legal drugs than were gangyouth, they were more likely than gang youth to report that their parentsdrank alcohol, although CBO youth were more likely to report light con-sumption. However, the groups did not differ in regard to the distributions oftheir reports of parental smoking or use of illegal drugs and drug type, if used. CBO youth were more likely than gang youth to think badly of people whouse drugs, and the groups differed in the substances they regarded as a drug(with more CBO youth than gang youth believing that heroin, crack, andcrack cocaine were drugs) and where they first encountered illegal drugs(with gang youth more likely to have seen such use throughout their neigh-borhood and CBO youth on the specific street on which they lived).
There was no difference between the two groups in regard to the peer-
relations item pertaining to having a friend to talk to. However, CBO youthwere more likely to report that they settled disputes with friends through talk-ing out their differences; gang youth were more likely to endorse the use ofguns or fists to settle disputes.
The groups differed in respect to all five items relating to school/education.
The CBO youth more often reported that they worked best by themselves onschool assignments and that they did theirhomework assignments. Theseyouth also had parents who were more likely to attend parent-teacher confer-ences. More CBO youth perceived their schools as having a diversified fac-ulty than was the case in regard to the perceptions of teacher diversity foundamong the gang youth. Finally, the groups differed in regard to their reasonsfor going to school. More gang youth attended because of parental demandsor to socialize with their friends. More CBO youth attended to get a better jobor to go to college.
Both of the two substantive items relating to religious activities/religiosity
were associated with significant differences between the groups. Fewer gangyouth were likely to attend church than was the case for the CBO youth. TheCBO youth were more likely than the gang youth to belong to a youth groupat church.
Race was a more frequently reported attribute of a friend for the gang
youth than for the CBO youth. The groups did not differ in regard to the termthat they used to refer to their race. Most youth labeled themselves either“Black” or “African American.”
The groups differed on one of the items related to neighborhood/safety.
More CBO youth saw their neighborhood as somewhat safer than was thecase with the gang youth. The groups did not differ in regard to their percep-tions of school safety, although slightly more CBO youth than gang youthreported that their family and friends had not experienced danger or violence.
The groups also did not differ in the frequencies of their reports of having
had sexual relations or in the age distribution associated with the beginningof this behavior. A majority of youth in both groups had had sex and hadfriends who began to be sexually active at age 15 or younger. More CBOyouth than gang youth reported using contraception.
Finally, the groups differed in regard to all items in the role models/
confidants category. The most frequent role model for the gang youth was a“rapper.” Fewer gang youth than CBO youth reported having a person towhom they could go foradvice, and more CBO youth than gang youthbelieved that the gender of a role model was not important.
In sum, of the item comparisons, the African American adolescent male
gang and CBO members differed substantially in regard to their family life,
JOURNAL OF ADOLESCENT RESEARCH / November 2003
TABLE 1: Means and Standard Deviations for Gang and Community-Based Or- ganization (CBO) Youth for the Positive Individual and Social Behavior Scores for the Nine Attribute Categories of the Interview
experience with and attitudes toward drugs, the ways in which disputes aresettled with friends and family members, orientation to school, religiosity,the practices of safe sex, and the nature of their role models. On the otherhand, there were similarities across the groups in respect to selected aspectsof social support from friends, racial identity, and sexual activity. That is,there was some overlap between the groups in regard to positive developmen-tal outcomes. It is useful to consider the results of interrelational analysescombining data across interview items that enable the identification of pro-files of responses that may be linked, at least, theoretically (e.g., as inMcAdoo, 1995, 1998, 1999; Spencer, 1990, 1995, 1999), to positive devel-opment among both the gang and CBO groups. Discriminating Between the Two Groups of Youth Based on Positive Individual and Social Characteristics
Table 1 presents means, standard deviations, and ranges for the two
groups of youth in regard to their respective scores for positive individual andsocial behavior characteristics for the nine substantive dimensions assessedin the interview. The tests of equality of group means appear in Table 2. Theresults in Table 2 suggest that the two groups of youth differ in each of thenine dimensions understudy. The Wilks’s Lambda scores indicate that themean differences account for 4.2% to 35.4% of the variance of the ninedimensions.
To determine the magnitude of the separability of the two groups based on
these nine categories, we performed a discriminant analysis. Specifically, we
TABLE 2: The Tests of Equality of Group Means TABLE 3: Classification Results From Discriminant Analysis Between the Gang and the Community-Based Organization (CBO) Groups Using Fisher’s Discrimination Criterion Estimating Prior Probabilities From the Sam- ple Sizes Using Separate-Group Covariance Matrices
discriminated between the two youth groups using Fisher’s discriminationcriterion, estimating prior probabilities from the sample sizes and usingseparate-group covariance matrices. The classification table appears inTable 3. The analysis indicates that, based on the nine categories, a total of81.1% of the original group memberships can be correctly classified. Inter-estingly, African American CBO members are more homogeneous thanAfrican American gang members. Specifically, 88% of the African Ameri-can CBO members are correctly classified, whereas only 73.3% of the Afri-can American gang members are correctly classified. In this regard andunderscoring the view that all young people have attributes pertinent to pos-itive youth development, it should be noted that 26.7% of the gang youthwere predicted to be in the CBO group. Finally, the overall Wilks’s Lambdawas .58. This significant value (χ2 = 48.21; df = 9; p < .01) indicates that 42%of the total variability in the data can be accounted for by the discriminantfunction.
JOURNAL OF ADOLESCENT RESEARCH / November 2003
TABLE 4: Standardized Canonical Discriminant Function Coefficients: The Mar- ginal Impact That a Category Has on the Discriminant Function While Holding the Other Categories Fixed
We next asked which of the nine categories is the most powerful for the
group discrimination. Table 4 displays the standardized discriminant func-tion coefficients. These coefficients can be interpreted in a fashion parallel toregression coefficients. Each coefficient indicates the marginal impact that acategory has on the discriminant function while holding the other categoriesfixed. The results in Table 4 show that the most potent discriminating variableis role model orientation. Specifically, role models most often identified byCBO youth are parents, family, and teachers, whereas role models most oftenidentified by gang youth are gangsters and rappers. The next two most pow-erful discriminating variables are sexual activity (CBO members are morelikely to use protection during sex) and neighborhood safety (gang membersreport that they live in less safe neighborhoods). Interrelational Analyses
To explore the covariation between domains of positive attributes, the
interrelational structure of the positive attributes in the two groups were com-pared. The correlation matrix appears in Table 5. The above-diagonal scoresin this matrix display the correlations for the gang youth; the below-diagonalscores display the correlations for the CBO youth. The correlations in Table 5suggest that many of the responses of the gang youth are strongly correlated. Only 4 out of 36 correlations are not significant. The responses of the CBOyouth show a different pattern. Less than half of the correlations are signifi-cant. All of the correlations between sexual activity and other variables aresignificant in the gang responses, and only one of these correlations, the onewith neighborhood safety, is significant in the CBO responses. TABLE 5: Correlations Among the Positive Attributes Between the Gang Youth (above diagonal) and the Community-Based Organization (CBO) Youth (below diagonal)
JOURNAL OF ADOLESCENT RESEARCH / November 2003
Figure 1. Mirrored bar chart of the frequency distributions of the total positive attribute scores for the gang and community-based organization (CBO) African American male adolescents.
In essence, the data summarized in Table 5 reflect greater behavioral cou-
pling (covariation) forgang than forCBO youth. However, otherdata indi-cate that the dispersion of these scores is comparable across the two groups.
Figure 1 shows that there is considerable overlap in the frequency distri-
bution of total positive scores for the two groups. In particular, in the areabetween 20 and 30 on the total positive score scale, there is a considerablenumberof gang youth. However, in the area below 15 on the total positivescore scale, the numberof CBO youth is 3 and the numberof gang youthis 25.
In addition to the contrasts between the gang and the CBO youth in regard
to the distribution and interrelation of attributes associated with positivedevelopment, there are some youth who possess more positive scores thanothers. Table 6 shows the percentage of gang youth who are above the meanfor the CBO youth in positive scores for the nine attributes. These percent-ages range from 4.4% (for sexual activity) to 53.3% (for racial identity) andaverage 27.6% across the nine attribute categories. Although the majority ofgang youth are below the average of CBO adolescents in regard to the posses-sion of positive attributes, a substantial minority of the gang youth exceeds
TABLE 6: Percentage of Gang Youth (N = 45) Who Are Above the Mean for the Community-Based Organization (CBO) Youth (N = 50) in Positive Scores for the Nine Attribute Categories
the average total positive score of the CBO youth in regard to possessingattributes that may afford the basis of positive youth development.
Are the gang youth who are above the mean in total number of scores for
positive development on one or more attributes studied in OTO likely to bethe adolescents who ultimately overcome the odds and show positive devel-opmental outcomes? How will theirdevelopment compare to CBO youthwho, as a group, tend to have more positive attributes in regard to the nineattribute categories assessed in this research? Although the data to ade-quately address these questions will require the longitudinal data that will becollected as part of the future waves of testing planned for the OTO project,these present results suggest that, consistent with theoretical ideas about theresources for healthy development that are present among all youth (Benson,1997; McAdoo, 1995, 1999; Spencer, 1990, 1995), both gang and CBO youthshave some commonality in regard to features of positive development. DISCUSSION
Developmental systems models pertinent to understanding the character-
istics of person-context regulations promoting adaptive modes among ado-lescents of colorand theirfamilies, especially in economically poorcommu-nities (e.g., Adams & Marshall, 1996; Grotevant, 1998; Luster & McAdoo,1996; Lutharet al., 2000; McAdoo, 1999; Spencer, 1990, 1999; Spenceret al., 1997), emphasize that all youth living in these settings have individualand contextual assets that may be used to promote positive behavior anddevelopment. This focus on positive youth development among pooryouth
JOURNAL OF ADOLESCENT RESEARCH / November 2003
of colorleads to an interest in identifying the individual and ecological char-acteristics that may result in healthy outcomes even among individualsinvolved in many of the high-risk behaviors linked to poverty (Dryfoos,1990; Huston, 1991; McAdoo, 1995, 1998, 1999; Spencer, 1990, 1995,1999). The study from which the present report is derived is framed by thisperspective about positive youth development.
Independent of the longitudinal data set, of which the data will eventually
be a part, the present cross-sectional data indicate that the African Americanmale adolescent gang and CBO youth are significantly distinct in regard toattributes of positive development associated with family, drug use, peers,school orientation, religiosity, racial identity, neighborhood and safety char-acteristics, sexual activity, and role models. Nevertheless, and underscoringthe theoretical basis of this research that all young people have the potentialfor the individual and ecological characteristics associated with positiveperson-context regulations, the discrimination between the groups is notabsolute. There is some overlap in the distribution of positive characteristicsacross the two groups of adolescents. Moreover, and consistent with thescholarship of McAdoo (1995, 1998, 1999; Luster & McAdoo, 1994, 1996),Spencer(1990, 1995, 1999; Spenceret al., 1997) and Taylor(1990, 1993,1996, 2000), there may be characteristics of the gang youth that could be cap-italized on by strength-building interventions.
First, scores for the nine attribute categories are more often interrelated
significantly among the gang youth than among the CBO youth. This greatercoupling of attributes among the gang youth may represent either a potentialresource for interventions into the developmental system or a challenge toefforts aimed at promoting positive development among gang youth.2 On onehand, it may be that, in a highly coupled system, targeting one domain ofbehaviors (e.g., one that was most positive) could make it more likely to havea health-promoting impact on other, less positive, but still coupled, attributes. On the other hand, interventionists targeting particular domains of behaviorcould have more difficulty affecting positive change because coupling couldact as a strong countereffect to an intervention.
The present data set does not provide information that allows discrim-
ination between these alternative ideas about the import of coupling forintervention with gang youth. It may be that both of the individual character-istics (e.g., related to integrated vs. differentiated cognitive styles or tempera-mental attributes such as adaptability) and features of the context (e.g., gangrules that foster nonindividualized behavioral repertoires) provide sources ofcoupling. A goal of ourfuture assessments of the gang and the CBO youthwill be to collect more information about such individual and contextualcharacteristics, to appraise the role of such variables in providing bases of
coupling, and to understand links between coupled characteristics and inter-vention effectiveness.
Second, despite the magnitude ormeaning of coupling among the gang
youth, it may be important that, across the nine attributes studied in thisresearch, about one quarter of the gang youth have a total attribute score thatis more positive than the average total attribute scores among the CBO youth. This finding suggests that there may be a rich pool of potential targets forinterventions aimed at capitalizing on the positive attributes of gang youth toenable them to enhance the overall probability of their health and thriving.
In sum, the data from this first wave of assessment of the OTO sample
have both basic and applied significance. Substantively, the data indicate thatthere are a subset of gang youth who, despite being embedded in a behavioraland social milieu marked by risks (e.g., gang violence, drugs, and poor famil-ial support), transcend the ambient problems of poverty and racism (i.e.,show some characteristics of positive development). This finding under-scores the theoretical notion that all groups of youth may possess the basesforpositive development (e.g., Benson, 1997; Damon, 1997; Leffert et al.,1998; McAdoo, 1995, 1999; Spencer, 1990, 1995, 1999; Taylor, 1996, 2000)and suggests that, if individual and ecological conditions are appropriatelyintegrated among gang youth, their positive development may be enhanced. Oursubsequent longitudinal analyses will assess, independent of similarityto the CBO youth, if there exist among gang youth particular profiles of indi-vidual and ecological resources (i.e., assets; Benson, 1997; Leffert et al.,1998) that are linked over time to favorable developmental outcomes, such asones associated with psychological and social thriving (Scales et al., 2000).
Certainly, such longitudinal data will have significance for application.
However, the present, unitemporal data also have import for policies and pro-grams. The fact that many gang youth have attributes of positive developmentabove the mean of the CBO youth suggests that there are useful targets foractions aimed at promoting positive development (e.g., Spencer, 1999;Spenceret al., 1997; Taylor, 1996). Such interventions may be conductedwith greater empirical support for the belief that one may actualize the poten-tial for positive development among African American adolescent male gangmembers.
JOURNAL OF ADOLESCENT RESEARCH / November 2003
APPENDIX The Interview Protocol
Where were your parents born?A. _______________________________________B. _______________________________________C. Don’t know
Where were you born?A. _______________________________________
Do you belong to a youth group at church?A. YesB. No
Do you have a friend that you can talk to about your problems?A. YesB. NoC. Sometimes
How do you settle disputes between your friends (crew, posse, set)?A. Talk it outB. Fight w/ fistsC. GunsD. KnivesE. Other
9. On school assignments, you work best with:
A. YourselfB. Other studentsC. Your friends
12. How would you rate your relationship with your parents?
13. Will your parents support you regardless of what you do in life?
A. SupportB. Will not supportC. Never support
14. How does your family settle disagreements?
A. Physical (violence)B. Verbal (violence)C. Verbal (peaceful)D. Don’t talkE. Do nothing
15. What does your family do together? (play sports, eat meals together, etc.)
A. Everything togetherB. Lots of things togetherC. Some things togetherD. Never together
16. Does your family care about your education?
17. How far did your parents go in school?
A. Elementary schoolB. Middle schoolC. High schoolD. Other school (trade)E. No school
19. Do they (parents) tell you they are proud of you?
JOURNAL OF ADOLESCENT RESEARCH / November 2003
a. Heavy (more than 4 drinks per day)b. Moderate (2 drinks per day)c. Light (occasional, once a week)
22. Do your parents use legal drugs? (like aspirin, blood pressure medication)
23. Do they (parents) use illegal drugs? (heroin, cocaine, weed)
Drug type:a. Cocaineb. Heroinc. Crack cocained. Marijuana
24. Do your parents go to parent/teacher conferences?
25. What name do you call yourself racially?
A. BlackB. AfricanC. African AmericanD. Other
26. Does it matter what race a person is that you hang with?
27. What do you think of people who use drugs?
A. Good things (character)B. Bad thingsC. Don’t thinkD. Don’t care
A. MarijuanaB. HeroinC. CocaineD. Crack cocaine
29. Where did you first see illegal drugs?
A. In my homeB. On my streetC. In my neighborhoodD. With my friendsE. My relatives outside my home
30. Does your school have diversified teachers? (Black, Latino, Asian,
Native American, other race groups)A. DiverseB. Some diverseC. All the same raceD. Don’t know
A. To get a better jobB. Because my parents demand that I goC. To socialize and hang out with my friendsD. To be able to go to college
35. Do you use protection when having sexual relations?
36. At what age did most of your friends begin having sex?
37. Whom do you look up to as a role model?
(a person you admire and would like to be like)A. Parents
38. Is there someone you can go to for advice?
JOURNAL OF ADOLESCENT RESEARCH / November 2003
39. Does it matter if they are male or female?
A. SafeB. Somewhat safeC. DangerousD. Very dangerousE. Don’t know
41. Have your family/friends experienced danger and violence?
1. Tables presenting the results of the χ2 analyses foreach item may be obtained from the
2. The authors are grateful to two anonymous reviewers for suggesting this interpretation. REFERENCES
Adams, G. R. (1997). Identity: A brief critique of a cybernetic model. Journal of AdolescentResearch, 12(3), 358-362.
Adams, G. R., Day, T., Dyk, P., Frede, E., & Rogers, D. (1992). On the dialectics of pubescence
and psychosocial development. Journal of Early Adolescence, 12, 348-365.
Adams, G. R., & Marshall, S. K. (1996). A developmental social psychology of identity: Under-
standing the person-in-context. Journal of Adolescence, 19, 429-442.
Allison, K. W. (1993). Adolescents living in “non-family” and alternative settings. In R. M.
Lerner (Ed.), Early adolescence: Perspectives on research, policy, and intervention (pp. 37-50). Hillsdale, NJ: Lawrence Erlbaum.
Benson, P. (1997). All kids are our kids: What communities must do to raise caring and responsi-ble children and adolescents. San Francisco: Jossey-Bass.
Damon, W. (1997). The youth charter: Howcommunities can work together to raise standardsfor all our children. New York: Free Press.
Dryfoos, J. G. (1990). Adolescents at risk: Prevalence and prevention. New York: Oxford Uni-
Fisher, C. B., Jackson, J. F., & Villarruel, F. A. (1998). The study of African American and Latin
American children and youth. In R. M. Lerner (Ed.), Handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed., pp. 1145-1207). New York: John Wiley.
Grotevant, H. (1998). Adolescent development in family contexts. In W. Damon & N. Eisenberg
(Eds.), Handbook of child psychology: Vol. 3. Social, emotional, and personality develop-ment (pp. 1097-1149). New York: John Wiley.
Hamburg, D. A. (1992). Today’s children: Creating a future for a generation in crisis. New York:
Huston, A. C. (Ed.). (1991). Children in poverty: Child development and public policy. Cam-
bridge, MA: Cambridge University Press.
Jarrett, R. L. (1998). African American children, families, and neighborhoods: Qualitative con-
tributions to understanding developmental pathways. Applied Developmental Science, 2(1),2-16.
Joe, K. (1993). Issues in accessing and studying ethnic youth gangs. Journal of Gang Research,
Leffert, N., Benson, P., Scales, P., Sharma, A., Drake, D., & Blyth, D. (1998). Developmental
assets: Measurement and prediction of risk behaviors among adolescents. Applied Develop-mental Science, 2, 209-230.
Lerner, R. M. (1995). America’s youth in crisis: Challenges and options for programs and poli-cies. Thousand Oaks, CA: Sage.
Lerner, R. M. (1996). Relative plasticity, integration, temporality, and diversity in human devel-
opment: A developmental contextual perspective about theory, process, and method. Devel-opmental Psychology, 32(4), 781-786.
Lerner, R. M. (1998). Theories of human development: Contemporary perspectives. In R. M.
Lerner (Ed.), Theoretical models of human development, Vol. 1. Handbook of child psychol-ogy (5th ed., pp. 1-24). New York: John Wiley.
Lerner, R. M. (2002). Concepts and theories of human development (3rd ed.). Mahwah, NJ: Law-
Lerner, R. M., Freund, A. M., DeStefanis, I., & Habermas, T. (2001). Understanding develop-
mental regulation in adolescence: The use of the selection, optimization, and compensationmodel. Human Development, 44, 29-50.
Lerner, R. M., & Galambos, N. L. (1998). Adolescent development: Challenges and opportuni-
ties for research, programs, and policies. In J. T. Spence (Ed.), Annual reviewof psychology,Vol. 49 (pp. 413-446). Palo Alto, CA: Annual Reviews.
Lerner, R. M., Sparks, E., & McCubbin, L. (1999). Family diversity and family policy:Strengthening families for America’s children. Norwell, MA: Kluwer.
Luster, T., & McAdoo, H. P. (1994). Factors related to the achievement and adjustment of young
African American children. Child Development, 65, 1080-1094.
Luster, T., & McAdoo, H. P. (1996). Family and child influences on educational attainment: A
secondary analysis of the High/Scope Perry Preschool data. Developmental Psychology,32(1), 26-39.
Luthar, S. S., Cicchetti, D., & Becker, B. (2000). The construct of resilience: A critical evaluation
and guidelines for future work. Child Development, 71(3), 543-562.
McAdoo, H. P. (1995). Stress levels, family help patterns, and religiosity in middle- and
working-class African American single mothers. Journal of Black Psychology, 21, 424-449.
McAdoo, H. P. (1998). African American families: Strength and realities. In H. C. McCubbin, E.
Thompson, & J. Futrell (Eds.), Resiliency in ethnic minority families: African American fam-ilies (pp. 17-30). Thousand Oaks, CA: Sage.
McAdoo, H. P. (1999). Diverse children of color. In H. E. Fitzgerald, B. M. Lester, & B. S.
Zuckerman (Eds.), Children of color: Research, health, and policy issues (pp. 205-218). NewYork: Garland.
McLoyd, V. C. (1990). Minority children: Introduction to the special issue. Child Development,
JOURNAL OF ADOLESCENT RESEARCH / November 2003
McLoyd, V. C. (1998). Children in poverty: Development, public policy, and practice. In
W. Damon (Ed.), I. E. Sigel, & K. A. Renninger(Vol. Eds.), Handbook of psychology: Vol. 4. Child psychology in practice (5th ed., pp. 135-208). New York: John Wiley.
Mincy, R. B. (1994). Conclusions and implications. In R. B. Mincy (Ed.), Nurturing young blackmales: Challenges to agencies, programs, and social policy (pp. 187-203). Washington, DC:The Urban Institute Press.
Muuss, R. E. (1996). Theories of adolescence (6th ed.). New York: McGraw-Hill. Neely, D. E. (1997). The social reality of African American street gangs. Journal of GangResearch, 4(2), 37-46.
Patton, P. L. (1998). The gangstas in our midst. Urban Review, 30(1), 49-76. Scales, P., Benson, P., Leffert, N., & Blyth, D. A. (2000). The contribution of developmental
assets to the prediction of thriving among adolescents. Developmental Psychology, 22, 342-347.
Scales, P., & Leffert, N. (1999). Developmental assets: A synthesis of the scientific research onadolescent development. Minneapolis, MN: Search Institute.
Spencer, M. B. (1990). Development of minority children: An introduction. Child Development,
Spencer, M. B. (1995). Old issues and new theorizing about African American youth: A
phenomenological variant of ecological systems theory. In R. L. Taylor (Ed.), Black youth:Perspectives on their status in the United States (pp. 38-69). Westport, CT: Praeger.
Spencer, M. B. (1999). Social and cultural influences on school adjustment: The application of
an identity-focused cultural ecological perspective. Educational Psychologist, 34, 43-57.
Spencer, M. B., Dupree, D., & Hartmann, T. (1997). A phenomenological variant of ecological
systems theory (PVEST): A self-organization perspective in context. Development &Psychopathology, 9, 817-833.
Taylor, C. S. (1990). Dangerous society. East Lansing: Michigan State University Press. Taylor, C. S. (1993). Girls, gangs, women, and drugs. East Lansing: Michigan State University
Taylor, C. S. (1996, January). The unintended consequences of incarceration: Youth develop-ment, the juvenile corrections systems, and crime. Paper presented at the Vera Institute Con-ference, Harriman, NY.
Taylor, C. S. (2000). Youth gangs. In N. J. Smelser & P. B. Baltes (Eds.), International encyclope-dia of the social and behavioral sciences. Oxford: Elsevier.
Taylor, C. S., Villarruel, F. A., & Lerner, R. M. (1998). Overcoming the odds: Understandingsuccessful development among African American and Latino male adolescents. FinalReport: 1996-1998. Boston, MA: Boston College, CenterforChild, Family, & CommunityPartnership.
Valdez, A., & Kaplan, C. D. (1999). Reducing selection bias in the use of focus groups to investi-
gate hidden populations: The case of Mexican-American gang members from south Texas. Drugs & Society, 14(1-2), 209-224.
Werner, E. E., & Smith, R. S. (1977). Kauai’s children come of age. Honolulu: University of
Werner, E. E., & Smith, R. S. (1982). Vulnerable but invincible: A longitudinal study of resilientchildren and youth. New York: McGraw-Hill.
Werner, E. E., & Smith, R. S. (1992). Overcoming the odds: High risk children from birth toadulthood. Ithaca, NY: Cornell University Press. Carl S. Taylor is the director of Community and Youth Development Programs at theInstitute for Children, Youth, and Families and a professor of family and child ecology atMichigan State University. He has extensive experience in field research aimed at thereduction of violence involving American youth. He has worked with communities, foun-dations, and government agencies in understanding gangs, youth culture, and violence. Some of the organizations that he has worked with include the Guggenheim Foundation,the C.S. Mott Foundation, the FBI Academy, and the Children’s Defense Fund. He servesas the principal investigator in the Michigan Gang Research Project. He also serves onthe Michigan Juvenile Justice Committee and advises various projects concerning youththroughout America.Richard M. Lerner is the Bergstrom Chair in Applied Developmental Science at TuftsUniversity. A developmental psychologist, Lerner received a Ph.D. in 1971 from the CityUniversity of NewYork. Lerner is the author or editor of 45 books and more than 300scholarly articles and chapters. He edited Volume 1, Theoretical Models of Human De-velopment, for the 5th edition of the Handbook of Child Psychology. He is the foundingeditor of the Journal of Research on Adolescence and of Applied Developmental Sci-ence. He is known for his theory of and research about relations between life-span humandevelopment and contextual or ecological change. He has done foundational studies ofadolescents’relations with their peer, family, school, and community contexts, and he is aleader in the study of public policies and community-based programs aimed at the pro-motion of positive youth development.Alexander von Eye is currently chair of the Developmental Psychology unit at MichiganState University. A developmental methodologist, he received his Ph.D. in 1976 from theUniversity of Trier, Germany. He is a fellowof Division 5 of the American PsychologicalAssociation. Prior to joining Michigan State University, he held positions at the Univer-sities of Trier and Erlangen-Nuernberg and at the Max Planck Institute for Human De-velopment and Education, all in Germany, and at Penn State. He is the author or editor of14 books and more than 300 scholarly articles and chapters. He focuses his research onstatistical methods for the analysis of longitudinal data and categorical data. He is inter-ested in classification methods and conducts computer simulation studies. Substantively,he is interested in cognitive development, aging, and adolescent development—all from alife-span perspective.Deborah L. Bobek is the managing director of the Applied Developmental Science Insti-tute in the Eliot-Pearson Department of Child Development at Tufts University. She hasexperience working on research projects related to applied developmental science andhas authored and published pieces on this work and other work around positive youth de-velopment. She received her master’s degree in child development from the Eliot-PearsonDepartment of Child Development at Tufts University and her bachelor’s degree in psy-chology, economics, and Spanish, also from Tufts. She is a member of the Tufts AlumniAdmission Program. Prior to receiving her master’s degree, she was a research associateand analyst in the real estate group at Fidelity Investments in Boston. She received herChartered Financial Analyst designation in 1998.
JOURNAL OF ADOLESCENT RESEARCH / November 2003
Aida B. Balsano is from Bosnia and Herzegovina. After earning a B.A. in psychologyfrom Grinnell College, Iowa, she earned her M.A. in child development from the Eliot-Pearson Department of Child Development at Tufts University in May 2001. She has con-tinued with her studies at Eliot-Pearson as the first doctoral student accepted in the pro-gram as a Jacobs Fellow. She is in her third year of the doctoral program. As part of herinterest in community-university collaborations at an international level, she has initi-ated collaboration between the Eliot-Pearson Department of Child Development and theyouth-serving NGO sector in Bosnia and Herzegovina. Her research interests involve is-sues of parenting among refugee populations in the Boston metropolitan area.Elizabeth Dowling is a fourth-year doctoral student in the Eliot-Pearson Department ofChild Development at Tufts University. She coauthored a grant proposal to form a col-laboration between Tufts University and the University of Sarajevo aimed at exchangingknowledge about child development. She taught a class at Boston College on early child-hood development and coteaches a cognitive development class at Tufts. She is focusedon studying the assets within communities that support the development of a sense ofspirituality in children. She is most interested in looking at the ways that schools supportthe spiritual life of children and believes that the dialogue around the current school re-form effort is detrimental to learning and spiritual development. She believes that societyneeds to advocate to recreate schools to support spiritual development.Pamela M. Anderson is currently a doctoral student at Tufts University in the Eliot-Pearson Department of Child Development. She received her bachelor’s degree in psy-chology from the University of California, Santa Cruz, and her master’s degree in childdevelopment from Tufts University. She has had experience working on various researchprojects related to positive youth development and has coauthored articles and presentedresearch related to this work at conferences. Her current research interests involvestudying howperson-context relations influence development in young people. Spe-cifically, she is interested in howparticipating in wilderness activities/programs mightfoster and support positive development and spirituality in young people.
J Vet Intern Med 2003;17:304–310 Diazepam as a Treatment for Metronidazole Toxicosis in Dogs: A Retrospective Study of 21 Cases Jason Evans, Donald Levesque, Kim Knowles, Randy Longshore, and Scott PlummerThe currently recommended treatment for metronidazole toxicosis is drug discontinuation and supportive therapy. Reported recoverytimes are 1–2 weeks. The records of 21 dogs with metr
Oxidative Coupling of 17 -Estradiol: Inventory of Oligomer Products and Configuration Assignment of Atropoisomeric C4-Linked Biphenyl-Type Dimers and Trimers Alessandro Pezzella,* Liliana Lista, Alessandra Napolitano, and Marco d’Ischia Department of Organic Chemistry and Biochemistry, University of Naples “Federico II” Complesso Universitario Monte S. Angelo, Via Cinthia 4, I-80