The role of developmental assets in predicting academic achievement: A longitudinal study

The role of developmental assets in predicting academic achievement: A longitudinal study

ARTICLE IN PRESS Journal of Adolescence Journal of Adolescence 29 (2006) 691–708 www.elsevier.com/locate/jado The role of developmental assets in pr...

235KB Sizes 3 Downloads 37 Views

ARTICLE IN PRESS

Journal of Adolescence Journal of Adolescence 29 (2006) 691–708 www.elsevier.com/locate/jado

The role of developmental assets in predicting academic achievement: A longitudinal study Peter C. Scalesa,, Peter L. Bensona, Eugene C. Roehlkepartaina, Arturo Sesma Jr.a, Manfred van Dulmenb a

Office of the President, Search Institute, 615 First Avenue NE, Suite 125, Minneapolis, MN 55413 b Department of Psychology, Kent State University, PO Box 5190, Kent, OH 44242

Abstract A sample of 370 students in the 7th–9th grades in 1998 was followed for 3 years through the 10th–12th grades in order to investigate the relation of ‘‘developmental assets’’—positive relationships, opportunities, skills, values, and self-perceptions—to academic achievement over time, using actual GPA as the key outcome variable. The greater the number of developmental assets students reported in the 7th–9th grades, the higher their GPA in the 10th–12th grades. Students who stayed stable or increased in their asset levels had significantly higher GPAs in 2001 than students whose asset levels decreased. Increases in assets were significantly associated with increases in GPA. Experiencing in 1998 clusters of specific assets increased by 2–3 times the odds of students having a B+ or higher GPA in 2001. The results offer promising evidence that a broad focus on building the developmental nutrients in young people’s lives may contribute to academic success. r 2005 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved. Keywords: Adolescent development; Developmental assets; Academic achievement; Positive youth development; School reform

Corresponding author. Search Institute, c/o 940 Chestnut Ridge Road, Manchester, MO 63021, USA. Tel./fax: +1 636 225 2112. E-mail address: [email protected] (P.C. Scales).

0140-1971/$30.00 r 2005 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.adolescence.2005.09.001

ARTICLE IN PRESS 692

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

Introduction A vast amount of studies has been conducted, model reform approaches tried, and heated political dialogue carried on in an effort to identify successful ways of reforming schools to promote achievement and lessen achievement inequities across different groups of students. Recent school reform, with some notable exceptions (Bredekamp & Copple, 1997; National Middle School Association, 2003) has focused almost exclusively on the goal of raising students’ standardized test scores, but generally without a larger vision of the schools’ role in promoting positive youth development more broadly (Oakes, Quartz, Ryan, & Lipton, 2000). Yet implementation of developmentally based ‘‘best practices’’ such as interdisciplinary team teaching, co-operative learning, and heterogeneous ability grouping has been associated with better school climate and student achievement, among other positive outcomes (Fellner et al., 1997).

Human development as an achievement strategy Weissberg and O’Brien (2004) describe the ‘‘broad mission’’ of schools as developing young people who are ‘‘knowledgeable, responsible, healthy, caring, connected, and contributing’’ (p. 87). They point to research suggesting that an integrated combination of social, emotional, and academic learning is the most effective means to achieve that developmental goal. Students’ school success may be viewed theoretically as a result of a complex interplay among numerous factors reflecting multiple levels of young people’s ecology (Bronfenbrenner, 1979). School success is promoted when developmental nutrients: Provide caring and supportive relationships in the school community; increase student motivation and engagement; increase the value that students attach to education; increase the effectiveness of students’ study habits; strengthen social norms and expectations that promote achievement; and increase parent involvement and student attendance (Starkman, Scales, & Roberts, 1999). Research shows numerous developmental influences playing a role in school success, including: family support (Gutman, Sameroff, & Eccles, 2002; Petit, Bates, & Dodge, 1997; Steinberg, 2001); relationships with non-family adults (Fletcher, Newsome, Nickerson, & Bazley, 2001; Wenz-Gross, Siperstein, Untch, & Widaman, 1997; caring school climate (Roeser, Midgely, & Urdan, 1996); providing children opportunities to feel useful, such as through service-learning (Araque, 2002; Billig, 2004); fairness of school discipline policies (Catterall, 1998); high expectations (Schmidt & Padilla, 2003); positive peer influence (Bagwell, Schmidt, Newcomb, & Bukowski, 2001; Mounts & Steinberg, 1995); participation in co-curricular and after-school programs (Barber & Eccles, 1997; Hofferth & Sandberg, 2001; Mahoney, Cairns, & Farmer, 2003; NICHD, 2004); achievement motivation and school engagement (Jessor, VanDen Bos, Vanderryn, Costa, & Turbin, 1995; Shiner, 2000); and social competencies (Arroyo & Zigler, 1995; Malecki & Elliot, 2002). Developmentalists have long noted that, as Ryan, Stiller, and Lynch (1994) put it, school is as much an interpersonal as a cognitive enterprise. Greenberg et al. (2003) reviewed a wide range of evidence that suggests the most effective school-based prevention and youth development approaches are those that ‘‘enhance students’ personal and social assets’’ and improve the school–community environment (p. 467). Effective approaches try simultaneously across multiple

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

693

contexts to build students’ health, character, citizenship, school orientation, and academic performance. A recent example of the impact of such integrative approaches on achievement is the study of Hanson, Austin, and Lee-Bayha (2003) of nearly 1700 California public schools with students in grades 7, 9, and 11. Schools where successively higher proportions of students reported experiencing ‘‘resilience assets’’ such as caring relationships with and high expectations from teachers, parents, community adults, and peers, and meaningful opportunities to participate in schools and communities (such as through service-learning), had successively higher standardized achievement test scores. Fellner et al. (1997) also concluded that attempts to change the school ecology were less successful when each of the recommended change strategies was seen as separable from the others, and more successful when each was seen as linked with the others in a grand whole. The efficacy of a holistic approach to school reform is consistent with developmental theory and research that the most significant positive youth outcomes are likely when individual and collective actions attempt to strengthen multiple developmental nutrients simultaneously across multiple contexts (Bronfenbrenner & Morris, 1998; Lerner, Wertlieb, & Jacobs, 2003). Positive academic outcomes, like all developmental outcomes, occur as a result of successful individual-context relations (Lerner, Lerner, De Stefanis, & Apfel, 2001) in which persons are embedded in and actively shape multiple, interconnected ecologies. Weissberg, Kumpfer, and Seligman (2003) note that several strategies repeatedly appear in reports of successful efforts. These include building students’ social–emotional learning repertoire, providing frequent opportunities for student participation, such as through community service, fostering caring, supportive relationships among students, teachers, and parents, and consistently rewarding positive social, health, and academic behaviour through school–parent–community collaborations. These principles, and numerous specific positive developmental influences such as those reviewed above, are reflected in the framework of Developmental AssetsTM (Benson, 1997). Rather than narrowly focusing on preventing the problems of adolescents (e.g. school dropout, unprotected sexual activity, delinquency, substance use), this framework identifies the kinds of positive connections and qualities necessary for healthy growth. Theoretically, young people who experience high levels of ‘‘developmental assets’’ should enjoy more positive developmental outcomes, such as achieving better in school, than students who experience fewer assets. Assets are defined as important relationships, skills, opportunities and values that help guide adolescents away from risk behaviours, foster resilience, and promote thriving. As displayed in Table 1, 40 developmental assets have been identified, arrayed across eight conceptually coherent categories. External assets comprise a set of experiences and relationships across multiple contexts of the youth’s life that adults (and peers) provide for young people: support; empowerment; boundaries and expectations; and constructive use of time. Internal assets comprise a set of individual qualities—values, skills, and self-perceptions—thought to help the young person become effectively self-regulating: commitment to learning; positive values; social competencies and positive identity. The asset framework is the conceptual progeny of a number of rich empirical traditions, from normative child and adolescent development to more applied areas of inquiry, such as prevention science and resilience research. Canvassing across these research areas, Benson and colleagues identified assets that contributed to at least one of three criteria: (1) reduction in risk behaviours;

ARTICLE IN PRESS 694

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

Table 1 Search Institute’s 40 developmental assets External assets Support

1. Family support—family life provides high levels of love and support 2. Positive family communication—young person and her or his parent(s) communicate positively, and young person is willing to seek advice and counsel from parents 3. Other adult relationships—young person receives support from three or more non-parent adults 4. Caring neighbourhood—young person experiences caring neighbours 5. Caring school climate—school provides a caring, encouraging environment 6. Parent involvement in schooling—parent(s) are actively involved in helping young person succeed in school

Empowerment

7. Community values youth—young person perceives that adults in the community value youth 8. Youth as resources—young people are given useful roles in the community 9. Service to others—young person serves in the community one hour or more per week 10. Safety—young person feels safe at home, at school, and in the neighbourhood

Boundaries and expectations

11. Family boundaries—family has clear rules and consequences and monitors the young person’s whereabouts 12. School boundaries—school provides clear rules and consequences 13. Neighbourhood boundaries—neighbours take responsibility for monitoring young people’s behaviour 14. Adult role models—parent(s) and other adults model positive, responsible behaviour 15. Positive peer influence—young person’s best friends model responsible behaviour 16. High expectations—both parent(s) and teachers encourage the young person to do well

Constructive use of time

17. Creative activities—young person spends three or more hours per week in lessons or practice in music, theater, or other arts 18. Youth programs—young person spends three or more hours per week in sports, clubs, or organizations at school and/or in the community 19. Religious community—young person spends one or more hours per week in activities in a religious institution 20. Time at home—young person is out with friends ‘‘with nothing special to do’’ two or fewer nights per week

Internal assets Commitment to learning

21. Achievement motivation—young person is motivated to do well in school 22. School engagement—young person is actively engaged in learning 23. Homework—young person reports doing at least one hour of homework every school day 24. Bonding to school—young person cares about her or his school

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

695

Table 1 (continued ) 25. Reading for pleasure—young person reads for pleasure three or more hours per week Positive values

26. Caring—young person places high value on helping other people 27. Equality and social justice—young person places high value on promoting equality and reducing hunger and poverty 28. Integrity—young person acts on convictions and stands up for her or his beliefs 29. Honesty—young person ‘‘tells the truth even when it is not easy’’ 30. Responsibility—young person accepts and takes personal responsibility 31. Restraint—young person believes it is important not to be sexually active or to use alcohol or other drugs

Social competencies

32. Planning and decision making—young person knows how to plan ahead and make choices 33. Interpersonal competence—young person has empathy, sensitivity, and friendship skills 34. Cultural competence—young person has knowledge of and comfort with people of different cultural/racial/ethnic backgrounds 35. Resistance skills—young person can resist negative peer pressure and dangerous situations 36. Peaceful conflict resolution—young person seeks to resolve conflict nonviolently

Positive identity

37. Personal power—young person feels he or she has control over ‘‘things that happen to me’’ 38. Self-esteem—young person reports having a high self-esteem 39. Sense of purpose—young person reports that ‘‘my life has a purpose’’ 40. Positive view of personal future—young person is optimistic about her or his personal future

(2) promotion of positive behaviours, and (3) fostering of resilience, or succeeding developmentally despite experiencing adversity. As such, the assets reflect fundamental developmental processes of connection, competencies, support, and efficacy (Benson, Scales, & Mannes, 2003; Scales & Leffert, 2004). Studies have shown that youth who report relatively more assets are less likely to report engaging in risk behaviour patterns, and more likely to report engaging in positive, socially constructive behaviours, including doing well in school (Leffert et al., 1998; Scales, Benson, Leffert, & Blyth, 2000; Taylor et al., 2003). This finding is the converse of one of the most robust and hardy findings in the psychological literature: the cumulative risk gradient (Sameroff, 1999). The concept of cumulative risk grew out of two consistent results: risk factors often do not occur in isolation, and the accumulation of many risk factors, not just one, acts to thwart developmental progress. Thus, what is typically depicted is a monotonic function of increasing risk factors with the diminution of adaptation with each successive level of risk. Few factors moderate this linear function, as it has been documented across age, gender, race/ethnicity, SES level, and

ARTICLE IN PRESS 696

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

cross-culturally (Rutter, 2000). A recent longitudinal study of poor, urban children followed from age 12 months to 16 years specifically tested for both linear and quadratic risk gradients, and found support only for the additive, not the exponential model of cumulative risk (Appleyard, Egeland, van Dulmen, & Sroufe, 2005). Similar findings have been reported about the operation of developmental assets on both risks and thriving (Benson, Roehlkepartain, & Sesma, 2004; Sesma & Roehlkepartain, 2003). A single asset or group of few assets is unlikely to set a youth on a positive developmental trajectory, but each successive level of asset-richness appears related to more positive youth development outcomes (discussed below). As displayed in Table 1, the 40 assets in the developmental assets framework touch on many potential direct and indirect predictors of academic achievement, including: a caring school climate, parent involvement in school, high expectations, adult role models, involvement in constructive after-school programs, achievement motivation, school engagement, social competencies, positive sense of the future, and more. Previous research on developmental assets and achievement Previous studies have reported promising connections between these specific 40 developmental assets and reports of academic achievement among both adolescents (reviewed in Scales & Leffert, 2004) and elementary-age students (reviewed in Scales, Sesma, & Bolstrom, 2004). For example, in two large aggregate surveys together involving more than 300,000 6th–12th graders from more than 500 US communities, the more assets young people report having, the more they report better school attendance, and getting mostly A’s (Benson, Scales, Leffert, & Roehlkepartain, 1999; Developmental assets: A portrait of your youth, 2001). The same significant positive relation of assets, attendance, and grades has been found in a study of low-income, urban, Hispanic and African-American high school students (Scales, Foster, Horst, Rutherford, & Pinto, 2005). Asset clusters that include time spent in youth programs and achievement motivation have been found to explain from 19%–31% of self-reported student grades across six racial/ethnic groups (African-American, American-Indian, Asian, Caucasian, Hispanic, and Multiracial), substantially more variance in grades than gender, race/ethnicity, grade level, and socioeconomic status explain (Scales et al., 2000). Purpose of this study The results of these studies point to the positive value of the assets framework as a theoretical and applied lens for understanding and promoting student achievement. However, these largely have been cross-sectional studies, and the school success variables studied most often have been measured by student self-reports. No study to date has three key features: (1) utilizes as independent variables the entire 40 assets outlined by Benson and colleagues, (2) employs a longitudinal design, and (3) connects students’ reports of assets to objective measures of school success such as actual grades. The present study has those features and so fills a significant gap in the literature. In this paper, we specifically address two critical questions: (1) Are developmental assets related to higher GPA in the same year? (2) Are developmental assets related to higher GPA over time?

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

697

Methods Sample The sample consisted of 370 students from St. Louis Park, MN, a suburb of Minneapolis, who were followed from 1997, when they were in the 6th–8th grades, to 2001, when they were in the 10th–12th grades. Actual GPAs were unavailable for considerable numbers of students in 1997, leaving an unacceptably small sample size. To use the full sample of 370, the present analysis focuses on the 3-year period from 1998–2001, or the years when students moved from 7th–9th grades, to 10th–12th grades. The sample was comprised of 40% 7th graders (N ¼ 148), 28% 8th graders (N ¼ 105), and 32% 9th graders (N ¼ 117). Fifty-three percent were female. The students largely were Caucasian (85%), with small numbers of Multiracial (7%) Asian (4%), African American (3%), and Hispanic students (2%). Seventy-five percent of the students lived with two parents. Parental education was used as a proxy for socioeconomic status. The families of these students included large proportions of Russian-Jewish immigrants for whom educational attainment is traditionally highly valued. Thus, unsurprisingly, the mean educational levels of students’ parents were quite high, suggesting a high SES: 60% of both fathers and mothers had either graduated from college or attended graduate or professional school. These demographic percentages were virtually unchanged from 1998–2001. A passive parental permission process was used. Students were included unless their parents or the students explicitly opted out of the study. A total of 85% of eligible students participated across the full 3 years of the longitudinal study.1 Measures Developmental assets: Developmental assets were measured through the Search Institute Profiles of Student Life: Attitudes and Behaviour survey (A&B). The 156-item survey includes measures of 40 developmental assets, 10 risk-taking behaviour patterns (e.g. problem alcohol use), 8 thriving indicators (e.g. self-reported grades), 5 developmental deficits (e.g. time alone at home), and standard demographic questions. The development and psychometric properties of the A&B survey have been discussed extensively elsewhere (Benson, Leffert, Scales, & Blyth, 1998; Leffert, et al., 1998). For the present study, only the 92 items measuring the 40 assets were utilized, along with demographic items. Most of the asset items are measured on a 5-point Likert scale (5 ¼ strongly agree, 1 ¼ strongly disagree). In general, students are scored as having an asset if 1

In 1997, all 760 students in the grades 6–8 cohort were eligible to participate, with 646 (85%) doing so. One year later, in 1998, 628 (97%) of those students completed surveys. Three years later, in 2001, 590 students remained in the cohort, now in 10th–12th grades, and 502 (78% of the original survey sample) completed surveys that year. From those cross-sectional totals, students were excluded if their surveys in all 3 years did not have matching student ID numbers or school records data, leaving a longitudinal sample of 370 (57% of the original survey sample from 4 years earlier). Data were available to compare refusers with participants in year 1 on grade, gender, race/ethnicity, and free lunch eligibility. There were no significant differences. The same data were also available to compare in 1998 and 2001 the longitudinal sample with students who completed surveys but were excluded because of non-matching IDs or missing data, and with the entire school cohort for those grades. In each of those 2 years, the longitudinal sample had a greater proportion of females, White students, and students from two-parent families, and a lesser proportion of students whose mothers had completed only a high school education.

ARTICLE IN PRESS 698

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

they average a 4 or above on the items comprising that asset. Individual asset measures were not used as independent variables. Rather, the categorical variable of students’ total number of assets was used, as reflected in their quartile level of assets (0–10, 11–20, 21–30, 31–40). The quartile distribution has been employed as the primary descriptive presentation of assets research since the introduction of the framework in 1990. The quartile variable has repeatedly been found to have significant positive linear associations with important developmental outcomes such as reported risk-taking behaviour patterns and indicators of thriving, including self-reported school grades (e.g. Benson et al., 1999; Leffert et al., 1998; Scales et al., 2000). Surveys were administered in the fall of 1997, 1998, and 2001 (i.e., a 1- and a 3-year interval). As noted earlier, most of the analyses in the present study focus on the 1998–2001 period. GPA: GPA was computed on all courses, not just core courses such as english, science, math, and social studies, and was calculated on a 12-point scale (A+ ¼ 12, A ¼ 11, A ¼ 10, etc.). Analysis plan Two primary kinds of analysis were employed. ANOVAs illuminated how mean GPA in high school varied by the four middle school asset quartiles. Growth curve analyses were used to show how increases in assets affected changes in GPA. In an additional analysis, asset factors derived from a different aggregate cross-sectional sample of more than 217,000 6th–12th graders were used as predictors, to determine how clusters of specific assets might be associated with later GPA.2 In that prior analysis, principal components analysis with Kaiser normalization was utilized. Inspection of the eigenvalues (i.e., greater than 1) suggested an 8-factor model, utilizing 34 of the 40 assets with factor loadings of .30 or greater. In addition, confirmatory factor analyses conducted for each of the six racial/ethnic groups of participants revealed that the eight-factor structure was an adequate representation of the 40 assets (detailed results, including factor loadings, available on request from authors). Table 2 displays the assets that loaded at .30 on each factor. Table 2 also displays reliabilities calculated on the St. Louis Park sample. All factors but connection to community had alpha reliabilities ranging from acceptable to good, but variations in young people’s interests and talents may well explain why this factor’s assets should not necessarily be expected to covary. These factors were utilized in a logistic regression analysis to investigate the odds that students experiencing those asset factors in 1998 would have a high (B+ or better GPA) in 2001. Logistic regression was used because we were interested in predicting group membership (high grades) from the asset clusters, reflecting the conceptual ideas behind the operationalization of a thriving construct such as GPA. A GPA of B+ or higher was chosen because it represents performance midway between all A’s (a perhaps unreasonable standard that would allow too few students to be described as having ‘‘high’’ grades) and all B’s (perhaps a too attainable standard that does not sufficiently differentiate those with truly outstanding performance). 2

The sample of 217,000 on which factor analysis was conducted was a cross-sectional aggregate of students from more than 300 US communities who completed the Search Institute A&B survey in the 1999–2000 academic year. Although not nationally representative, the sample was weighted to reflect Census figures for race/ethnicity and urban residence. The sample for the current study was not a subset of the large aggregate. Rather, the current sample came from one community surveyed in different academic years than the large sample, and was a longitudinal, not crosssectional sample.

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708 Table 2 Asset factors and alpha reliabilities for St. Louis Park sample Factor and assets

a

Family—4 assets Family support Positive family communication Parental involvement in school Family boundaries

.73

School—3 assets Caring school climate High expectations Bonding to school

.80

Positive identity—4 assets Self-esteem Sense of purpose Personal power Positive view of personal future

.83

Youth perceptions of community—3 assets Community values youth Caring neighbourhood Youth as resources

.80

Social competence—4 assets Interpersonal competence Planning and decision making Cultural competence Resistance skills

.71

Positive values—5 assets Equality and Social Justice Caring Integrity Responsibility Honesty

.80

Connection to community–6 assets Youth programs Religious community Service to others Creative activities Reading for pleasure

.51

Norms of responsibility–5 assets Positive peer influence Restraint Time at home Peaceful conflict resolution School engagement

.72

N ¼ 370.

699

ARTICLE IN PRESS 700

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

In addition, person-centered analyses were conducted to examine the relation of student asset trajectories to GPA trajectories. Students were counted as going ‘‘up’’ in assets or GPA if by 2001 they increased a half standard deviation (S.D.) of the mean asset level or GPA in 1998. They were counted as going down if they decreased a half S.D., and as relatively stable if they changed less than 7.5 S.D. over those 3 years. ANOVAs were then run to examine 2001 GPA differences among those asset trajectory groups. Finally, latent growth curve analysis also illuminated the relation of increases in assets to increases in GPA over time. Latent growth curve analyses have the advantage over traditional methods for studying change in that they capture individual variation in change and are more flexible in handling missing data than, for example, repeated measures ANOVA designs. Latent growth curve analyses were conducted in LISREL 8.53 using Maximum Likelihood estimation. Missing data was handled using multiple imputation with an Expectation Maximization algorithm.

Results Concurrent relation of developmental assets to GPA Consistent with previous research using student self-reports of grades, the more assets students reported, the greater their actual GPA the same year. An ANOVA showed that higher quartile levels of assets are related to significantly higher GPA, as displayed in Table 3 (F ð3; 325Þ ¼ 15:88, pp:0001). Longitudinal relation of developmental assets to GPA The more assets students reported in 1998, the higher students’ actual GPA in 2001. For example, the correlation between the total number of assets students experienced in 1998 and their GPA in 2001was .33 (N ¼ 326, pp:01). An ANOVA showed that 1998 asset quartile levels were strongly related to 2001 GPA (F ð3; 325Þ ¼ 15:78, pp:0001). As expected, when an ANCOVA was run controlling for the strong effect of earlier GPA on later GPA, the relationship between 1998 asset levels and 2001 GPA became weaker. Nevertheless, as Fig. 1 shows, the relation between 1998 asset level and 2001 GPA remained statistically significant (F ð3; 325Þ ¼ 2:54, pp:05). For

Table 3 Relation of 1998 asset levels to 1998 GPA 1998 asset level

N

Mean 1998 GPA

S.D.

0–10 11–20 21–40 31–40

26 134 121 45

6.23 7.98 9.00 9.71

3.17 2.61 2.13 1.37

N ¼ 326; GPA measured on a 12-point scale.

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

12

Relation of 1998 Assets to 2001 GPA

10 GPA

8

701

9.22

8.5

9.9

6.36

6 4 2 0

0 to 10

11 to 20

21 to 30

31 to 40

Asset Quartiles

Fig. 1. Relation of 1998 asset levels to 2001 GPA. Note: N ¼ 326. N’s and S.D.’s form lowest (0–10 assets) to highest (31–40 assets) asset quartile: 26(2.78), 134(2.44), 141(2.06), 45(1.80). Grades measured on a 12-piont scale.

example, the adjusted mean 2001 GPA for those who reported only 0–10 assets in 1998 was 6.36, versus an adjusted mean 2001 GPA of 9.9 for those with 31–40 assets in 1998. These results suggest that experiencing a higher level of assets does uniquely contribute to higher later levels of GPA. The difference each quartile increase in assets makes over 3 years is approximately equal to the difference between going from a C average to a B/B, B, and B+ average as the quartile level of assets increases from asset-depleted to average, above average, and asset-rich (see Fig. 1).

How much difference do specific asset clusters make in GPA? The above results suggest that the total number of assets has both a concurrent and longitudinal relation to GPA. But do some assets have a more significant relation to GPA than other assets? Using the empirically derived asset factors described above, logistic regression was conducted, using the eight 1998 asset factors to predict a high 2001GPA (i.e., a B+ average or better). Table 4 displays the results, showing that students with high levels of 1998 assets reflecting adherence to social norms of responsibility and connection to community were much more likely to have 2001 GPAs of B+ or higher, even after controlling for 1998 GPA levels. Table 4 shows that GPA in 1998 is the biggest predictor of whether a student achieved a high GPA 3 years later. For every higher point in 1998 GPA, students were four times more likely to be in the high GPA group in 2001. But clusters of 1998 assets also strongly improved the odds of being a B+ student in 2001. For example, for every higher point on the 1998 asset factor reflecting connection to community (youth programs, religious community, service to others, creative activities, reading for pleasure), students were three times more likely to be in the high GPA group in 2001. Moreover, for every higher point on the 1998 assets reflecting adherence to norms of responsibility (school engagement, positive peer influence, restraint, time spent at home, peaceful conflict resolution), students were two times more likely to be in the high GPA group in 2001.

ARTICLE IN PRESS 702

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

Table 4 Logistic regression: 1998 asset factors as predictors of 2001 GPA Predictor

B

SE B

eB

GPA—1998

1.47***

.18

4.32

Asset factors Family School Youth perceptions of community Connection to community Norms of responsibility Positive values Social competence Positive identity

.38 .10 .06 1.14** .80* .14 .47 .17

.39 .42 .37 .41 .40 .32 .39 .32

.69 1.10 .94 3.13 2.23 .87 .62 1.19

Constant

14.06

w2 ðdf Þ

256.73(9)

Note: eB ¼ exponentiated B. *po:05, **po:01, ***po:001.

Relation of change in assets to subsequent GPA We also looked at how changes in assets affected GPA in 2001. Students who remained stable or increased by at least .5 S.D. in their number of assets between 1998 and 2001 had significantly higher mean GPAs in 2001 than students who declined at least .5 S.D. in their number of assets (GPA of 9.12 vs. 8.51, t ¼ 4:90, df ¼ 306, pp:02). In addition, students who stayed stable or went up in their Commitment to Learning asset category scores by at least .5 S.D. between 1998 and 2001 had significantly higher 2001 GPAs than students who went down by at least .5 S.D. in Commitment to Learning across those years (GPA of 9.07 versus 8.31, t ¼ 8:31, df ¼ 353, pp:004).3 Latent growth curve analyses also showed that increases in assets over time were related to increases in GPA. More specifically, a growth curve model was analyzed (see Fig. 2). Results showed that the model represented the data well, w2 ð10Þ ¼ 15:50, po:01, RMSEA ¼ .04. Results further indicated that the level of assets at 1997 was associated with changes in assets from 1997 to 2001 (B ¼ :70, po:05). In addition, results showed that GPA at 1997 was associated with changes in assets from 1997 to 2001 (B ¼ :37, po:05). Change in the number of assets experienced had a small but significant relationship to change in GPA over the 4 years from 1997–2001 (.31 3

Students who went down in assets between 1998 and 2001 were twice as likely to also go down in GPA as students who stayed stable or went up in assets. The actual numbers of students going down in GPA were quite small, so these results must be treated extremely cautiously. Nevertheless, they are provocative and suggest the need for research with larger samples to determine if these patterns hold. Of the 127 students who went down in assets from 1998–2001, seven went down in GPA, for odds of .055. Of the 180 students who stayed stable or went up in assets, five went down in GPA, for odds of .027. The odds of going down in GPA were thus .055/.027 ¼ 2.04 times greater for those who went down in assets than for those who stayed stable or went up in assets.

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

Intercept Assets

- 0.70*

703

Slope Assets

- 0.15* 0.37* 0.43*

Intercept GPA

0.31*

- 0.03*

Slope GPA

χ2 (10) = 15.50, p<.01, RMSEA = .04

Fig. 2. Growth curve model investigating the relationship between GPA and assets from 1997 to 2001.

pp:05). The slope of the growth curve indicates that the greater the increase in students’ assets over that period, the greater the increase in their GPA, such that, for approximately every five assets students increased, they increased a half-letter grade.

Discussion These different analyses, from simple correlations to longitudinal modeling, collectively describe a positive relationship between developmental assets and GPA. The higher students’ asset levels, the higher their current GPA and their GPA several years later. In addition, increases in assets are related to increases in GPA. In this sample, the modal developmental path for students moving from middle school through high school was a decline in assets. In a practical sense then, if school-based positive youth development strategies can simply help maintain students’ earlier asset levels, much less increase them, subsequent GPA may be favourably affected. Our longitudinal results were modest but significant, noteworthy since the sample size was relatively small. The strength of the relationship of assets to achievement compares favourably with that of numerous demographic, classroom, and school variables. For example, comprehensive school reform models, such as Success for All and the Comer School Development Program, have average effect sizes on achievement test scores of d ¼ :15 (Borman, Hewes, Overman, & Brown, 2002). That is, the effects represented about 1/8th of a S.D., 2.5 Normal Curve Equivalents on a percentile basis. For example, this reflects the difference in a student moving from a 70th percentile to a 73rd percentile test score, certainly an improvement, but not a stunning one. Similarly, in the national samples used to norm the Stanford Achievement Test, 5th graders scored only 1/3 of a S.D. than 4th graders in reading. In other words, ‘‘everything that happens to a student between the end of 4th grade and the end of 5th grade—a whole school year of full-day classroom instruction, interactions with family, conversations with friends, and homework—is associated with an important but not huge gain on an achievement test’’ (Granger & Kane, 2004, p. 76). Our findings of significant small to moderate relationships between developmental assets and GPA are consistent with such previous research.

ARTICLE IN PRESS 704

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

From an applied standpoint, even small effects can change lives. We found a correlation of .31 between the number of developmental assets students experienced in 1997and their GPA the following year, and .24 between their 1997 number of assets and their GPA 4 years later. By comparison, meta-analyses have found that the correlation between use of aspirin and reduced death due to heart attack is just .02, between anti-hypertensive medication and reduced stroke is just .03, and between parental divorce and later child well-being is only .09. Uncorrected correlations above .30 are sufficiently rare in medicine, education, and the social sciences that one research team suggested that researchers should be ‘‘rather satisfied’’ with correlations in the teens, ‘‘pleased’’ with those in the upper twenties, and ‘‘rejoice’’ at those above .50 (Meyer et al., 2001). Study limitations and needed research A recent study illustrates the challenge inherent in attempting to connect developmental asset levels and achievement longitudinally: Academic achievement measures typically do not show much change over time. The National Assessment of Educational Progress mathematics tests showed that the average percent gain made by a student moving from 4th grade in 1996 to 8th grade in 2000 was just 1.6%, or only .4%/year (Viadero, 2003). Similarly, GPA for most children, especially those at low risk and at least average intelligence, is fairly stable, particularly over relatively brief periods such as 1–3 years (Gutman, Sameroff, & Cole, 2003). Similarly, over our 3-year study period the great majority (85%) of students were relatively stable in their GPA, with only 4% going down and 12% up a meaningful amount. The mean change in this sample was only .33 points on a 12-point grade scale. These results suggest the need for large sample sizes and long study periods to reliably connect such small changes to predictor variables. Moreover, although there was more variation in assets levels than in GPA, a decline in assets was developmentally normal as students moved from being in 7th–9th grades to being in 10th–12th grades. The single largest group of students decreased their total number of assets (41%). A total of 34% of the students were stable (34%), and the smallest group, 24% of the students, went up in their asset totals over 3 years.4 The combination of those facts added difficulty to the task of detecting whether assets are connected to change in academic achievement. Our sample also over-represented white students from highly educated and intact families. Significant concurrent relations have been reported between assets and educational outcomes for youth of colour, urban youth, and youth from single-parent families (Scales et al., 2005; Sesma & Roehlkepartain, 2003). However, the analyses discussed here need to be conducted with similarly diverse samples to understand whether these longitudinal linkages between assets and GPA also obtain for non-white, urban students, and those from single-parent families. Then too, this sample looked only at middle school students, followed into high school. More study is needed of school districts implementing asset building efforts throughout the K-12 grades, and following children for all those years, to better understand how building students’ 4

Overall, the average asset levels of the students declined slightly across the 3 years reported here, from 20.97 assets in 1998, to 19.06 in 2001. By asset quartile in from 1998 to 2001, the percentage with just 0–10 assets went from 7% to 11%, those with 11–20 assets went from 36% to 44%, those with above average levels of assets (21–30) were stable at 33%, and the percentage of students who were asset-rich (31–40 assets) declined from 12% to 5%.

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

705

developmental assets affects achievement not just in middle and high school, but across the school-age years. Finally, although the results we report here are promising, insufficient data are available to explain exactly how they may have occurred. More documentation is needed of what schools and communities do to intentionally use developmental asset building as an achievement strategy, so that the implications for suggesting policy and practice improvements become clearer.

Conclusion Promoting positive youth development and traditional school reform strategies are not two separate paradigms. For example, in an 18-month in-depth case study of four Southern California schools, researchers concluded that the biggest ‘‘problem’’ in the schools, as seen ‘‘from the inside,’’ was the quality of human relationships, between teachers and students, parents and teachers, students and students, and so on (Poplin & Weeres, 1992). In results consistent with the findings of the present study, they concluded that focusing on the challenge of raising student achievement without focusing on improving human relationships and human development within the school community was not likely to produce the kind of transformation schools need to realize higher achievement goals. The results of the present study offer promising evidence of how a broad focus on building the developmental nutrients in young people’s lives may contribute to promoting their academic achievement. Building developmental assets thus merits consideration as one of the strategies districts and communities can use to positively affect achievement. Especially if careful attention is paid to how strategies for promoting this positive youth development can be infused to strengthen classroom practices and curriculum and instruction (e.g. Starkman et al., 1999; Taccogna, 2003), then the already apparent link between developmental assets and achievement likely can be strengthened. The data presented here suggest that the benefits to students, their families, schools, and communities are likely to be realized in both the short term and in years to come.

References Appleyard, K., Egeland, B., van Dulmen, M. H. M., & Sroufe, L. A. (2005). When more is not better: The role of cumulative risk in child behavior outcomes. Journal of Child Psychology and Psychiatry and Allied Disciplines, 46, 235–245. Araque, J. (2002). The A.P.I. study: Academics and prevention inseparable. Costa Mesa, CA: Orange County Department of Education, Learning Support Instructional Services Division. Arroyo, C. G., & Zigler, E. (1995). Racial identity, academic achievement, and the psychological well-being of economically disadvantaged adolescents. Journal of Personality and Social Psychology, 69, 903–914. Bagwell, C. L., Schmidt, M. E., Newcomb, A. F., & Bukowski, W. M. (2001). Friendship and peer rejection as predictors of adult adjustment. New Directions for Child and Adolescent Development(91), 25–49. Barber, B., & Eccles, J.S. (1997). Student council, volunteering, basketball, or marching band: What kind of extracurricular involvement matters? Symposium paper presented at the biennial meeting of the Society for Research on Child Development, April 1997, Washington, DC. Benson, P. L. (1997). All kids are our kids: What communities must do to raise caring and responsible children and adolescents. San Francisco: Jossey-Bass.

ARTICLE IN PRESS 706

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

Benson, P. L., Leffert, N., Scales, P. C., & Blyth, D. A. (1998). Beyond the ‘village’’ rhetoric: Creating healthy communities for children and adolescents. Applied Developmental Science, 2, 138–159. Benson, P. L., Roehlkepartain, E. C., & Sesma, A., Jr. (2004h). Tapping the power of community: The potential of asset building to strengthen substance abuse prevention efforts. Search Institute Insights and Evidence, 2(1). Benson, P. L., Scales, P. C., Leffert, N., & Roehlkepartain, E. C. (1999). A fragile foundation: The state of developmental assets among American youth. Minneapolis: Search Institute. Benson, P. L., Scales, P. C., & Mannes, M. (2003). Developmental strengths and their sources: Implications for the study and practice of community-building. In R. M. Lerner, F. Jacobs, & D. Wertlieb (Eds.), Handbook of applied developmental science. Vol. 1, Applying developmental science for youth and families—Historical and theoretical foundations (pp. 369–406). Thousand Oaks, CA: Sage. Billig, S. (2004). Heads, hearts, and hands: The research on K-12 service-learning. In National Youth Leadership Council, Growing to greatness: The state of service-learning project. St. Paul, MN: NYLC. Borman, G. D., Hewes, G. M., Overman, L. T., & Brown, S. (2002). Comprehensive school reform and student achievement: A meta-analysis. Baltimore: Center for Research on the Education of Students Placed At Risk, Johns Hopkins University. Bredekamp, S., & Copple, C. (1997). Developmentally appropriate practice in early childhood programs. Washington, DC: National Association for the Education of Young Children. Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press. Bronfenbrenner, U., & Morris, P.A. (1998). The ecology of developmental processes. In: Damon, W. (Series Ed.), Lerner, R.M. (Vol. Ed.), Handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed., pp. 993–1028). New York, NY: Wiley. Catterall, J. S. (1998). Risk and resilience in student transitions to high school. American Journal of Education, 106, 302–334. Developmental assets: A portrait of your youth. (2001). Minneapolis: Search Institute (1999–2000 school year aggregate database). Fellner, R., Jackson, A., Kusak, D., Mulhall, P., Brand, S., & Flowers, N. (1997). The impact of school reform for the middle grades: A longitudinal study of a network engaged in turning points-based comprehensive school transformation. In R. Takanishi, & D. Hamburg (Eds.), Preparing adolescents for the twenty-first century: Challenges facing Europe and the United States (pp. 38–69). Cambridge: Canbridge University Press. Fletcher, A. C., Newsome, D., Nickerson, P., & Bazley, R. (2001). Social network closure and child adjustment. Merrill-Palmer Quarterly, 47, 500–531. Granger, RC., & Kane, T. (2004b). Improving the quality of after-school programs. Education Week, 23(23) [76, 52]. Greenberg, M. T., Weissberg, R. P., O’Brien, M. U., Zins, J. E., Fredericks, L., Resnik, H., et al. (2003). Enhancing school-based prevention and youth development through coordinated social, emotional, and academic learning. American Psychologist, 58, 466–474. Gutman, L. M., Sameroff, A. J., & Cole, R. (2003). Academic growth curve trajectories from 1st grade to 12th grade: Effects of multiple social risk factors and preschool child factors. Developmental Psychology, 39, 777–790. Gutman, L. M., Sameroff, A. J., & Eccles, J. S. (2002). The academic achievement of African American students during early adolescence: An examination of multiple risk, promotive, and protective factors. American Journal of Community Psychology, 30, 367–399. Hanson, T. L., Austin, G., & Lee-Bayha, J. (2003). Student health risks, resilience, and academic performance: Year 1 report. Los Alamitos, CA: WestEd, Health and Human Development Program. Hofferth, S. L., & Sandberg, J. F. (2001). How American children spend their time. Journal of Marriage and the Family, 63, 295–308. Jessor, R., VanDen Bos, J., Vanderryn, J., Costa, F. M., & Turbin, M. S. (1995). Protective factors in adolescent problem behavior: Moderator effects and developmental change. Developmental Psychology, 31, 923–933. Leffert, N., Benson, P. L., Scales, P. C., Sharma, A., Drake, D., & Blyth, D. A. (1998). Developmental assets: Measurement and prediction of risk behaviors among adolescents. Applied Developmental Science, 2, 209–230.

ARTICLE IN PRESS P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

707

Lerner, R. M., Lerner, J. V., De Stefanis, I., & Apfel, A. (2001). Understanding developmental systems in adolescence: Implications for methodological strategies, data analytic approaches, and training. Journal of Adolescent Research, 16, 9–27. Lerner, R. M., Wertlieb, D., & Jacobs, F. (2003). Historical and theoretical bases of applied developmental science. In R. M. Lerner, D. Wertlieb, & F. Jacobs (Eds.), Handbook of applied developmental science. Vol. 1. Applying developmental science for youth and families—Historical and theoretical foundations (pp. 1–28). Thousand Oaks, CA: Sage. Mahoney, J. L., Cairns, B. D., & Farmer, T. W. (2003). Promoting interpersonal competence and educational success through extracurricular activity participation. Journal of Educational Psychology, 95, 409–418. Malecki, C. K., & Elliot, S. N. (2002). Children’s social behaviors as predictors of academic achievement: A longitudinal analysis. School Psychology Quarterly, 17, 1–23. Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies, R. R., et al. (2001). Psychological testing and psychological assessment: A review of evidence and issues. American Psychologist, 56, 128–165. Mounts, N. S., & Steinberg, L. (1995). An ecological analysis of peer influence on adolescent grade point average and drug use. Developmental Psychology, 31, 915–922. National Middle School Association. (2003). This we believe: Successful schools for young adolescents. Westerville, OH: NMSA. NICHD Early Child Care Research Network. (2004). Are child development outcomes related to before- and afterschool care arrangements? Results from the NICHD study of early child care. Child Development, 75, 280–295. Oakes, J., Quartz, K.H., Ryan, S., & Lipton, M. (2000). Civic virtue and the reform mill. Education Week, 19(24), 68, 43. Petit, G. S., Bates, J. E., & Dodge, K. A. (1997). Supportive parenting, ecological context, and children’s adjustment: A seven-year longitudinal study. Child Development, 68, 908–923. Poplin, M., & Weeres, J. (1992). Voices from the inside: A report on schooling from inside the classroom. Claremont, CA: Claremont Graduate School, Institute for Education in Transformation. Roeser, R. W., Midgely, C., & Urdan, T. C. (1996). Perceptions of the school psychological environment and early adolescents’ psychological and behavioral functioning in school: The mediating role of goals and belonging. Journal of Educational Psychology, 88, 408–422. Rutter, M. (2000). Resilience reconsidered: Conceptual considerations, empirical findings, and policy implications. In J. P. Shonkoff, & S. J. Meisels (Eds.), Handbook of early childhood intervention, (2nd ed) (pp. 651–682). New York, NY: Cambridge. Ryan, R. M., Stiller, J. D., & Lynch, J. H. (1994). Representations of relationships to teachers, parents, and friends as predictors of academic motivation and self-esteem. Journal of Early Adolescence, 14, 226–249. Sameroff, A. J. (1999). Ecological perspectives on developmental risk. In J. D. Osofsky, & H. E. Fitzgerald (Eds.), WAIMH handbook of infant mental health: Vol. 4. Infant mental health groups at risk (pp. 223–248). New York, NY: Wiley. Scales, P. C., Benson, P. L., Leffert, N., & Blyth, D. A. (2000). Contribution of developmental assets to the prediction of thriving among adolescents. Applied Developmental Science, 4, 27–46. Scales, P. C., Foster, K., Mannes, M., Horst, M., & Rutherford, A. (2005s). School-business partnerships, developmental assets, and positive outcomes among urban high school students: A mixed-methods study. Urban Education, 40, 144–189. Scales, P. C., & Leffert, N. (2004). Developmental assets: A synthesis of the scientific research on adolescent development (2nd ed). Minneapolis: Search Institute. Scales, P. C., Sesma, A., Jr., & Bolstrom, B. (2004). Coming into their own: How developmental assets promote positive growth during middle childhood. Minneapolis: Search Institute. Schmidt, J. A., & Padilla, B. (2003). Self-esteem and family challenge: An investigation of their effects on achievement. Journal of Youth and Adolescence, 32, 37–46. Sesma, A., Jr., & Roehlkepartain, E. C. (2003). Unique strengths, shared strengths: Developmental assets among youth of color. Search Institute Insights and Evidence, 1(2), 1–13. Shiner, R. L. (2000). Linking childhood personality with adaptation: Evidence for continuity and change across time into late adolescence. Journal of Personality and Social Psychology, 78, 310–325.

ARTICLE IN PRESS 708

P.C. Scales et al. / Journal of Adolescence 29 (2006) 691–708

Starkman, N., Scales, P. C., & Roberts, C. (1999). Great places to learn: How asset-building schools help students succeed. Minneapolis: Search Institute. Steinberg, L. (2001). We know some things: Parent–adolescent relationships in retrospect and prospect. Journal of Research on Adolescence, 11, 1–19. Taccogna, J. (Ed.). (2003). Powerful teaching: Developmental assets in curriculum and instruction. Minneapolis: Search Institute. Taylor, C. S., Lerner, R. M., von Eye, A., Bobek, D. L., Balsano, A. B., Dowling, E., et al. (2003). Positive individual and social behavior among gang and non-gang African American male adolescents. Journal of Adolescent Research, 18, 496–522. Viadero, D. (2003). Study finds higher gains in states with high-stakes tests. Education Week, 22(31), 10. Weissberg, R. P., Kumpfer, K. L., & Seligman, M. E. P. (2003). Prevention that works for children and youth. American Psychologist, 58, 425–432. Weissberg, R. P., & O’Brien, M. U. (2004). What works in school-based social and emotional learning programs for positive youth development. Annals of the American Academy of Political and Social Science, 591, 86–97. Wenz-Gross, M., Siperstein, G. N., Untch, A. S., & Widaman, K. F. (1997). Stress, social support, and adjustment of adolescents in middle school. Journal of Early Adolescence, 17, 129–151.