Please take our brief survey

Blueprints Programs = Positive Youth Development

Return to Search Results

Promising Program Seal

Project Northland

Blueprints Program Rating: Promising

Provides classroom curricula, peer leadership, youth-driven extra-curricular activities, parent involvement programs, and community activism to reduce teen alcohol use, improve parent-child communication about alcohol use, increase students' self-efficacy to resist alcohol and understanding of alcohol use norms, and reduce students' ease of access to alcohol in their communities.

  • Alcohol

    Program Type

    • Academic Services
    • Alcohol Prevention and Treatment
    • Community Mobilization
    • School - Individual Strategies

    Program Setting

    • School

    Continuum of Intervention

    • Universal Prevention (Entire Population)

    Provides classroom curricula, peer leadership, youth-driven extra-curricular activities, parent involvement programs, and community activism to reduce teen alcohol use, improve parent-child communication about alcohol use, increase students' self-efficacy to resist alcohol and understanding of alcohol use norms, and reduce students' ease of access to alcohol in their communities.

      Population Demographics

      Project Northland targets 6th, 7th, and 8th grade students, although program components have been extended into the 9th, 11th and 12th grades. Applied initially to school districts in six rural, lower-middle-class to middle-class counties in Minnesota, it has been extended to urban minority school districts in Chicago, schools in a Croatian city in Europe, and schools in a mid-sized city in Illinois.

      Age

      • Early Adolescence (12-14) - Middle School
      • Late Adolescence (15-18) - High School

      Gender

      • Male and Female

      Gender Specific Findings

      • Female

      Race/Ethnicity

      • All Race/Ethnicity

      Race/Ethnicity/Gender Details

      Project Northland reduced alcohol use among primarily white students from rural, lower-middle- to middle-class counties in Minnesota. However, it failed to have much influence on primarily African-American and Latino students in Chicago. In a study in Croatia, the intervention did not significantly affect male students, but was highly significant for females on the Tendency to Use Alcohol Scale that combines intentions with behavior.

      • Individual
      • Peer
      • School
      • Family
      • Neighborhood/Community
      Risk Factors
      • Individual: Favorable attitudes towards drug use, Substance use
      • Peer: Peer rewards for antisocial behavior, Peer substance use
      • Family: Parental attitudes favorable to drug use
      • Neighborhood/Community: Community disorganization, Laws and norms favorable to drug use/crime*, Low neighborhood attachment, Perceived availability of drugs*
      Protective Factors
      • Individual: Perceived risk of drug use, Prosocial involvement, Refusal skills
      • Peer: Interaction with prosocial peers
      • Family: Opportunities for prosocial involvement with parents, Parental involvement in education, Rewards for prosocial involvement with parents
      • School: Opportunities for prosocial involvement in education, Rewards for prosocial involvement in school
      • Neighborhood/Community: Opportunities for prosocial involvement, Rewards for prosocial involvement

      *Risk/Protective Factor was significantly impacted by the program.

      See also: Project Northland Logic Model (PDF)

      Project Northland is a six-year intervention delivered over seven academic years from middle to high school, but a shortened, 3-year version may also be used in grades 6, 7, and 8. It is a multi-level intervention to include demand (individual level) and supply (environmental level) reduction strategies. Its main intervention components include classroom curricula, peer leadership, youth-driven extra-curricular activities, parent involvement programs, and community activism. By intervening on multiple levels, Project Northland strives to teach students skills to effectively negotiate social influences to drink, while at the same time directly modifying the social environment of youth (i.e., peers, parents, school, and community).

      Project Northland provides six years of comprehensive programming for students beginning in sixth grade. Phase I provides programming for 6th, 7th, and 8th graders. Each of the three years of programming has a specific theme and incorporates individual, parent, peer, and community participation. The students receive skills training in communicating with their parents about alcohol (6th grade), dealing with peer influence and normative expectations about alcohol (7th grade), and understanding methods that bring about community-level changes in alcohol-related programs and policies (8th grade).

      Phase II of the program is designed to help maintain the effects through high school. Phase II intervention strategies include community organizing, parent education, youth development, media, and school curriculum. The five intervention strategies are designed to increase community efficacy to enact changes in policies and practices related to high school students' alcohol use.

      Sixth grade: The "Slick Tracy Home Team Program" consists of four weekly sessions of activity-story books which the students complete as homework with their parents. The books also include information on young adolescent alcohol use for parents. Small-group discussions about the books are held during school, and an evening fair at which students' posters and projects are displayed. Community-wide task forces are created and include members from a cross section of the communities: government officials, school and business representatives, law enforcement personnel, health professionals, youth workers, parents, concerned citizens, clergy and adolescents.

      Seventh grade (1992-93): The "Amazing Alternatives! Program" includes the following: an eight-week peer- and teacher-led classroom curriculum using interactive activities related to themes of why young people use alcohol and alternatives to use, influences in terms of drinking, strategies for resisting those influences, normative expectations that most people their age do not drink, and intentions not to drink; a peer participation program called T.E.E.N.S. creates alternative alcohol-free activities outside the classroom using adult volunteers; parental involvement through the use of booklets mailed to parents includes behavioral prescriptions for parents and activities for parents and children to complete as well as updated information on the program and its events.

      Eighth grade: "PowerLines" consists of an eight-session classroom curriculum that introduces students to the "power" groups within their communities that influence adolescent alcohol use and availability and teaches community action/citizen participation skills. Students interview parents, local government officials, law enforcement personnel, school teachers and administrators, and retail alcohol merchants about their beliefs and activities concerning adolescent drinking and conduct a town meeting to make recommendations for community action to prevent alcohol use. A theater production is performed at each school for classmates, parents and community members. T.E.E.N.S. continues to provide alternative activities in all intervention school districts. A newsletter written by and for eighth grade students is sent to parents and peers. Last, the community-wide task force continues in its activities to limit access to alcohol.

      Ninth grade: During the interim Phase (1994-1996), a brief five-session classroom program entitled "Shifting Gears" is implemented. This program focuses on pressures to drink and drive or to ride with a drinking driver, the influences and tactics of alcohol advertising, and ways to deal with those influences. No programs are implemented in grade ten.

      Eleventh and twelfth grades (1997-1998): Phase II consists of five components designed to reinforce and complement the previous goals of Project Northland. The "Class Action" program is implemented during Phase II, containing a substance abuse prevention curriculum that can be used as a booster session for the Project Northland series or as a stand-alone curriculum. "Class Action" builds upon the early adolescent interventions with new strategies for the cohort's last years in high school while emphasizing changes in the social environment of young people. The 6-session curriculum is based on the social influences theory of behavior change, where students are asked to debate and discuss the social influences to use alcohol and the negative consequences those influences have not only on the individual teen, but on the community as a whole. Through an innovative, civil-trial approach, Class Action challenges high-school students to examine the real-world consequences, both legal and social, of teen alcohol use. The program is peer-led and uses interactive methods to accomplish its instructional goals. Students debate and discuss the consequences of substance abuse, thus changing the social norms around alcohol use and changing negative peer pressure into positive peer pressure.

      Community and parent components of Project Northland are also continued during the implementation of "Class Action." Eleven Action teams, representing all communities and ranging in size from 5 to 12 members are formed and asked to encourage community adoption of institutional or policy solutions to underage drinking through a number of sponsored activities and awareness campaigns. For the parent component, a postcard campaign is designed to increase awareness about alcohol-related issues and encourage specific actions to keep adolescents alcohol free. A total of 11 postcards designed to correspond to events sponsored by the action teams are mailed to the cohort's parents at 6-week intervals. A new parent-child intervention, "Sound OFF!", is used in the senior year to encourage parents and their seniors to communicate about alcohol. A series of three mailings are sent to parents and seniors in November, January, and February with discussion questions regarding teen drinking. Students are asked to answer the questions and return the mailer to the researchers.

      Print media campaigns are also implemented. The first targets young adults with the theme of "Don't provide to those under age 21." Other print media include calendars for alcohol merchants, newsletters for students and adults, and a celebration poster for the many participants in Project Northland over the years of the study.

      In addition, youth action teams are formed in Phase II. Young people are recruited for participation in teams to affect alcohol use among their peers. The teams meet as an extracurricular activity, assisted by adult coordinators.

      Project Northland is a multi-component, community-wide intervention that targets the environmental, intra-personal, and behavioral factors that may influence adolescent drinking. Both demand and supply reduction approaches are targeted in the home and community. Phase II (grades 11 & 12) is based on social cognitive theory, with an emphasis on changes in social environmental factors (such as changing social norms). Each of the five Phase II intervention strategies was designed to make changes in these factors, with normative changes having priority.

      • Cognitive Behavioral
      • Normative Education
      • Person - Environment
      • Self Efficacy
      • Skill Oriented
      • Social Learning

      The main evaluation of Project Northland, a multi-level, multi-year intervention program for youth, began with a 7-year study of six counties and 24 school districts in northeast Minnesota. School districts (and adjoining communities) were randomized to an intervention or reference condition. Students in grade 6 at the baseline assessment in fall 2001 (n = 2,351) were followed through grade 12 in spring 1998 (with 67.8% of the original sample completing the assessment). Outcome measures came from surveys of students about the use of alcohol and other substances, surveys of parents about changes at home in alcohol use, and surveys of community leaders about alcohol policy changes.

      Additional 3-year evaluations of the project came from samples in Chicago, Croatia, and Kankakee County, Illinois (none are strong enough methodologically to meet the criteria for replication). The Chicago evaluation randomly assigned 61 schools to intervention and control conditions. Beginning in fall 2002, a racially mixed sample of 5,298 grade 6 through 8 students in the schools was examined over the three years. The Croatian evaluation randomly assigned 26 schools to intervention and control conditions. Rather than follow the same sample over time, the study assessed all grade 6 students in the first year (2003), all grade 7 students in the second year (2004), and all grade 8 students in the third year (2005). The analysis then examined aggregate changes in alcohol use for 1,981 students. The Kankakee evaluation selected 14 schools for the intervention in 2002 but, rather than assign other schools to a control condition, it used previous data from 1999 on another program intervention for comparison. The study followed grade 6 students in 2002 for three years to grade 8, but did not report the number of students participating.

      Alcohol Use: Students in the intervention districts had significantly lower scores on the Tendency to Use Alcohol Scale by the end of eighth grade. The scale score was also significantly lower among baseline nonusers in the intervention districts, and lower but not significantly so for baseline users at eighth grade. The percentages of past month and past week alcohol users were also significantly lower for intervention groups at the end of eighth grade. For baseline nonusers, students in the intervention districts consistently showed significantly lower onset rates at eighth grade. Students in the intervention schools were also significantly less likely to increase their Tendency to Use Alcohol and binge drinking during Phase II (grades 11 and 12), and marginally less likely to increase past month alcohol use.

      Cigarette, Smokeless Tobacco, and Marijuana Use: For baseline nonusers of alcohol, those reporting cigarette use and marijuana use were significantly lower in the intervention districts, and smokeless tobacco approached significance at the end of eighth grade. Students in the intervention districts reported a significant reduction in the combination of alcohol and cigarette use.

      Psychosocial Factors: According to growth curve modeling, students in the intervention schools had significantly lower scores on the Peer Influence Scale and Perceived Access to alcohol by the end of 8th grade (Phase I of intervention). Using regression models, normative expectations about how many young people drink, parent-child communication about the consequences of alcohol use, and the importance of reasons for not using alcohol were also impacted favorably by the program at the end of grade 8. Among baseline non-users, intervention students had significantly lower scores by 8th grade on the Peer Influence Scale and higher scores on the Self-Efficacy Scale. None of the psychosocial factors were significant at the end of Phase II.

      Chicago Study (Komro et al., 2008): The results for inner-city, largely minority students showed no statistically significant differences between the intervention and control schools in the growth rate for any of the primary outcome measures (drug use, alcohol use, alcohol intentions) or mediating outcome measures (norms supportive of use, perceived outcomes supportive of use, lack of resistance self-efficacy, parental involvement, and limited access to alcohol).

      Croatia Study (West et al., 2008): Intervention schools demonstrated significantly smaller increases in the Tendency to Use Alcohol score than did the control group from baseline to year 2. Although there was a similar difference for the increases from baseline to year 3, it was non-significant.

      Kankakee Study (Center for Prevention Research and Development, 2004): In all three subgroups (white, non-white, and low-income non-white), Project Northland participants had significantly more favorable outcomes related to alcohol use in the past year and getting drunk in the past month than non-Project Northland students. Low-income, non-white Project Northland students had significantly more favorable outcomes related to alcohol use in the past month, cigarette use in the past year, and marijuana use in the past year than non-Project Northland students. Non-white and low-income Project Northland students had a significantly smaller increase in the rate of alcohol use from sixth to eighth grade as compared to non-Project Northland students.

      Minnesota Study (Perry et al., 1996, 2002)

      • Students in the intervention drank significantly less and reported less alcohol onset than control students at the end of 8th grade.
      • Students in the intervention group who were never-drinkers at the beginning of sixth grade not only drank significantly less than students in the control group, they also smoked fewer cigarettes and used less marijuana at the end of the eighth grade.
      • Project Northland was effective in changing Peer Influence to use alcohol and Perceived Access to alcohol by the end of Phase I (8th grade), but these psychosocial variables were not affected during Phase II (grades 11 and 12).
      • Students in the intervention schools were significantly less likely to increase their Tendency to Use Alcohol and binge drinking, and marginally less likely to increase past month alcohol use during grades 11 and 12.

      Significant Program Effects on Risk and Protective Factors:

      • Normative expectations about how many young people drink, parent-child communication about the consequences of alcohol use, and the importance of reasons for not using alcohol were also impacted by the end of grade 8.
      • The intervention reduced the ability to purchase alcohol in off-sale outlets during Phase II of implementation (grades 11 and 12).

      Measures of peer influence, self-efficacy, perceived access, alcohol norms, parent communication about alcohol, and functional meaning attributed to alcohol can be seen as mediators between the intervention and alcohol and drug use outcomes. Although a full mediation analysis was not done, evidence of program effects on these mediators provided some support for the posited theoretical mechanisms.

      No effect sizes were presented.

      Although the original study of Project Northland was conducted among a predominantly White, rural, lower-middle- to middle class population, additional studies in Chicago and Kankakee had ethnically diverse samples, and another study was conducted among Croatian children. Positive results were found for the White sample in Minnesota, but not for the minority sample in inner-city Chicago. In the Croatian implementation (West et al., 2008), the intervention did not significantly affect male students, but was highly significant for females. Females in the intervention group had significantly smaller scores on the Tendency to Use Alcohol index at years 2 and 3 compared to females in the control group.

      The initial Project Northland study in Minnesota would have been strengthened by greater equivalence between intervention and reference districts at baseline, although the school districts were randomized to condition. Sample sizes between groups were equivalent; however, there were more White students in the reference districts and there was greater alcohol use in the intervention districts. The Chicago study had high attrition from students moving to other schools and low participation of students and parents in the program, as well as an absence of results. The Kankakee study has no control group. Rather, data is compared to a countywide youth survey with similar measures. In the Croatian study, the program was not effective for males and although there are some small significant differences in alcohol use at year 2, they disappear by year 3.

      Project Northland has evolved over the years to become a six-year intervention to be delivered over seven academic years. The original Project Northand was designed as a three-year program to be implemented in grades six through eight. Given the development and change in curriculum of Project Northland, none of the additional evaluations of Phase I (Slick Tracy, Amazing Alternatives, and Powerlines) delivered in grades six-eight should be considered replications of the newer, six-year program. Additionally, the components delivered during the high school years have not been evaluated as stand-alone programs and should not be considered evidence-based, unless delivered in combination with the middle-school program.

      • Blueprints: Promising
      • Crime Solutions: Promising
      • OJJDP Model Programs: Promising
      • SAMHSA: 2.9-3.4

      Kankakee/Iroqois ROE
      Brenda Wetzel
      189 E Court St, Suite 600
      Kankakee, IL 60901
      815-936-4630
      bwetzel@i-kan.org

      Center for Prevention Research and Development (2004). Project Northland evaluation for Iroquois-Kankakee Regional Office of Education. Unpublished report. Institute of Government and Public Affairs, University of Illinois.

      Komro, K. A., Perry, C. L., Veblen-Mortenson, S. V., Farbakhsh, K., Toomey, T. L., Stigler, M. H., ... Williams, C. L. (2008). Outcomes from a randomized controlled trial of a multi-component alcohol use preventive intervention for urban youth: Project Northland Chicago. Addiction, 103(4), 606-618.

      Perry, C. L., Williams, C. L, Komro, K. A., Veblen-Mortenson, S., Forster, J. L., Bernstein-Lachter, R., ... McGovern, P. (2000). Project Northland high school interventions: Community action to reduce adolescent alcohol use. Health Education and Behavior, 27(1), 29-49.

      Perry, C. L., Williams, C. L, Komro, K. A., Veblen-Mortenson, S., Stigler, M. H., Munson, K. A., ... Forster, J. L. (2002). Project Northland: Long-term outcomes of a community action to reduce adolescent alcohol use. Health Education and Research Theory and Practice, 17(1), 117-132.

      Perry, C. L., Williams, C. L., Veblen-Mortenson, S., Toomey, T. L., Komro, K. A., Anstine, P. S., ... Wolfson, M. (1996). Project Northland: Outcomes of a communitywide alcohol use prevention program during early adolescence. American Journal of Public Health, 86(7), 956-965.

      Stigler, M. H., Perry, C. L. Komro, K. A., Cudeck, R., & Williams, C. L. (2006). Teasing apart a multiple component approach to adolescent alcohol prevention: What worked in Project Northland? Prevention Science, 7, 269-280.

      West, B., Abatemarco, D., Ohman-Strickland, P. A., Zec, V., Russo, A., & Milic, R. (2008). Project Northland in Croatia: Results and lessons learned. Journal of Drug Education, 38, 55-70.

      Project Northland
      Hazelden Publishing and Education
      P. O. Box 176
      Center City, MN 55012-0176
      Phone: 1-800-328-9000 ext. 4324
      Email: kmcelfresh@hazelden.org
      www.hazelden.org/web/go/projectnorthland

      Study 1

      Perry, C. L., Williams, C. L, Komro, K. A., Veblen-Mortenson, S., Stigler, M. H., Munson, K. A., ... Forster, J. L. (2002). Project Northland: Long-term outcomes of a community action to reduce adolescent alcohol use. Health Education and Research Theory and Practice, 17(1), 117-132.

      Perry, C. L., Williams, C. L., Veblen-Mortenson, S., Toomey, T. L., Komro, K. A., Anstine, P. S., ... Wolfson, M. (1996). Project Northland: Outcomes of a communitywide alcohol use prevention program during early adolescence. American Journal of Public Health, 86(7), 956-965.

      Stigler, M. H., Perry, C. L. Komro, K. A., Cudeck, R., & Williams, C. L. (2006). Teasing apart a multiple component approach to adolescent alcohol prevention: What worked in Project Northland? Prevention Science, 7, 269-280.

      Perry, C. L., Williams, C. L., Veblen-Mortenson, S., Toomey, T. L., Komro, K. A., Anstine, P. S., ... Wolfson, M. (1996). Project Northland: Outcomes of a communitywide alcohol use prevention program during early adolescence. American Journal of Public Health, 86 (7), 956-965.

      Perry, C. L., Williams, C. L, Komro, K. A., Veblen-Mortenson, S., Stigler, M. H., Munson, K. A., ... Forster, J. L. (2002). Project Northland: Long-term outcomes of a community action to reduce adolescent alcohol use. Health Education and Research Theory and Practice, 17 (1), 117-132.

      Stigler, M. H., Perry, C. L. Komro, K. A., Cudeck, R., & Williams, C. L. (2006). Teasing apart a multiple component approach to adolescent alcohol prevention: What worked in Project Northland? Prevention Science, 7, 269-280.

      Evaluation Methodology

      Design: Of 30 invited school districts in six Minnesota counties, 24 agreed to participate in the program. Four smaller school districts were combined with nearby districts (to ensure an adequate sample size in each unit), and these 20 combined districts were then blocked by size (small, medium, large and very large) and randomized to an intervention condition (n = 10) or a reference condition (n = 10). The primary study cohort was the class of 1998 in these school districts; these students were sixth graders at baseline in fall 1991 (n = 2351). Students were surveyed in their classrooms at baseline (fall 1991) and follow-up (each spring 1992-1998).

      Phase I was implemented during the 6th-8th grades (1991-1994), a brief five-session "Shifting Gears" program was implemented in 9th grade, and Phase II was implemented during the 11th and 12th grades (1996-1998).

      The reference districts continued their usual alcohol and other drug education programs, almost all which included Project DARE.

      Sample: Project Northland, conducted in northeast Minnesota, involved mostly rural, lower-middle-class to middle-class communities primarily of European ethnic backgrounds. The participants were 94% White and 5.5% Native American. Of the 2,351 students present at baseline, 93% (n = 2,192) completed surveys at the end of 6th grade, 88% (n = 2,060) completed surveys at the end of the 7th grade, 81% (n = 1,901) completed surveys at the end of the 8th grade. In 1998, the final assessment included 67.8% of the original cohort.

      Counting subjects who were not present at every data point, including baseline, the sample size for the entire study was 3,151 and included students who completed at least one survey. There were 198 students who moved between school districts during the 7 years of data collection and these students were dropped from the analysis, leaving an analysis sample size of 2,953. Of the 2,953 students who completed at least one survey, 33.4% (n = 987) had all eight data collection points.

      Measures: The student questionnaire contained the following measures: alcohol, tobacco and other drug use; peer influences; self-efficacy; functional meanings of alcohol use (reasons not to use alcohol); communication with parents; normative expectations concerning alcohol use; perceptions of ease of access to alcohol; attendance at activities with/without alcohol; and demographic factors. Other measures included a phone survey of parents in half the households of the cohort; alcohol merchant surveys; and interviews with community leaders.

      In addition, as a process measure to assess the success of the community action teams' efforts to reduce teens' access to alcohol, alcohol purchase attempts by young-looking buyers were made at the beginning of grade 6, at the end of grade 8, and grade 12 in the summers of 1991, 1994, and 1998. Telephone surveys were also conducted with parents and guardians of the participants at the end of grades 10 and 12 (1996 and 1998), to assess any changes in homes that might have resulted from the community action teams, parent postcards, print media, the "Sound OFF!" campaign and youth development programs.

      Analysis: A mixed model, intent-to-treat analysis of covariance was used to compare intervention and reference group differences at the appropriate level. Within-group differences were incorporated in each analysis by adjusting for baseline measures and race. Differences between the intervention and comparison conditions were tested using a three-level linear mixed-effects regression model for repeated measures data at the end of Phase II. School district was the unit of analysis. The main outcomes of the study were analyzed at the end of Phase II using growth curve analyses. Alcohol purchase attempt rates were analyzed using mixed-model regression methods with community specified as a nested random effect. The procedure SAS GLIMMIX was used to conduct the analysis, specifying a binomial distribution. Differences between parents in the treatment and control conditions at the beginning and end of Phase II were analyzed using mixed-model regression methods with school district as a nested random effect (with two-tailed .05 significance levels).

      Stigler et al. (2006): Growth curve analyses were used to estimate the effects of the five intervention components over time. Specifically, mixed-effects regression models appropriate for group-randomized trials were extended to accommodate repeated measures data. The analysis was conducted in two steps. First, a model was constructed to represent change in an outcome variable over time, or its "normative trajectory." Then, sets of time-varying covariates were added to the models to evaluate how participation in/exposure to one or more component(s) altered a "normative trajectory" over time.

      Outcomes

      Baseline Equivalence: At baseline, more students reported using alcohol in the intervention districts than the control districts. Baseline measures of alcohol use were therefore treated as covariates in the longitudinal models. Also, there were fewer white students in the intervention than control districts, requiring controls for race in the models.

      Differential Attrition. Of those lost to follow-up, 62% moved out of the area, 19% were refusals, 9% moved across treatment conditions, 7% were absent, and 3% were deleted because of inconsistent reporting. The attrition does not appear to bias the sample with regard to alcohol use. First, there were no significant differences in baseline alcohol use between those lost to follow-up in the intervention condition and those lost to follow-up in the control condition. Second, no significant differences were found in baseline alcohol use between those who were lost to follow-up and those who remained. However, tests for differential attrition by sociodemographic and other outcome variables were not discussed.

      Perry et al. (1996)

      Alcohol Use: Among all students (n = 1901), those in the intervention districts had significantly lower scores on the Tendency to Use Alcohol Scale by the end of eighth grade than did students in the reference groups: the mean for intervention students increased from 11.5 to 16 compared to an increase from 11 to 17.5 for reference groups (where 8 = no tendency to use alcohol and 48 indicates high tendencies). The scale score was also significantly lower among baseline nonusers in the intervention districts, and lower but not significantly so for baseline users. The percentages of past month and past week alcohol users was also significantly lower for intervention groups at the end of eighth grade as compared to reference students (for past month use, intervention students increased from 6.9 to 23.6 occasions, compared to an increase from 3.9 to 29.2; for past week use, intervention students increased from 3.8 to 10.5, compared to 2.0 and 14.8). For baseline nonusers, students in the intervention districts consistently showed significantly lower onset rates at eighth grade. For baseline users, there were more users in the intervention districts at baseline and at the end of sixth grade, but fewer by the end of seventh and eighth grades (though not statistically so).

      Cigarette, Smokeless Tobacco, and Marijuana Use
      : Among all students, there were no significant differences in the percentages of cigarette users, smokeless tobacco users, or marijuana users between conditions, although the percentage of cigarette users (defined as more than two or three lifetime occasions of use) approached significance (p < .08) at the end of eighth grade (intervention students increased from 6.9 to 24.8, compared to an increase from 4.7 to 30.7 for the reference students, a difference of 19%). For baseline nonusers of alcohol, those reporting cigarette use and marijuana use were significantly lower in the intervention districts, and smokeless tobacco approached significance (p < .06) at the end of eighth grade.

      A polydrug use measure was created by calculating the prevalence of combinations of alcohol, cigarette, and marijuana use. None of the combinations involving marijuana were significant. However, there was a significant reduction when using the combination of alcohol and cigarette use only (without marijuana included); the difference indicated a 27% reduction in "gateway" drug use.

      Psychosocial Factors: Among all students, those in the intervention districts had significantly lower scores on the Peer Influence Scale by the end of eighth grade. There were no significant differences in the Self-Efficacy or Perceived Access scales. However, the intervention students were more likely to report that they could resist alcohol at a party or dance or when offered it by a boyfriend or girlfriend (both measured as part of the Self-Efficacy scale) than their reference group counterparts. Among baseline nonusers, students in the intervention districts had significantly lower scores by eighth grade on the Peer Influence Scale and higher scores on the Self-Efficacy Scale than those in the reference schools. Among all students at the end of eighth grade, those in the intervention districts were significantly more likely to perceive that "not many people my age drink alcohol" than students in the reference districts. By the end of eighth grade, intervention students were significantly more likely to report that their parents had told them what would happen if they were caught drinking, and were marginally more likely to report that their parents talked with them about the problems involved with drinking and that their families had rules against young people drinking (p < .06 and .08, respectively) than were reference students. At the end of eighth grade, intervention students were also significantly more likely than reference students to view 9 of 10 reasons as important for not using alcohol: parents have rules, hurts reputation, fear of becoming an alcoholic, sports eligibility, costs too much money, school rules, bad for health, hurts performance, and not giving in to peer pressure. There were no significant differences between the two groups in the perception of their influence on their communities in terms of alcohol-related issues. Finally, intervention students were significantly more likely to report a greater likelihood of being disciplined by the school if they were caught driving after drinking (but there were no differences in the other six consequences of driving after drinking) than were reference students.

      At the end of 10th grade, after 2 years without a substantive intervention program, there were no significant differences between the intervention and control groups (the results of Phase I had attenuated). By the end of 11th grade, after 1 year of Phase II intervention activities, students in the intervention group were drinking less than were students in the control group, but not statistically significantly so.

      Perry et al. (2002)

      Using three-level growth curve analyses, students in the intervention schools were significantly less likely than students in the control schools to increase their Tendency to Use Alcohol, past month alcohol use, and binge drinking during Phase I of Project Northland. Students in the intervention schools were also significantly less likely to increase their Tendency to Use Alcohol and binge drinking during Phase II, and marginally less likely to increase past month alcohol use. During the Interim Phase, students in the intervention schools were significantly more likely to increase their Tendency to Use Alcohol, past month alcohol use, past week alcohol use and binge drinking.

      With regard to the psychosocial factors, during Phase I, students in the intervention schools were significantly less likely to increase their perceptions of Peer Influence to use alcohol and their Perceived Access to alcohol, but there were no differences in Self-Efficacy. During the Interim Phase, students in the intervention schools were significantly more likely to experience increased perceptions of Peer Influence to use alcohol and decreased Self-Efficacy to refuse alcohol. There were no differences in the trajectories of these scales in Phase II.

      In the reference communities, 25.4% of all outlets and 30.7% of off-sale outlets sold beer without identification. This compares with 13.6% and 5.6% in the intervention communities, a reduction of 46% in all outlets and a 81.7% reduction in potential underage purchases in off-sale outlets.

      Parents in the intervention communities had significantly less permissive norms regarding teen alcohol use at the end of Phase II than parents in the control communities.

      Stigler et al. (2006)

      Program component analysis: A secondary analysis was conducted to estimate the effects of the program's five intervention components (classroom curriculum, peer leadership, extra-curricular activities, parent program, community activism) on four relevant outcome variables (tendency to use alcohol, self-efficacy, peer influence, perceived access). None of the interaction terms introduced in the models were statistically significant, suggesting there was no synergistic effect between the components tested. The impact of the five intervention components on the four outcome variables over time appears to have been differential. Only three components significantly contributed to altering the normative trajectory of alcohol use and other psychosocial risk factors in Phase I: (1) classroom curricula, (2) extra-curricular activities (planners only), and (3) the parent involvement programs.

      Overall, the program component analysis showed that the parent involvement program had the most consistent and positive effect. Participation in the parent programs was associated with a significant decrease in the rate of increase in the Tendency to Use Alcohol and Peer Influence scales over time and a significant increase in Self-Efficacy. Being a planner for one or more extracurricular activities was also associated with a significant decrease in Tendency to Use Alcohol and increase in Self-Efficacy, but had no significant effect on Peer Influence or Perceived Access. The coefficients for planning extracurricular activities and participating in the parent programs can be compared as they are on the same "dose" scale and indicate that the effect of planning extra-curricular activities was stronger than participation in the parent involvement programs. Being exposed to the classroom curricula was associated with a significant decrease in Peer Influence and marginally significant decreases in both Tendency to Use Alcohol and Perceived Access. The level of community activism was not associated with changes in any of these outcome variables. Being a peer leader, too, does not appear to have contributed to changing Tendency to Use Alcohol, Self-Efficacy, or Peer Influence, but was associated with an increase in Perceived Access. Participation in (rather than planning) the extra-curricular activities was not associated with changes in Tendency to Use Alcohol or Perceived Access over time but was associated with a marginally significant decrease in Self-Efficacy as well as a marginally significant increase in Peer Influence over time.

      Summary: The intervention appears to have met most of its goals, and seems most successful for students who are nonusers of alcohol at the beginning of sixth grade than with those who had already initiated use.

      Komro, K. A., Perry, C. L., Veblen-Mortenson, S. V., Farbakhsh, K., Toomey, T. L., Stigler, M. H., ... Williams, C. L. (2008). Outcomes from a randomized controlled trial of a multi-component alcohol use preventive intervention for urban youth: Project Northland Chicago. Addiction, 103 (4), 606-618.

      This study evaluated an adapted version of the original Project Northand. It was designed as a three-year program to be implemented in grades 6 through 8. The intervention included the Slick Tracy, Amazing Alternatives, and Powerlines curricula described in detail above but made some changes to fit a multiethnic population. It should be noted that Project Northand has since developed into a six-year program, and as such this study does not constitute a full replication.

      Evaluation Methodology

      Design: This randomized controlled trial of schools and surrounding community areas in the city of Chicago selected its sample from a list of all public schools. For inclusion, the schools needed to teach grades 5 through 8, have mobility rates less than 25%, and have more than 30 students per grade. From this pool, 66 schools agreed to participate in the project for 3 years. To obtain sufficient sample sizes per study unit, the 66 schools were grouped into 22 units by combining schools that were geographically close and within the same city-defined community area. After matching the 22 units on demographic and academic characteristics, 10 units were randomly assigned to the intervention and 12 units to the control condition.

      Before the pretest but after randomization, four control schools and one intervention school dropped out and were no longer included in the study. Once randomized, the schools should still be followed, but it appears that the schools would not allow students to be assessed. Thus, the 10 intervention units included 29 schools and the 12 control units included 32 schools (or 61 schools in total).

      The program started in grade 6 and continued to grade 8. The pretest survey of students occurred in fall 2002 for 6th graders and was followed by three additional survey assessments. The first occurred in spring 2003 (near the end of 6th grade), the second occurred in spring 2004 (near the end of 7th grade), and the third occurred in spring 2005 (near the end of 8th grade). Because the program was ongoing during these years, none of the assessments examined the sustainability of the program after it ended.

      In the 61 schools participating in the study, 4,259 students responded to the baseline survey (91% response rate). By the third survey, 39% had been lost. The loss of cases came primarily from two schools closing and from students leaving for non-study schools. Also, investigators removed inconsistent responders (with four or more inconsistent responses) and those who moved between study conditions. However, because new students entered the schools after baseline, the total sample size over all four surveys reached 5,698 and included all persons who completed at least one assessment – even those not present at baseline.

      Surveys of parents and community leaders were done only at baseline and at the last survey (T1 and T4). For the parent survey, 70% answered at baseline and 67% responded at the last survey.

      The 3-year intervention followed the original 3-year version of Project Northland by implementing 1) 6-10 sessions per year using a peer-led classroom curriculum, 2) four home-based sessions per year to educate parents and involve them in some school and community activities, 3) peer leadership and youth-planned community involvement activities, and 4) community organizing and environmental neighborhood change. Major changes to the Chicago program included changes in the curricula to fit inner-city students, expanded home programs for parents, peer-led community service projects rather than social activities, and more emphasis on community organizing.

      Many control schools also implemented an alcohol and/or drug prevention curriculum. During the study, 69% did so in the first year, 50% in the second, and 49% in the third. The controls schools thus represented something more than a no-treatment condition.

      Sample Characteristics: The sample was gender balanced (50% each), ethnically diverse (43% black, 29% Hispanic, 13% white, 15% other), and low socioeconomic status (72% received free or reduced-price lunch). Less than half (47%) lived with both their parents, and 74% reported English as their primary language at home.

      Measures: The primary measures came from surveys of students, but supplementary measures came from assessments of parents, community leaders, and commercial access to alcohol by underage youth.

      The student surveys, which were read aloud during the assessment session, included five items to create an alcohol scale: alcohol use in the last 12 months, alcohol use in the last 30 days, alcohol use in the last 7 days, five or more drinks in a row during the last two weeks, and ever been drunk. Other items were used to create an alcohol intentions scale and a drug use scale based on alcohol, tobacco, and marijuana use. In addition, mediating student measures included perceptions of norms supportive of alcohol use, perceived outcomes supportive of alcohol use, lack of self-efficacy to resist alcohol, parental involvement, and limited access to alcohol.

      The parental survey included items on monitoring of children, family alcohol discussion, alcohol access at home, perceived community action, support for policies to prevent illegal sales, support of advertisement policies, and perceived neighborhood problems.

      The community leaders survey, done like the parental survey at baseline and the last time point, measured neighborhood strength, neighborhood and police prevention action, and organization prevention efforts.

      Measures of alcohol purchase attempts came from the attempted purchase of alcohol without age identification by young-appearing women.

      Analysis: For data from all sources, three-level mixed-effect regression models included random effects to capture variation 1) over time, 2) across individuals, and 3) across school units. The models estimated growth rates in the outcome variables for control and intervention groups and used chi-square tests of differences across the groups.

      The models allowed for an intent-to-treat analysis that used cases with incomplete data across the four time points.

      Outcomes

      Implementation Fidelity. Direct observation of the classroom programs by research staff occurred in two to four sessions each year. Across all 3 years, scores indicated 82-87% completeness in coverage of the curricula and indicated high engagement of students (mean of 11-12 on a scale ranging from 5, no engagement, to 15, very high engagement). However, participation of students was less than ideal. Due to student attrition, average cumulative exposure to the class components reached was only 53% among the study cohort. Also, completion of the parent program by 73% in year 1 was followed by lower completion of 53% in year 2 and 51% in year 3. About 40% of the intervention cohort participated in at least one community service project. About half of the community projects were rated as having low levels of productivity and effectiveness. The authors concluded that mobilizing inner-city communities around alcohol issues proved considerably harder than in rural Minnesota.

      Baseline Equivalence. The authors reported (but presented no figures) that there were no statistically significant differences between the intervention and control groups at baseline on demographic characteristics.

      Differential Attrition. Among those completing the baseline survey, white students who lived with both parents were more likely to complete three or four surveys than others. The completers also had lower values on the alcohol use scale at baseline. Yet, attrition rates overall or by demographic characteristics and alcohol use did not differ across conditions.

      Posttest. The results examined changes in outcomes during and at the end of the 3-year intervention. Tests revealed that the models did not differ by gender, ethnicity, family composition, or baseline alcohol and drug use. This result justified analyzing all subjects together and generalizing across subgroups.

      The models showed no statistically significant differences between the intervention and control schools in the growth rate for any of the primary outcome measures (drug use, alcohol use, alcohol intentions) or mediating outcome measures (norms supportive of use, perceived outcomes supportive of use, lack of resistance self-efficacy, parental involvement, and limited access to alcohol).

      Dose-Response. Analysis of the relationships between the over-time exposure to each intervention component and alcohol and drug use outcomes offered a few positive results. Higher participation in the home program was related significantly to lower growth over time in drug use and lower growth over time in alcohol use. Exposure to the other program components (i.e., classroom curricula, classroom peer leader status, community service projects, or community involvement) failed to have significant relationships with growth in drug or alcohol use.

      Mediating Effects: None of the mediating outcomes was significantly affected by the intervention.

      Effect Size: With no significant intervention benefits, effect sizes can be assumed to be near zero.

      Bulleted Outcome:

      • Inner-city, largely minority students in intervention schools did no better than those in control schools on any of the primary outcome measures (drug use, alcohol use, alcohol intentions) or mediating outcome measures (norms supportive of use, perceived outcomes supportive of use, lack of resistance self-efficacy, parental involvement, and limited access to alcohol).

      Generalizability: The negative results apply to inner-city communities with largely minority populations.

      Limitations: The study is fairly strong methodologically. The major problem was difficulty in obtaining student and parent involvement in the program. A related problem was the high level of attrition from students moving to other schools. Additionally, five schools dropped after randomization, and the schools did not allow assessments of these students. Furthermore, the study showed no significant effects.

      West, B., Abatemarco, D., Ohman-Strickland, P. A., Zec, V., Russo, A., & Milic, R. (2008). Project Northland in Croatia: Results and lessons learned. Journal of Drug Education, 38, 55-70.

      This study is an assessment of the impact of Project Northland in Croatia. The evaluation design included both quantitative and qualitative methods. Pretest and posttest data were collected to compare intervention and control schools.

      Evaluation Methodology

      Design: Project Northland in Croatia was launched in 2002 after a series of community forums designed to introduce the concept to teachers, administrators, and parents. Following the forums, the project's Croatian partners systematically recruited 26 schools to participate in Project Northland. All program materials were translated into Croatian and were culturally adapted with assistance from age-appropriate youth in the Croatian city of Split who worked with the researchers on the translation. The first year of the three-year intervention was implemented in the spring of 2003, year two was implemented in the spring of 2004, and year three in the spring of 2005. A total of 13 primary schools within Split were selected to receive the intervention and 13 schools were selected to act as controls. Primary schools (which included grades 1 through 8 in Croatia) were selected for comparability of size and geographic distribution across the city.

      In the first year of the intervention, all sixth grade students within each school marked to receive Project Northland were included in the intervention. In the second year, all seventh graders were included, and in year three all eighth grade students were included. A total of 948 students received the intervention and 1,033 acted as controls. The student survey was administered at baseline and following completion of each curriculum year. Qualitative data collected included information from facilitated focus groups with teachers (n = 7 and n = 4) and parents (n = 4). Students provided comments about their experience at a city-wide conference and notes were taken to provide more information on their views regarding Project Northland.

      Sample: The intervention sample was 48.5% female and the control sample was 49.4% female. Although no additional specific information was provided regarding the race or ethnicity of the sample, the authors do state that within the sample of schools in Split, the student population is relatively homogeneous with respect to race/ethnicity and socio-economic status and reflects the Croatian population overall, which consists of 90% Croats, with Serbs being the largest national minority making up only 4.5% of the total population.

      Measures: The student survey included questions concerning alcohol use, as well as tobacco and other drug use, peer usage and normative expectations involving alcohol use, reasons not to use alcohol, parental communication with child, perception of ease of access to alcohol, attendance at activities with/without alcohol, and demographic factors. Most questions were either true/false or Likert scale responses. The "Tendency to Use" index was computed based on eight questions that focused on actual use and intention to use alcohol.

      Three focus groups were conducted with teachers and parents of students who participated in Project Northland at the completion of the three-year curriculum. Two focus groups were held with teachers and one was held with parents, in which they were asked to identify strengths and weaknesses of the program and their overall reactions to the program. In addition, they were also asked to suggest ways to improve the program in the future.

      Analysis: The intervention outcome was modeled using a mixed linear model, with random effects included to account for correlation between responses within a school. Intervention group and class year were included as predictors in the model as well as an interaction between the predictors. An F -test of the interaction was used to test whether the changes over time differed between the intervention and control schools. Two contrasts were used to compare the changes: 1) from baseline to year 2; and 2) from baseline to year 3 between the intervention and control schools. These analyses were repeated for each gender separately.

      Qualitative analysis focused on three major areas: 1) strengths and 2) weaknesses of the curriculum, and 3) recommendations for changes that might improve the program in the future. In the parent focus group, a fourth area focused on the extent of knowledge regarding the curriculum was explored, while in the teacher focus group the fourth area focused on problems in implementing the curriculum.

      Outcomes

      Fidelity Implementation. A process evaluation came from the qualitative results described below.

      Baseline Equivalence: Comparisons of intervention and control groups showed no significant differences on demographic variables, but differences at baseline on the outcome variables were not compared.

      Differential Attrition: Since the project did not attempt to follow a set of individuals over time but compared all grade 6 students in year 1, all grade 7 students in year 2, and all grade 8 students in year 3, differential attrition is not an issue. However, the study reports no information on the comparability of the samples across the 3 years. Further, two schools dropped out of the program but were not examined for differential attrition.

      Quantitative Analyses: The composite effect of the intervention was significant overall. Intervention schools demonstrated significantly smaller increases in the Tendency to Use Alcohol scale than did control group from baseline to year 2. Although there was a similar difference for the increases from baseline to year 3, it was non-significant.

      Results by Gender: The intervention did not significantly affect male students, but was highly significant for females. Females in the intervention group had significantly smaller scores on the Tendency to Use Alcohol index at years 2 and 3 compared to females in the control group. In terms of the individual items on the Tendency to Use Alcohol index, girls in the intervention group had significantly smaller scores than girls in the control group on reported lifetime drinking (year 2 only), past year (years 2 and 3), past 30 days (years 2 and 3), and past week (year 2 only); and on intentions to use alcohol as an adult (years 2 and 3), in the next 12 months (years 2 and 3), in the next 30 days (year 3 only), and in the next week (year 3 only). In contrast, boys in the treatment group only scored significantly better than the control group on lifetime alcohol use (year 2 only).

      Qualitative Analyses: One focus group was held with parents of children who had participated in Project Northland for all three years. Parents identified four major strengths of the curriculum: the program encouraged more parent/child discussions on the subject of alcohol use, encouraged parent/child discussions on other subjects, their child now placed pressure on them not to drink, and made them feel more reluctant to drink in front of their child. In terms of weaknesses, parents felt the program had a more obvious impact on younger children and its impact was less discernible as their child grew older. Many parents noted that their child seemed bored with the program in the eighth grade, and spoke about the challenges of preventing alcohol use within the Croatian culture. Parents offered a number of recommendations for improving the curriculum, including feelings that the program should be initiated earlier so children would be less likely to "age-out," making the program longer, and expanding the topic areas covered to include discussion of drugs and involving sexuality.

      Two focus groups were held with teachers who participated in Project Northland. Teachers identified seven major strengths of the program: 1) enhanced teacher-student communication; 2) encouraged older student mentoring of younger students; 3) employed active as opposed to passive learning; 4) integrated the community into the work of the schools; 5) engendered anti-alcohol sentiment among students; 6) encouraged students to challenge school and community policies involving alcohol use; and 7) provided emotional support for students experiencing trauma. Teachers also identified four weaknesses within the curriculum: 1) students were ready for the program at an earlier age (starting in fifth grade); 2) although younger students were engaged, students in the eighth grade became bored with the program and failed to take it seriously; 3) girls seemed more interested in the program than boys; and 4) parents were seen as more involved with children in the younger grades and less involved in the last year of the program. Teachers also indicated three problems in implementation: 1) they indicated they needed more time to implement the program, 2) they felt rushed to accomplish everything in the given time period, and 3) they commented on resistance from some teachers who felt over-burdened and resented having additional work added to the program. Finally, teachers offered a number of recommendations including starting the program at an earlier age, lengthening the program, and expanding the program to include discussion of larger issues such as civil society and values.

      Mediating Effects: None reported.

      Effect Size: None reported.

      Bulleted Outcome:

      • Intervention schools demonstrated significantly smaller increases in the Tendency to Use Alcohol score than did control group from baseline to year 2, but there were no significant differences at Year 3.
      • The intervention did not significantly affect male students, but was highly significant for females: Females in the intervention group had significantly smaller scores on the Tendency to Use Alcohol index at years 2 and 3 compared to females in the control group

      Generalizability: This implementation of Project Northland was conducted with a representative sample of Croatian demographics within the city of Split. The program was more effective for girls than for boys. It is unknown to what degree the results may be generalizable to the larger population of Croatia or to the U.S.

      Limitations: Students were not matched from baseline to subsequent data collection points (years 2 and 3). Subsequently, data was analyzed in the aggregate, which limits the ability to identify behavior change. The authors indicate that two intervention schools dropped out during the course of the study, although no analysis of differential attrition was performed. Finally, the intervention did not appear to be effective in reducing alcohol use or intentions to use alcohol among males.

      Center for Prevention Research and Development (2004). Project Northland evaluation for Iroquois-Kankakee Regional Office of Education. Unpublished report. Institute of Government and Public Affairs, University of Illinois.

      This is an evaluation of the original Project Northand, which was designed as a three-year program to be implemented in grades six through eight. The program included the Slick Tracy, Amazing Alternatives, and Powerlines curricula described in detail above. It should be noted that Project Northand has since developed into a six-year program, and as such this study does not constitute a replication of the outcomes presented above.

      Evaluation Methodology:

      The Center for Prevention Research and Development (CPRD) worked with the Iroquois-Kankakee Regional Office of Education (I-KAN ROE) to evaluate the impact of implementing Project Northland in 14 Kankakee County schools during fiscal years 2002-2004.

      Design: 14 schools in Kankakee County, Illinois, were selected for participation in an unspecified manner. A comparison group was not utilized in this evaluation, as the Illinois Department of Human Services Bureau of Substance Abuse's (BSAP) transition to the use of comparison groups occurred after the evaluation had begun. Instead, the Communities Can! School Youth Survey data collected from Kankakee County sixth graders in 1999 and eighth graders in 2001 (three years prior to the Project Northland data collection) was identified as the best available comparison data for the evaluation. It was possible to use this data set as a comparison group for the Project Northland evaluation because the youth who were surveyed were not exposed to the curriculum at the time of the survey. In addition, the Communities Can! School Youth Survey includes questions related to substance use that are comparable to the questions on the Drug Perceptions and Use Survey, which allows the comparison of substance use outcomes between the Project Northland participants and non-Project Northland participants.

      Students participated in the first level of the curriculum, Slick Tracy, in sixth grade, the second level of the curriculum, Amazing Alternatives, in seventh grade, and the third level of the curriculum, Powerlines, in eighth grade. At each grade level, students participated in Project Northland program sessions during class, completed program-related homework assignments that required parental involvement, and worked on projects in peer-led groups.

      The evaluation followed students in Kankakee County schools that began participating in Project Northland in sixth grade in 2002 and that completed the program in eighth grade in 2004. Data for the evaluation was collected through pretest and posttest surveys administered in 2002, 2003, and 2004 to sixth, seventh, and eighth grade students, respectively. Because Project Northland outcomes are expected only after completion of all three years of the program, data analysis was limited to the results of the pretest administered to a sample of sixth grade students prior to the beginning of the Project Northland curriculum in 2002, and the posttest administered to a sample of eighth grade students who had completed the final year of the program in 2004.

      Sample: No mention is made regarding the total number of participants in this evaluation. Information was provided, however, as to the racial/ethnic composition of the sixth and eighth grade Project Northland and Non-Project Northland participants. The sixth grade Project Northland participants were 38% Caucasian, 32% African American, 6% Latino, and 24% Other, while their eighth grade counterparts were 40% Caucasian, 40% African American, 8% Latino, and 12% Other. The sixth-grade Non-Project Northland participants were 59% Caucasian, 27% African American, 5% Latino, and 9% Other, while their eighth grade counterparts were 64% Caucasian, 5% African American, 5% Latino, and 26% Other.

      Measures: The survey asked questions for a period of the past month and a period of the last year on frequency of using alcohol, getting drunk, using cigarettes, and using marijuana.

      Analysis: Unspecified statistical tests determined if any differences in use rates were statistically significant. In cases where use rates were lower among the Project Northland eighth graders, additional unspecified statistical tests were conducted to rule out factors unrelated to the program, such as between-group differences at pretest. Additionally, use behavior trends were examined from sixth to eighth grade.

      Outcomes

      Implementation Fidelity: The Project Northland curriculum was provided to students by their health and physical education teachers, or by their regular teacher in schools that do not switch classes throughout the day. Life Education Center staff from the Regional Office of Education facilitated training of the school teachers who implemented Project Northland each year and worked with them throughout the implementation of the program.

      Baseline Equivalence: When the Project Northland sixth and eighth graders and Communities Can! sixth and eighth graders were compared, it was determined that the demographic distributions of the four groups differed significantly. Specifically, the comparisons showed that the racial/ethnic composition differed among the four groups and the distribution of free/reduced lunch status differed among the four groups. It was determined that the best way to address the demographic differences between the samples was to remove race/ethnicity as a factor by conducting separate analyses for the different racial categories. Thus, three separate analyses were conducted: one including only white students, one including only non-white students, and one including only non-white students with free lunch status. White Project Northland participants reported a significantly lower rate of alcohol use in the past month and cigarette use in the past year and past month at baseline than non-Project Northland participants. Non-white Project Northland participants reported a significantly lower rate of alcohol use in the past year at baseline than non-Project Northland participants. Non-white Project Northland participants with free lunch status reported significantly lower rate of alcohol use in the past year and month at baseline than non-Project Northland participants.

      Differential Attrition. No information was provided on the degree of attrition or the comparability of the sixth and eighth grade samples.

      Post-test (eighth grade):

      White students: Project Northland students reported a significantly lower rate of alcohol use in the past year and getting drunk in the past month than did non-Project Northland students. Rates of alcohol use from sixth to eighth grade increased among both groups, although Project Northland students reported a (non-significant) lower increase during this time. Similarly, rates of cigarette use increased among both groups from sixth to eighth grades, and unexpectedly Project Northland students demonstrated a (non-significant) greater increase in cigarette use than did non-Project Northland students (7% vs. 4%). Project Northland did not significantly impact cigarette use in the past year or month or marijuana use in the past year or month among white students.

      Non-white students: Project Northland students demonstrated a significantly smaller increase in the rate of alcohol use from sixth to eighth grade, a significantly lower reported rate of alcohol use in the past year and month, and a significantly lower rate of getting drunk in the past month compared to non-Project Northland students. Project Northland did not significantly impact cigarette use in the past year or month or marijuana use in the past year or month among non-white students.

      White students with free lunch status: Project Northland students demonstrated significantly lower rates of alcohol use in the past year and month and significantly lower rates of getting drunk in the past month than non-Project Northland students. Despite baseline differences in past-year and month alcohol use that favored the Project Northland students, Project Northland students had a significantly smaller increase in the rate of alcohol use from sixth to eighth grade as compared to non-Project Northland students. Project Northland students reported a significantly lower rate of cigarette use in the past year than non-Project Northland students. Finally, Project Northland students reported a significantly lower rate of marijuana use in the past year than non-Project Northland students.

      Long-term: No long-term data was analyzed in this evaluation.

      Mediating Effects: None reported.

      Effect Size: Not reported.

      Bulleted Outcomes:

      • In all three subgroups, Project Northland participants had significantly more favorable outcomes related to alcohol use in the past year and getting drunk in the past month than non-Project Northland students.
      • Past month alcohol use was lower for non-white students and white low-income students.
      • Non-white as well as low-income white Project Northland students had a significantly smaller increase in the rate of alcohol use from sixth to eighth grade as compared to non-Project Northland students.
      • White low-income Project Northland students had significantly more favorable outcomes related cigarette use in the past year, and marijuana use in the past year than non-Project Northland students.

      Generalizability: Because three separate subgroup analyses were conducted in this evaluation, the results among any one subgroup of students cannot be generalized to the entire population of participants. It is possible, however, to examine the patterns of results among the subgroups. Some results were consistent across all three subgroup analyses, which suggest that these results are the most robust findings of the evaluation.

      Limitations: There are serious questions regarding the equivalency of the treatment and non-treatment groups. Specifically, there were numerous significant differences in both racial/ethnic and socioeconomic status between the Project Northland and non-Project Northland participants, as well as significant differences in reported use patterns at baseline. The inability to include a control group in the design of the study is a significant weakness, as is the failure to collect equivalent comparison data at the same time that intervention data was collected. It is unknown to what degree, if any, the three-year difference in collection times may have affected the outcomes. No mention is made as to how the 14 Kankakee schools were recruited or selected for participation, and no mention of the final sample size in the evaluation. Finally, no specifics were mentioned as to which statistical analyses were performed on the data, so it is difficult to assess the rigor with which the results can be interpreted.