Ripple Effects | Fact Sheet: Research

Fact Sheet: Research

Both outside researchers and Ripple Effects researchers have conducted several studies on Ripple Effects software over the past four years. All have shown promising results, indicating that Ripple Effects software can boost academic performance, reduce problem behavior, increase prosocial behavior, and positively impact attitudes towards drugs. Study findings are summarized here. To download a one page summary of all research studies, click here. Contact us for more detailed information or full articles.

Social Capital Development Corporation/Ripple Effects, NIH/NIDA-funded quasi-experimental study (2003): better grades, behavior, and attitudes towards drugs
Key findings: Experimental group students had statistically significant increases in GPA (+.4), perceptions of risk of alcohol, and disapproval of alcohol and marijuana, and lower rates of fatalistic thinking. Data also showed dramatically less absenteeism, and fewer discipline referrals, than in the comparison group.

Summary: This randomized controlled study, conducted in the Fall of 2003, included 527 urban and rural, African-American (40%), Latino (31%), European American (23%) and Asian American and American Indian (6%) middle and high school students from six regular, alternative, charter, and continuation schools. The treatment studied was a culturally inclusive, multimedia behavior-training program from Ripple Effects.

Students were asked to individually explore 42 computer-based, interactive tutorials (10-14 hours total) over seven weeks. They were required to complete three of the eight to ten possible exercises for each tutorial. Facilitators were not all teachers; they also included two secretaries, a school custodian, a volunteer and a social worker. Topics for half of the tutorials were assigned by Ripple Effects, and the other half were chosen by adult facilitators from the individual schools, based on their perception of students’ needs and site-specific goals. In all schools, students were able to pursue topics beyond those assigned. Implementation rates averaged 60%. While still insufficient, this rate was obtained in “real world” conditions, in some of the lowest performing schools in a troubled school district.

Both the overall effects of the Ripple Effects software at sites where implementation was high (90% or higher), and the dosage-correlated effects at sites where overall implementation was low (30% or lower), support the hypothesis that Ripple Effects software has positive effects on risk and protective factors, including attitude (reduced fatalistic thinking, increased perception of risk of alcohol), behavior (discipline referrals, absenteeism, and tardiness) and performance (GPA and citizenship grades). Pooled effects–high implementation schools, compared to the control group:

GPA + .40 (scale 0-4)Fate -.14Norms (higher disapproval of drugs/alcohol): +.07, .22, .14. .19Risk alcohol (perception of risk): +15, +21, -.08
Risk marijuana (perception of risk): -.08, -.10, -.39, -.11 = counter hypothesis

Pooled correlation effects, low implementation schools, compared to the control group:

Unit increase in dosage (0-100) =
1. GPA increase of 1.07 (scale 0-4)Disapproval of trying marijuana increase by .047 (scale of 1-3)
2. Increase in perception of harm from smoking marijuana regularly .50 (scale 1-4)

Behavioral outcomes: Every school with high implementation had clinically important positive differences on behavioral parameters. Absenteeism. Pooled results at high implementation schools showed 42% fewer days absent in the treatment group than in the comparison group.This is a financially, as well as clinically, important number. Not only is absenteeism a marker of school bonding (or the lack of it), it is a major factor in school funding. Schools receive payment averaging $35 per student per day, or roughly $6300 per year, based on attendance. Reducing overall absenteeism by just two students per day could more than pay for the school-wide cost of this intervention.

Problem behavior. A factor that has even more direct social consequences than absenteeism, and increasingly is a measure of school safety, is problem behavior by students. The data from high-implementation sites show the trend is toward reduced problem behavior at every site, on every parameter, and the magnitudes of these reductions are clinically important. Experimental group discipline referrals were 60% lower, and detentions more than 30% lower, than the comparison group.

Eastern Kentucky University experimental study (2001-2002): positive effects on violence
Key findings: This study looked at the level of violence in three high schools and tested the effects of universal and targeted strategies to reduce this violence. Student reports of perpetration at the intervention school were significantly lower than the combined scores at the control school. Students, teachers, and administrators reported improved awareness, knowledge, and skills to implement violence prevention programs.

Summary: Conducted by faculty from Eastern Kentucky University, this study was designed to describe the level of violence in three high schools and to test the effects of universal and targeted strategies to reduce this violence. A repeated-measures design with two baseline scores and two intervention scores was used during a two-year period. Two rural high schools served as control schools with a single intervention high school. All freshmen at the three high schools completed a self-report measure of school violence; 420 completed the study at the end of their sophomore year. Multiple universal and targeted interventions to prevent school violence were used for slightly more than one year. Ripple Effects for Teens computer-based training program from Ripple Effects was the principal targeted intervention. It was used in an In School Suspension program. Scores on student Victimization and Perpetration, gathered one year apart, were compared using a pretest, posttest model. Student reports of perpetration at the intervention school were significantly lower than the combined scores at the control school. Students, teachers, and administrators reported improved awareness, knowledge, and skills to implement violence prevention programs. Researchers concluded that a comprehensive program that includes university and high school partnerships has the capability to reduce school violence during a short period. Outcomes attributable to Ripple Effects’ training software could not be separated from outcomes attributed to other parts of the comprehensive program (Renfro, Huebner, Ritchey, 2003).

New York experimental study (2000): gains in educational outcomes and prosocial behavior, reductions in aggressive behavior
Key findings; After using Ripple Effects' program for three months, two randomly assigned experimental groups both showed substantially more pro-social behavior (77%/32%), less aggressive behavior (32%/22%), and fewer remedial summer school referrals (42%/62%) when compared to a control group.

Summary: Faculty from Columbia University and New York University conducted this experimental, three-group pilot study. 54 students were randomly assigned to a control group that followed normal classroom practices, or to one of two experimental groups, who received a 12 session, self-directed intervention covering 24 tutorials in Ripple Effects for Teens, over 12 weeks. Topics were ones research has shown are correlated with reduced aggression, focusing on empathy, impulse control anger management and conflict resolution. A tutorial on learning styles was also included. One experimental group had an additional, adult-facilitated discussion and role-play session, once a week. The other experimental group did not. Researchers hypothesized that some observable behavior changes might occur in the group that worked on their own, but the most significant changes would occur in the group that had adult facilitation of discussion and role-plays in addition to the independent computer-based learning. “Blind” researchers measured social behavior of all groups post-intervention, using the Student Observation Assessment Tracking Form, which included both positive and negative behaviors. The negative behaviors were those commonly cited in school incident reports, such as pushing, shoving and name-calling. The positive behaviors were aspects of empathy, impulse control, problem solving and anger management. All have been correlated with reduced aggression in teens. Inter-rater reliability was high. Observers agreed 93% of the time when observing 24 students for five-minutes each using the 53 observation categories of the Student Observation Assessment Tracking Form. Pos-test data on remedial summer school referrals, as a measure of academic outcomes, was also collected. After using Ripple Effects' program for three months, both experimental groups showed substantially more pro-social behavior and less aggressive behavior when compared to the control group. In addition, both experimental groups had fewer referrals to remedial summer school. This was the first evidence that computer-based training focusing on social behavior could affect academic performance as well. Despite the apparently large program effect, statistical significance could not be demonstrated because of the small sample size of the study. More research was needed.

One striking finding was contrary to the original hypothesis. Students who used the program on their own had more than twice the level of positive behavior, and less aggressive behavior, compared to students who used the program along with supplemental adult-led sessions.

	Stand alone CBT vs. control	CBT + adult facilitation vs. control
Pro social behavior	+77%	+32%
Aggressive behavior	-32%	-22%
Remedial summer school referrals	-42%	-62%

This directly countered researchers' original hypotheses. It led to the formulation of new hypotheses about why this might be true, including adolescents reliance on peers rather than authority figures for guidance on personal matters; the safety provided by the private computer-based learning; and/or the better modeling provided in the program videos compared to teacher delivery (Stern & Repa, 2000).

San Francisco quasi-experimental study (1999): gains in assertiveness Key findings: The experimental group showed statistically significant positive gains in assertiveness after only one hour of training, while a control group showed no such change.

Summary: In an internally funded, quasi-experimental study, with experimental and comparison groups, comprised of two computer classes (n = 17 each) equated at pretest, ethnically diverse 9th and 10th graders in a San Francisco public schools were assigned five assertiveness related topics from Ripple Effects for Teens, to explore in small peer groups of two to five students. They were required only to do the interactive parts, what developers had identified as core components of the learning process, specifically journal writing exercises and interactive games that assessed concept mastery. They were tested before and after using the Children’s Assertiveness Behavioral Scale (CABS). CABS has been shown to posses discriminative validity and could accurately differentiate students who received 16 versus 8 hours of skill training in previous studies. The intervention studied here was only one hour, and used only 1/3 of the items measured on the CABS test.

The experimental group showed statistically significant positive gains in assertiveness after only one hour of training, while a control group showed no such change. This suggested the software had the power to transfer basic concepts. (Ray, 1999)

ImplicationsThese studies have landmark significance. Mounting evidence suggests that there is a flexible, scaleable way to positively affect social behavior among today's youth. Educators who have been buckling under the burden of teaching sensitive social content and skills they often have no training in, can be relieved of that strain without denying students critically needed skill training. Researchers and others interested in doing further studies should contact Sarah Berg. Find out more at the Research section of Ripple Effects web site.

Contact:
Sarah Berg
415-227-1669 x307