2009 Research Report for G2

Research conducted by: Richard Ryan, Ph.D, Professor of Psychology, University of Rochester; Neta Weinstein, Ph.D, University of Rochester; Inger Williams, Ph.D, G2 Consultant.

In our ongoing effort to understand the impact of Generation 2 on students, their classroom teachers and our volunteers, at the direction of the Board of Directors the research committee of G2 initiated a longitudinal research strategy for the 2008-2009 academic year. In the prior year’s reports we had documented that children who were receiving G2 were showing greater interest and enjoyment of school on those days when the program occurred compared to other week days, and that this positive effect irradiated to their classroom attitudes and their relationships with their classroom teachers.

pull-researchTo build on these results in the current year we assessed children’s well being and school attitudes in the beginning of the year (October) and again at the end of the G2 year (late May) to examine whether G2 was having any long-term effect on the children. We also wanted to see how children in G2 classrooms were faring relative to those who did not have the benefit of the program. We decided to use multiple reporters, so we had not only the children’s self-reports on the two occasions, but also classroom teacher ratings on each individual child both pre- and post-program. In addition, thanks to the cooperation of the Fairport Schools, we were able to assess several non-G2 classrooms to use as a comparison group in order to anchor these change scores. In addition to teacher and child ratings, we also obtained a separate set of ratings from volunteers, although because their data was collected anonymously we could not do the pre/post comparisons on individual children. In the case of volunteer ratings we therefore used aggregated (or group averaged) data to compare G2 children pre- and post-program.

Data from each of these sources is presented separately in the pages that follow.


Teacher Ratings

The first four graphs in this report show results based on the pre- and post- program ratings of individual children provided by their classroom teachers. Because teachers were rating individual children rather than programs we suspect minimal bias, in addition to the fact that the ratings were separated by over seven months. The data combine ratings from all G2 classes in both the Rochester City School District (RCSD) and the Fairport School District. The non-G2 data comes exclusively from the Fairport comparison group, because we were not able not obtain comparable data from the city schools.

Because the city means on all variables are lower, these graphs most likely underestimate the positive effects of the program, since the non-G2 data would likely be worse if an RCSD comparison were included. The inclusion of RCSD data in the G2 group in fact explains the lower baseline means of the G2 group on several outcomes. Moreover, when we remove the city G2 data from these graphs, we see even more powerful trends, so we have retained all participants in these analyses in order to be most conservative.

These data are very promising. They suggest that in terms of school enthusiasm and prosocial attitudes classroom teachers are reporting gains in G2 children across the year. By contrast, non-G2 children are not becoming more prosocial, and they are losing enthusiasm for school (which by the way is a trend found in most studies of elementary school children over time). Moreover, non-G2 children showed declines in wellbeing over the year, and increases in problematic behaviors. Children in the G2 program, in contrast, evidenced no declines in well-being, and no increases in problematic behaviors.

Each of these results is graphed below, and is accompanied by a more specific explanation of the moderation analyses we performed (in italics).

School enthusiasm. Results show no change in school enthusiasm across the school-year F(1, 157) = .24, p > .05 for the sample as a whole. On the other hand, interactions of time with the G2 intervention, F(1, 157) = 4.84, p < .05, showed that teachers of classrooms who participated in G2 viewed their children as more enthusiastic about school at the end of the year (p < .05), while those whose students did not participate in G2 were less enthusiastic about school as the year wore on (p < .05).

Well-being. Main effects were present showing that, overall, children’s well-being decreased throughout the school-year, F(1, 157) = 31.15, p < .01. Marginal moderation effects were present, F(1, 157) = 3.48, p < .06, showing that decreases in well-being only took place for students who did not participate in G2, p < .05 (G2 students, p > .05).

Well-being. Main effects were present showing that, overall, children’s well-being decreased throughout the school-year, F(1, 157) = 31.15, p < .01. Marginal moderation effects were present, F(1, 157) = 3.48, p < .06, showing that decreases in well-being only took place for students who did not participate in G2, p < .05 (G2 students, p > .05).

Problematic behaviors. Levels of problematic behaviors did not change throughout the year, F(1, 157) = 1.21, p > .05 for the sample as a whole. However, a marginal interaction, F(1, 157) = 3.31, p < .07, showed that children who participated in G2 engaged in fewer problematic behaviors at the end of the year (p < .05), though no change was present for those who did not participate in G2.


Volunteer Reports

As noted volunteers data could not be matched on a child to child basis, so instead we present below the aggregated data comparing G2 children both pre and post. Obviously, there are no volunteer reports on the non-G2 comparison children because they did not interact with them. Again specific analyses are described in the caption printed in italics.

Main effects showed that G2 students increased in enthusiasm during G2 activities, F(1, 95) = 9.92, p < .01, well-being, F(1, 95) = 17.20, p < .01, intrinsic motivation, F(1, 95) = 12.11, p < .01, tendencies to engage in prosocial behavior, F(1, 95) = 30.28, p < .01, and proactive behaviors, F(1, 95) = 48.54, p < .01, from the start to the end of the school year.


Student Self-Reports

Finally we developed a very simple rating scale for use by children using a pictorially anchored rating scheme, which is a standard approach in this pre- or early reader developmental range. We inquired about feelings of happiness, liking of school and liking of teachers, since our expectation was that the G2 program positively affects both affective outcomes and classroom attitudes (results supported by the teacher-ratings reported above). Non-G2 comparisons are showing declines in happiness, liking of school and liking of teacher across the year, whereas G2 program participants show no declines in happiness or liking of school. They also showed increased liking of their classroom teachers. Again the inclusion of RCSD pupils in the G2 group lowers their overall means so these interaction results may underestimate effects. Specific analyses are described in italics.

I like school. In general, students reported liking school less throughout the school-year, F(1, 178) = 6.38, p < .01. An interaction, F(1, 95) = 7.21, p < .05, showed that in fact students who participated in G2 reported no change in their feelings toward school (p > .05), while students who did not participate carried this negative effect (p < .01). I like my teacher. Across students, no change was found in student perceptions of teachers, F(1, 178) = 0.04, p > .05. However, moderation analyses, F(1, 178) = 5.73, p < .05, showed that students who participated in G2 had stronger relationships with their teachers as the year progressed, p < .05, whereas students who did not participate in G2 liked their teachers less, p < .05.

I am happy. Collectively students reported feeling less happy throughout the school-year, F(1, 178) = 5.46, p < .05. Moderation analyses showed that this decline in happiness took place only for students who were not part of the G2 program, p < .05; whereas no change was present for students who participated in the program, p > .05.

To Top