Florida Learn & Serve 1997-98 Outcomes and Correlations with 1994-95, 1995-96 and 1996-97
In 1999 Florida Learn & Serve produced a report of the fourth year of research on quantitative outcomes of K-12 students who participate in service-learning projects in Florida. This webpage presents the results of that study, which are also available as a PDF file for download. The projects studied took place in 1997-98 throughout the state as part of Florida Learn & Serve K-12, supported by the Corporation for National Service as part of the Learn & Serve America program under grant number 97LSFFL003.
As the report marks our fourth year of data collection and analysis, we have been able to compile a multi-year report on the effects of service-learning on key elements including students’ absences and discipline referrals. The study, involving 117,187 youth participating in 382 sub-grants, provides increasingly comprehensive evidence of the positive effects of service learning on student outcomes in Florida.
Executive Summary
Introduction
Methodology
Results and Outcome Data
Levels of Youth Service-Learning Participation
Magnitude and Value of Youth Service-Learning Participation
Impacts on Attendance of Participating Students
Impacts on Discipline Referrals
Limits of the Data
Conclusions
Executive Summary
Florida Learn & Serve’s (FL&S) K-12 school-based projects support youth service learning through the Corporation for National Service’s Learn & Serve America program. In 1997-98, FL&S awarded 82 sub-grants. Of that total, 77 school-based sub-grants engaged an estimated 31,482 students in service that applied academic learning. In addition, there were five adult volunteer initiatives.
School-based 1997-98 efforts resulted in an estimated 415,514 hours of youth service; each sub-grant involved an average of 355 students. Calculated at $5.15 per hour, this service would be worth $2,139,897 or 2.34 times the $913,652 awarded. In key areas of attendance and conduct, majorities of participating students made improvements while participating:
- Attendance improved in 83% of reporting sub-grants
- 80% of reporting sub-grants had fewer discipline referrals.
FL&S has also collected outcome data from its 1994-1995, 1995-96, and 1996-97 sub-grants. From an overview of the four years of data collection, parallel trends are evident. The graphs that follow highlight the key areas of interest and include side-by-side comparisons of final report outcomes from the past three years. The report also examines four-year trends and addresses issues raised or left unresolved by the figures.
In selecting the data elements to be addressed, FL&S balanced the need for verifiable results with the hazard of asking for more information than sub-grantees had time or tools to provide. Data are self-reported, although the main elements are systematically collected by schools. The range of projects precludes their results being combined. Key data elements are therefore presented in the form of trends--percentages of sub-grantees responding to each variable--and sub-grants should be examined on an individual basis to determine what impact an individual service-learning project had at a single school.
Introduction
In 1997-98, Florida Learn & Serve K-12 (FL&S) awarded 82 sub-grants to Florida public schools and school districts for activities in which students provided volunteer service as an application and/or means of learning. Fifteen of the sub-grants created district or school youth service-learning councils that awarded an estimated 250 additional service-learning minigrants. Another 5 sub-grants were adult volunteer projects.
The total award amount was $913,652, or an average of $11,142 per sub-grant. The average minigrant was $750. Funds were awarded to Florida from the Corporation for National Service and supplemented with state match and in-kind (see “Background on the National Service Program” on p. 6). Applicants were required to provide match or in-kind contributions on the following sliding scale: 30% for first -, 50% for second -, and 100% for third-year applicants. Most activities took place between November 1997 and May 1998, but a number of sub-grants received extensions until December 31, 1998. A few projects were extended until June 30, 1999.
Methodology
During the grant period, project staff sent outcome data forms to be completed and submitted by sub-grantees in their final reports. The forms supplemented a project narrative requirement with nine questions concerning numbers of participants, hours of service; before-and-during comparisons of absences and referrals, and other items(see Appendix). Participants are defined as youth who actively provided service. Data were self-collected and self-reported, but based on information that is collected by schools.
A total of 39 of the projects submitted the outcome data forms in time for inclusion for this report. Project staff then transcribed information from final reports to a spreadsheet. Not all sub-grantees responded to every item, so a spreadsheet was developed for each of the key variables. The three independent spreadsheets (variables) compared before-and-during measurements of participating students’ (1) service hours, (2) attendance, and (3) discipline referrals.
Results and Outcome Data
Youth Participants
All reporting sub-grantees (100%) responded to this item. 13,847 students
participated directly (i.e., provided service) on these 39 projects, for
an average of 355 students per sub-grant. Extrapolating that average to the
other 28 school-based projects, an estimated 23,787 students participated.
In addition, 15 youth service learning councils involved an estimated 7,695
students. Therefore, the total estimate of youth participants in 1997-98
was 31,482.
Inclusion of previous years results in a total of 117,187 youth who have participated from 1994-1998 for an average of 29,297 participants per year. On average, 307 students participated on each of the 382 sub-grants.
Hours of Service Performed by Students
35 of 39 sub-grants responded, reporting an average of 5,053 student
service hours (176,868 total) per sub-grant. Extrapolating the hours of service
to the other 32 school-based projects results in an estimate of 338,564 service
hours. Also, the youth service-learning councils involving 7,695 students
at 10 hours per student produced an estimated 76,950 hours for a total of
415,514 student service hours. Calculated at the current minimum wage of
$5.15 per hour, this student service is equivalent to $2,139,897 worth of
work or 2.34 times the total amount awarded (see Graph
1).
From 1994-98, an estimated 1,389,906 hours of student service were provided, for an average of 347,477 hours of service per year. In addition, service students provided during the four-year period “cost” $3,559,208 (the total amount awarded) but would be “worth” $7,158,016 (2.01 times as much) if calculated at $5.15 per hour.
Graph 1 - Florida
Learn & Serve Student Service Hours,
Funds Awarded, and Service Values*
(Project Years 1994-1998)
* Student service hours valued @ $5.15 per hour.
Attendance of Participating
Students
29 sub-grantees responded to the questions comparing participating student
absences before and during their sub-grant. 16 of 29 (55%) showed a decrease
in student absences. One (3%) reported no change in attendance, and 12 (41%)
reported an increase in absences (see Graph 2).
Graph 2 - Percent
of Florida Learn & Serve Projects
with Decrease in Absences of Participating Students
(Project Years 1994-1998)
Numbers
of Discipline Referrals
26 sub-grantees provided information on numbers of students receiving
discipline referrals before and during their service-learning sub-grant.
19 of 26 (73%) sub-grantees indicated a decrease in student referrals (see Graph
3). One of the 26 or 4% showed no change, while 6 (20%) reported an increase
in referrals.
Graph 3 - Percent
of Florida Learn & Serve Projects
with Fewer Discipline Referrals of Participating Students
(Project Years 1994-1998)
Limits of the Data
School-based service-learning activities do not occur in a test tube. Other factors impact participating students’ grades, attendance, and conduct. These studies neither identify those factors nor measure their impact on participating students. It is fair to question the correlation between participating in service learning and academic and behavioral improvements.
Sub-grant activities ranged greatly in type, scope/scale, duration, number of activities, and level of curricular integration. The key data elements (attendance and discipline referrals) are formally collected by schools, which lends credibility to their accuracy. However, the fact that the outcomes are self-reported--usually by already overextended teachers--is a methodological concern. The researcher does not know, for example, the precise length of the before- and during reporting periods used by sub-grantees or how they were calculated.
In designing the data elements and final report forms, project staff walked a fine line. On the one hand, data elements need to be precise enough to ensure valid reporting and outside scrutiny as well as measure intended outcomes of diverse sub-grants. On the other, requesting too much detail and verifiability would burden sub-grantees to the point that most would not return the forms or would leave major portions blank. The form used (see Appendix) is a compromise that achieved a 48% return rate by the date of publication of this report.
For these reasons, key data are presented in terms of trends--percentages of sub-grantees responding within each element--instead of being combined. To determine what kind of impact one project had on individual students at a certain school, the data must be examined school-by-school.
Conclusions
The clear trend from 1997-98 Florida Learn & Serve sub-grant final reports is that there are quantifiable positive impacts on K-12 students engaged in service learning. By clear majorities, participating students showed improvements in attendance and conduct. Moreover, the 1997-98 outcomes are similar to and validate results obtained from the previous three years, suggesting these outcomes have validity and may be replicated elsewhere.
The large statistical sample over four years—117,187 students, 382 projects, multi-grade levels, a wide variety of projects, different project durations, degrees of curricular integration, etc.--lends weight to the trends observed but also highlights caveats and raises interesting questions:- What impact did activities have on service recipients?
- Are there particular elements of these sub-grants that brought positive results, or are the trends the result of a constellation of factors associated with the service-learning efforts?
- Would impacts be greater with greater curricular integration? (Data from 1994-95 found no such correlation, but there was also no scale for assessing the level of curricular integration.)
- How far can the data be trusted when it was not collected under rigorous conditions?
- What if most of the sub-grantees who did not submit data were the less or the more successful ones? Does that mean the remainder skew the overall trends?
This study cannot answer such questions; more long-term research is needed. In the absence of hard data, service learning has leaned uncomfortably on qualitative measures. Because of its combination of service with education, the truest measure of the value of service learning will need to be a combination of quantitative and qualitative measures over time.
(Some
of the above files require the free Adobe
Reader.
