The service learning bandwagon is in full swing. According to Kielsmeier (2000), the findings of a U.S. Department of Education 1999 survey revealed that 83% of high schools and 64% of K-12 public and private schools had students participating in some type of service learning activity. Further, the May, 2000 issue of the Kappan and the August, 2000 issue of The School Administrator are both devoted to articles and research on service learning. Service learning projects are increasingly being promoted as a tool to increase student achievement. Hornbeck (2000) stated that students' SAT-9 achievement scores in reading, math, and science have gone up 40 percent over the past four years. Billig (2000) stated that while research is limited, there is evidence that service learning helps students academically and socially. Researchers at Brandeis University (1999) in an evaluation of 17 middle and high schools found a significant improvement in math grades and a modest improvement in grade point average.

The impetus behind this movement is a national organization based in Washington D.C. called Learn and Serve America which distributes grant money to state departments of education. The Department of Education in the State of Georgia has used these funds to implement a number of service learning projects that are funded through school and/or district grants. One such grant was received by Harmony Leland Elementary School, a K-5 school in Mableton, Georgia, with a predominantly Afro-American population.

The faculty and students were enjoying their first year as a Leonard Bernstein School of the Arts. The Leonard Bernstein approach combines the fundamentals of artistic process and classical education with effective teaching practices and curriculum design. As part of the curriculum all students are issued a violin and receive instruction once a week.

One other factor of significance was a change of leadership the previous year. The school had been classified as an at-risk school with poor student performance and attendance. The new principal was assigned a mentor, who had been a principal in the school system, but is currently retired. The new principal and the mentor worked very closely together. There was a dramatic improvement in the climate of the school during that year, due in part to the leadership provided by them.

As part of the plan to improve the learning environment, a decision was made to apply for a service learning grant. The purpose of the grant was to secure funds for a project that might improve students' writing skills and attitudes toward school. The service learning project was to research and write the history of the Mable House. The history would be written, illustrated and published both in book format and on a website. The Mable House was the center piece of a 19^th century plantation with a lot of local history. The name of the area "Mableton" is a direct result of the name of the house/plantation.
 
 

The research design is classified as causal comparative since an existing situation was analyzed and random assignment of subjects was not employed. All fourth and fifth grade students (N = 106) participated in some way as data gatherers, interviewers, illustrators, readers, writers, etc. The majority of the writing fell to a group of fifth graders (N = 35). The person who was assigned as the mentor to the new principal assumed responsibility for carrying out the project.

The following measures pre-and post- were administered to determine if the objectives of the grant were met:
 

- Iowa Test of Basic Skills (ITBS)
- Attitudes toward school

Table # 1

A Comparison of Fourth Grade Language Arts Scores with Students' Fifth Grade Scores.
 

Year	Mean	SD	N	t-score	Probability
Fourth	4.15	2.14	31	7.36	.000
Fifth	6.60	3.12	31

df = 30
 

As a check to see if this kind of gain occurred in other areas, these students' scores in reading were also compared. The gain in reading scores was .97 or about one year of gain. This was what it should have been. All students are expected to gain one full year or grade equivalent each year. Since they gained so much more than that in language arts, it is possible that the learn and serve project was responsible for the gain.

The measures of attitudes toward school (Student Opinions About School) were administered in the fall and again in the spring to determine if a significant difference in attitudes occurred as a result of the service learning project. The instrument consists of 20 items with three of the items collecting demographic data as follows: grade level, gender, and type student. A Cronbach Alpha revealed that the instrument has a correlation coefficient of + .50. Students respond to each item with "most of the time," "some of the time," or "almost none of the time." Scoring is a 1.0 for "most," a 2.0 for "some," and a 3.0 for "almost none."

A factor analysis was conducted on the data and it was determined that there were three factors measured by the instrument as follows: (1) five items that measured trouble behavior that could happen to a student at school, e.g., students get into trouble in the lunchroom; (2) five items that deal with helping behavior, e.g., teachers help with homework; and (3) seven behaviors that deal with behavior that is liked, e.g., students like to learn how to read. There is a negative correlation between the "trouble" factor and the other two factors. The correlation is very weak (.062) and it is not significant. However, this could account for the low reliability (.50) of the instrument.

A graph for each of the factors with pre- and post-scores shows very little difference between scores--see Appendix A. The behaviors students "like" resulted in an average score of 1.68 in the fall of the school year and a score of 1.64 in the spring. The behaviors that get students in "trouble" resulted in an average score of 2.03 in the fall and a score of 2.01 in the spring. The behaviors that involve "helping" resulted in an average score of 1.56 in the fall and a score of 1.56 in the spring.

Multiple analysis of variance (MANOVA) was used to compare the overall scores and the scores for each item. The mean pre-learn and serve project score for all 17 items was 1.74 and the post was 1.72. There was very little change in student attitudes toward school for the overall or for the three factors. In looking at each of the behaviors, there was a significant change for the behavior "students like to learn how to read." The mean pre-score was 1.71 compared to 1.5 for the post-score indicating that students were more positive on this behavior after the learn and serve project (p < .05) --see Table 2.

Table #2

A Comparison of Students' Attitudes Toward Reading Before and After the Learn and Serve Project
 

Condition	Mean	SD	N	F-score	Probability
Before	1.71	.70	106	5.18	.024
After	1.50	.57	105

p < .05
 

A comparison of students' attitudes toward school was also made for the demographic factors. The effect of gender on attitudes toward school created significant differences on two behaviors­see Table 3. Female students were more positive about their teachers helping with homework. With a score of 1.77, females were reporting that teachers helped with homework "some of the time" with 36% of females reporting "most of the time" compared to only 13% of the males. Male students, on the other hand, with a score of 2.28 had 41% of the students reporting that teachers helped "none of the time" with their homework compared to 13% of the females.

Male students reported more positive behavior on "wanting to do better."

Fifty-seven percent of the males reported that they wanted to do better compared to only 34% of the girls on this behavior. There were no significant differences as a result of grade level or the type of student.
 

Table #3

A Comparison of Students' Attitudes on Selected Behaviors Before and After the Learn and Serve Project
 

Behavior	Boy Mean	Girl Mean	F-score	Probability
Teachers help	2.28	1.77	13.9	.000
Want to do better	1.50	1.81	5.56	.020

N Males = 46
N Females = 53
 

Qualitative data were also gathered. Students were asked to complete the following open-ended sentences as many times as they wish:

I like going to this school because . . . . .

I would like going to this school better if . . . . . .

The gathering of this data was flawed, however, since the post-data could not be compared to the pre-data because the lead partial sentences differed as follows:

I like the students at this school because . . . . .

I would like the students at this school better if . . . . . .

The original intent was to compare the frequency with which likes and dislikes were recorded with the premise being that at the end of the project there would be more likes and fewer dislikes. Nevertheless, an analysis of the data does show that there is a supportive learning environment for the students. The pre-data concerning what students like about the school revealed that there were 180 positive comments or 1.68 per student. The majority of the comments (126) were school and instruction related. For example, 49 comments dealt with school being fun, exciting and giving you what you want to study. The most frequently occurring single comment dealt with 30 positive comments about their teacher.

The pre-data concerning what they would like to see changed revealed that there were 122 suggestions for improvement or 1.14 per student. However, there was no consensus regarding what needed to be changed. For example, only nine students wanted the students to be more polite; eight students wanted more recesses; six wanted nicer teachers; six were complaints about the conditions in the bathrooms, etc.

The post-data regarding what they like about the students at Harmony Leland revealed that there were 226 positive comments or 2.26 per student. The majority of the comments (127) dealt with manners. For example, 49 comments dealt with students being nice;13 with being friendly; 10 with being caring, etc.

The post-data regarding what they would like to see changed about the students at Harmony Leland revealed that there were 157 negative comments or 1.57 per student. The majority of the comments (107) dealt with manners and conduct. For example, 12 comments dealt with students fighting and not behaving; 12 comments dealt with wanting students to be nicer; nine dealt with wanting students to have better manners, etc. Fourteen students stated that they wished students did not pick on each other. All of these comments are closely related and deal with unwanted and undesirable behaviors.
 

The failure to improve students' attitudes toward school could be attributed to a number of factors. First, students' attitudes toward school are almost universally better in the fall because students are glad to be back in school with their friends. Second, many students have not yet experienced failure on a report card. At the end of the school year there are more failures and many students are tired of school and want it to be over. Third, since this instrument measures whether students get into trouble, there are more students in trouble at the end of the year than at the beginning. The preceding observations by Bulach are based on 26 years of experience as a teacher, principal, and superintendent in the public schools of Ohio.

The finding that there was a significant difference in students' attitudes toward reading was an unexpected outcome of the project. An improvement in writing was expected and it did occur. The ITBS data show that a normal pattern of growth (one year) in reading occurred. It is quite possible that the change in attitudes could be attributed to the project as it involved a lot of reading to research archives, old newspapers. Further, they had to critique/edit the written material that went into the book and on the website and this also required reading.

A major finding of this project is that teachers spend more time helping females with their homework than they do males. In related research, Bulach (2000) reported the following: (p.3-4)

- More males than females reported that people do not care about each other;
- More females than males reported that teachers help them to understand and get along with other students;
- More females than males reported that teachers help them to feel good about their school; and
- More males than females reported that students do not show respect for each other.

The fact that more males want to do better than females could be linked to the above behavior that teachers help more females with their homework. The extra help could result in more females receiving "A's" and the belief that they are doing good enough versus fewer males receiving "A's" and the need to do better.
 

Conclusions

The qualitative data about likes and dislikes validates the earlier statement concerning the positive learning environment and climate at Harmony Leland. However, it would be difficult to say that this condition is attributed to the impact of learn and serve project. The fact that this was their first year as a Leonard Bernstein School of the Arts could certainly have impacted students' attitudes and the climate of the school. Of greater impact perhaps, was the new principal who was in her second year at the school. She is very dynamic and exudes a positive energy force and the belief that all students can learn. Another factor perhaps was the presence of the mentor principal who spent a lot of time in the building and directed the learn and serve project. In retrospect, it appears that all of these forces came together to change an at-risk school into one where students and teachers are happy and a lot of teaching and learning is taking place.
 

Billig, S. H. (2000). The effects of service learning. The School Administrator, 57,(7), 14-19.

Brandeis University (1999). Summary report: National evaluation of learn and serve America. The Center for Human Resources: Waltham, MA

Bulach, C. R., Penland, J., & Williams, R. Bullying Behavior at the Middle School Level: Are There Gender Differences? Paper presented at the American Educational Research Association on 4-27-2000 at New Orleans, LA.

Hornbeck, D. W. A central piece of district-wide reform. The School Administrator, 57,(7), 8-11.

Kielsmeier, J. C. (2000). A time to serve, a time to learn: Service learning and the promise of democracy. Kappan, 81(9), 652-657.