May 27, 2009


Project SPARK (Sports, Play, and Active Recreation for
Kids) is a physical education program designed to be implemented by physical
education specialists for elementary students in the 4th and 5th
grades. The goal is to engage students in more moderate to vigorous physical
activity during the week to maintain their health and promote a lifetime of

Seven schools were randomly assigned. Two were led by
specialists, two by classroom teachers, and three served as control schools. It
was hypothesized the program would reduce obesity and improve academic

Engagement in moderate to vigorous physical activity was
found to be significantly different depending on the condition. When taught by
specialists the students engaged in the most number of minutes per week of
moderate to vigorous physical activity, followed by the program taught by
teachers, and then the control condition of regular physical education.

Project SPARK was not found to significantly impact
skinfold measurements (body fat). Spending greater amounts of time in physical
education was not found to undermine achievement. Indeed, student performance
was often better in the trained teacher condition, though not in the specialist


Target population: Upper elementary students, 4th
and 5th grades.

In the following studies, the specialist condition and
teacher trained condition implement Project SPARK. The specialist condition is
implemented with three physical education specialists who have state credentials
through a 5-year program. One specialist trains volunteer teachers in the
teacher trained condition schools to implement Project SPARK. The control
condition schools are asked to continue with their usual physical education

The specialist condition was the Project SPARK intervention
led by trained fitness specialists. The teacher trained condition is the same
curriculum, but led by teachers trained by one of the certified education
specialists from the specialist condition.

Project SPARK consists mainly of a self-management
curriculum. Students learn to self-monitor, self-evaluate, and self-reinforce.
They keep records of their activities, evaluate whether they have meet a weekly
goal as well as meeting to discuss goals and why the child failed to meet them,
if that is the case, and then they reinforce their active lifestyle by praising
themselves or making more time for physical activity or making physical activity
more enjoyable. This program encourages family involvement and physical activity
as part of one’s lifestyle and not only done during school hours.

Project SPARK is implemented three times a week for 30
minutes each time. Class is separated in half with the first 15 minutes
consisting of health-fitness activities, and the second half consisting of 15
minutes of skill-fitness activities. There is a fourth day per week (30 minutes)
where students learn how to incorporate activities and skills into their lives
to remain active. Units lasted about four weeks and students progressed to more
advanced skill within each unit.

Project SPARK consists of a physical education curriculum
and a self-management curriculum. The two curricula were separated into three
parts during a class. First, children warmed up and performed exercises targeted
toward maintaining strength and endurance. Second, children participated in a
fitness activity for cardiovascular strength. Third, children learned a sports
skill, followed by cool-down with stretching.

The teachers who volunteered to implement the
teacher-trained condition were about 20 fourth and fifth grade classroom
teachers. They participated in an in-service program that lasted about 23 hours
over 11 sessions. In the second year of the program, there were five sessions
for a total of 15 hours of training. The teacher-trained condition lasts for
about 11 classes during the school year, for a total of about 32 hours of
activity for students. In the following studies, Project SPARK time decreases to
nine hours in the second year of the study and six hours in the third year.

The control condition was the regular physical education
curriculum, but schools were also provided with the same equipment that the
Project SPARK groups received.


Sallis, J. F., T. L.
McKenzie, et al. (1993). Project SPARK: Effects of Physical Education on
Adiposity in Children. Annals of the New York Academy of Sciences, 699,

Evaluated Population: Fourth grade children (N=745
at baseline) in seven suburban elementary schools in Southern California. Mean
age of the students at year 1 was 9.25 years. The children were 85% non-Hispanic
white, 6% Asian and Pacific Islander, 7% Latino, 1% African American, and 1%
other race.

Approach: Schools were stratified by race and size,
then two were randomly assigned to one of three conditions, specialist, teacher
trained, and control. The authors added a third school to the control condition.

This study includes two measures, body mass index (BMI) and
a skinfold score. BMI is a combination of height and weight reported as kg/m2.
The skinfold score was a sum of mid-calf and mid-upper arm skinfolds measured
with calipers. Measurements were taken three times and the mean was recorded.

Assessment staff attended three four-hour workshops where
they learned the proper technique for measures, practiced measuring each other
while receiving feedback and guidance, practiced on elementary-age children who
were not in the study, and passed 90% of items on a checklist for each

Results: There were no significant differences among
treatment or control groups for boys or girls (genders analyzed separately in
this study) on skinfold measurements after two school years.

The authors did not account for clustering in their
analyses. They used children as the unit of analysis without accounting for the
fact that they are nested within schools.

McKenzie, T. L., J. F.
Sallis, et al. (1997). Long-Term Effects of a Physical Education Curriculum and
Staff Development Program: SPARK. Research Quarterly for Exercise and Sport
4, 280-291.

Evaluated Population: Fourth grade students in seven
elementary schools in one Southern California school district.

Approach: These analyses focused on observing the
quantity and quality of physical education class time through direct
observation. Quantity and quality were measured using the SOFIT (System for
Observing Fitness Instruction Time). SOFIT captures energy expenditure
associated with physical activity, frequency and duration of physical education,
time spent in various lesson contexts, and time spent in teacher behavior

Trained assessors spent entire school days observing at
schools in the three different conditions for one week during the fall and
spring semesters of each school year. They began collecting data in the class
when 51% of the students reached the instructional area and ended when 51% had
left. Instructors were paced by an audio tape to make assessments at certain

Assessor training included classroom lectures, video
assessment, and field practice. Certification was tested by the assessor
reaching 85% agreement criterion for variables on precoded videotaped lessons.
Assessors were retested on precoded videotapes of lessons before each
measurement period.

Results: Engagement in Moderate to Vigorous Physical
Activity (MVPA) and Very Active categories varied significantly by condition;
students in the specialist condition engaged in 40.2 minutes of MVPA per week,
compared with 32.7 and 17.8 minutes in the teacher-trained condition and control
condition schools, respectively. Lesson context varied significantly by
condition as well. Students in the schools assigned to the specialist condition
spent significantly more minutes in fitness activities and skill drills than
students in schools assigned to the teacher-trained condition. Students in
schools assigned to the teacher-trained condition spend significantly more
minutes than students in schools assigned to the control condition. Children in
the control condition spend significantly more minutes a week in free play than
children in the specialist and teacher-trained conditions. For teacher behavior,
teachers in schools assigned to the Specialty condition provided students
significantly more minutes per week of attention to physical fitness promotion
and general instruction than the teachers assigned to the control condition,
while instructors in the teacher-trained condition fell in between the two.

Sallis, J. F., T. L.
McKenzie, et al. (1999). Effects of Health-Related Physical Education on
Academic Achievement: Project SPARK. Research Quarterly for Exercise and
Sport 70,
2, 127-134.

Evaluated population: Students (N=759) were from a
single school district drawing from an affluent suburb in Southern California.
In the schools, the percentage of ethnic minorities ranged from 10 to 19. The
students were fourth graders recruited at the beginning of two consecutive
school years (2 cohorts).

Approach: Before random assignment, schools were
stratified by percentage ethnic minority and size of student population. The
authors randomly assigned schools to one of two treatment conditions or to the
control condition. Cohort 1 took the Metropolitan Achievement Test 6 and Cohort
2 took the Metropolitan Achievement Test 7.

The main outcome was academic
achievement measured by the Metropolitan Achievement Tests (MAT6 and MAT7). This
test has four parts, a basic battery, a language section, a mathematics section,
and a reading section. The basic battery tests overall achievement. Scores are
reported as national percentile rankings.

Results: Initial scores on the MAT tests were high
for both Cohort 1 and Cohort 2, and scores tended to decline in posttest.

For Cohort 1, there were no significant impacts of the
Project SPARK intervention on the basic battery or mathematics scores on the
MAT. For the language scores, the decline in percentile ranking was
significantly less in the Trained Teacher condition than the control condition.
For reading scores, the specialist condition increased in percentile ranking
while the control students declined (p < 0.05).

For Cohort 2, basic battery and reading scores of students
in the trained teacher condition declined significantly less in percentile
ranking than students in the other two conditions. Declines in language scores
were significantly greater in the specialist condition compared with the other
two conditions.

An important note is that the authors did not account for
clustering in the data. The individual students were the units of analysis, and
because measurements were different for the two cohorts, they were analyzed
separately. In addition, achievement data were only available for about half of
the students.



McKenzie, T. L., J. F. Sallis,
et al. (1997). Long-Term Effects of a Physical Education Curriculum and Staff
Development Program: SPARK. Research Quarterly for Exercise and Sport 68,
4, 280-291.

Sallis, J. F., T. L. McKenzie,
et al. (1999). Effects of Health-Related Physical Education on Academic
Achievement: Project SPARK. Research Quarterly for Exercise and Sport 70,
2, 127-134.

Sallis, J. F., T. L. McKenzie,
et al. (1993). Project SPARK: Effects of Physical Education on Adiposity in
Children. Annals of the New York Academy of Sciences, 699, 127-136.

Curriculum materials:


SOFIT (System for Observing Fitness Instruction Time)

KEYWORDS: Middle Childhood, Children, Elementary, Suburban, Co-ed, School-Based, Manual Is Available, Cost Information Is Available, Education, Academic Achievement/Grades, Reading/Literacy, Mathematics, Other Education, Physical Health, Health Status/Conditions, Obesity, Other Physical Health


Program information last updated on 5/27/09.