Methodology: How we did the research for DODDS-Europe test scores
Stars and Stripes examined standardized test scores and
student-teacher ratios for 115 Department of Defense Dependents Schools-Europe schools. 
The schools were bracketed into categories for elementary, middle
school and high school education.
Elementary education included all elementary schools, and in schools
that had students from all grades, Stripes examined test scores only for pupils in
kindergarten through sixth grade.
Middle-school education examined middle schools, and in combined
schools, examined test-scores for students in grades seven, eight and nine.
High school education analyzed data from high schools, and in
combined schools, examined the test scores from pupils in grades 10, 11 and 12.
Each school was ranked according to how well students scored on the
Comprehensive Test of Basic Skills Terra Nova/Multiple Assessment in reading, language
arts, social studies, math and science; by the percentage of students that performed on
the distinguished level on the Department of Defense Education Agency writing assessment
test; and on student-teacher ratio for the 1998-1999 academic year.
Data for the 1999-2000 year came out in May after the Stripes
analysis was complete. However, a spot check did not reveal any significant changes in
scores.
High schools also were ranked on SAT verbal and math scores and the
percentage of students at the school that took the test.
Some schools did not report scores to DODDS in some categories
because the numbers of students taking the tests were too small to be considered accurate
indicators of performance. These schools were given the average score for the DODDS system
in that category so the school wouldn’t be helped or hurt in the overall rankings.
The ranks for individual categories were then averaged to give an
overall score.
Larissa High School in Greece, Livorno Elementary/High School in
Italy, and Sevilla Elementary/Junior High School in Spain were not included in the
rankings.
Larissa just opened last year and did not have complete data. Livorno
and Sevilla have too few students in every grade — for example, last year there were
just two students in the sixth grade at Sevilla — to have reportable numbers.
Statistics professor Andrew Rothman at New York University said Stars
and Stripes’ method was a fair way to compare the standardized test results at
individual schools.
“You are looking at apples and apples,” Rothman said.
A parent trying to compare performance of a student at one school
with a student at another school would get an accurate snapshot by using Stars and
Stripes’ data, he said.
— David Josar
Back to August stories
Page Two news roundup
Stories from July, 2001
Stories from June, 2001
Stories from May, 2001
Stories from April, 2001
Stories from March, 2001
Stories from February,2001
Stories from January, 2001
Stories from December, 2000
Stories from November, 2000
Stories from October, 2000
Stories from August and September, 2000
Stories from June and July, 2000
Home |
