Anderson.GIS assignment 1

Ina Anderson
Intro GIS- Assignment 1
9/18/09
Examining Regional Science Attainment and School- Teacher Resources Using GIS
Mark C. Hogrebe, Lydia Kyei-Blankson, Li Zou; Washington University in St. Louis
Education and Urban Society, Volume 40 Number 5, July 2008, 570-589 © 2008 Corwin Press, Inc.
Purpose
This is a complex study that attempts to show the relationship between the attainment of science
proficiency of 3rd and 10th grade students in the greater St. Louis region with a wide number of school,
teacher and student variables. The researchers hoped to demonstrate how the broad effects of school factors
and teacher characteristics impacted student achievement in science. Previous findings on this type of
research conducted around the country to date have yielded mixed results, making it important for St. Louis
to attempt its own analysis. Findings from this research could assist in making education policy decisions
where the relationships of certain variables on student achievement are clear.
The study contains both a statistical analysis and a GIS component. First, a statistical analysis was
conducted to investigate the relationship of science attainment of students in a number of school districts
in the St. Louis region with school variables of enrollment, percentage of students
receiving free/reduced lunch, instructional expenditures per student, teacher/pupil
ratio, and teacher variables of percentage with master’s degrees, salary, and years
of experience. Second, GIS was used to produce spatial and geographic representations of the
relationships between the school, teacher and science attainment variables and St. Louis area school
districts. This presents a geospatial picture of “how educational data are distributed across the
region and shows district variation in student science attainment in relationship to
where
school and teacher resources are located”(Hogrebe et al. 2008, 570).
Data
Data for this study was gathered between 2000-2005. Data on science attainment was obtained from
the 30 public school districts with the highest concentration of school age children based on the 2000


census. The districts are located in St. Louis City and St. Louis, St. Charles and Jefferson counties,

Missouri. The indicator of student attainment was the percentage of students attaining proficient or advance
level scores in science as assessed by the Missouri Assessment Program (MAP) for the 3 rd and 10th grade
levels. The researchers calculated the median test score to summarize each district’s percentage of
proficient/advance students on the science MAP. Data on teacher and school indicators were obtained from
the Missouri department of Education website in 2005.
Methodology
The researchers used Pearson correlation analysis to determine the strength of the relationship
between the selected school and teacher resources and the percentage of
proficient/advanced students in science. To demonstrate the variation in science
attainment across the school districts as it related to the selected school and teacher
factors, the data was presented spatially and geographically using GIS. Only
variables with significant relationships to percentage of proficient/advanced students
were used to create GIS maps. Maps were then created to display both the
Correlation of Instructional Expenditures and Free/Reduced Lunch With
Percentage of Proficient/Advanced Students as well as Correlation of Teacher
Variables With Percentage of Proficient/Advanced Students.
To display the Correlation of Instructional Expenditures and
Free/Reduced Lunch With Percentage of Proficient/Advanced Students, the
researchers created a series of maps (Figures 2 and 3). To indicate the percentage of proficient
/advanced science attainment across districts the researchers used the summary percentages for each district
and divided the percentages into lower, middle and upper thirds, assigning each level a varying shade of
gray or black on the map. This created the background against which the variables were shown. They
created one map for 3rd grade and one for 10th grade. The first variable, instructional expenditures per
student, was also divided into thirds and coded with an icon. This layer was added to the background to
create the maps in Figure 2. When juxtaposed with the proficiency data, it is easy to see the relationship
between the expenditure per student with the percentage of science attainment geographically across
districts. The same process was repeated with Figure 3, this time adding the variable of percentage of free/
reduced lunch students, divided into three levels and coded with an icon, then added to the separate 3 rd and
10th grade maps with the proficiency percentages in the background.
In Figure 4, the researchers displayed geographically the Correlation of Teacher

Variables With Percentage of Proficient/Advanced Students. To create these
maps, similar shades were used to designate the science proficiency of students
across districts in the three categories lower, middle and upper thirds. Three teacher
variables (percentage of teacher’s with masters, teacher experience and teacher
salary) where then represented together as a three- bar graph. This was a neat way
of displaying what would be three separate layers in one layer, within one icon. The
left bar is for percentage of teachers with master’s degrees, the middle bar for
teacher experience, and the right bar for teacher salary. The height of the bars
corresponds to the position in the variable’s distribution, with a low bar indicating the
district was in the lower third of all districts on this variable, and the converse for the
high bars. This icon was then layered over the background map to show the effects of
these teacher variables geographically across districts against student proficiency.
Results
The results presented in this study (far too many to discuss in the context of
this report) were mixed as to the magnitude and consistency of the findings across
variables in relationship to student achievement. The researchers point out that this
is consistent with their literature review of past studies of this type. The statistical results
did show that one of the school variables (instructional expenditure/student) and two of
the teacher variables (years of experience and salary) had significant relationships
with 10th-grade percentage of proficient/advanced students, although not at the 3rdgrade level. At the 10th grade level, a greater percentage of proficient/advanced
students in science was associated with higher instructional expenditures per
student, lower percentage participation in the free/reduced lunch program, teachers
with more years of experience,
higher salaries, and greater percentage with master’s degrees (Hogrebe et al, 2008, 18).
The maps created for the study, however, do show geographic relationships that could still be of
importance to education decision makers have great implications in terms of geographic equity in funding,
distribution of teachers with certain skills and student’s educational attainment. There relationships are
more easily apparent visually in map form than in tables. If nothing else, the visual patterns can spark

questions or eliminate previous assumptions that might help inform educational policy and reform. I would


like to see further study about how and why proficiency drops drastically between the 3rd and 10th grade in
all districts, and what the variables are that could effect this drop. This is essentially a non-spatial question,
but could be answered spatially by comparing regions and resources as the authors have done in the present
study, and adding a time factor, along with other potential variables that effect achievement over time, such
as mobility, truancy and other social factors. In conclusion, what I liked most about this study is
that it uses GIS to take a large, complex series of statistical variables and condense
them into simple, visual icons presented spatially and geographically, allowing for
comparisons that would be difficult to see and comprehend in tables. This is
particularly useful for social science, placing the ‘numbers’ into a more useful context
familiar to those who actually work in the districts studied.


QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.


QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.


QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.