My oldest sister Laurie is a superintendent for an Ohio school district and she sent me data about all of schools and school districts in Ohio. Ohio uses a system of 16 different measures of school performance. To understand more about the different metrics used in Ohio, go here. The data presented in this report came from my sister Laurie, which originally received the data from the State of Ohio. Additional information was added to this, such as the revenue spent per pupil, which came from the Ohio Department of Education's website here.
The data as provided was challenging to work with in a number of ways. Achieving the desired capabilities required merging various tables of data together, specifically for the District Comparison report to function properly. Additionally, the raw data provided for the rankings of the measures were in a format of A-F grading as well as a number of "NR" designations, meaning that certain schools and districts were not measured on that particular category. In order to create meaningful insights into the data, these letter grades were translated to numeric values 0-5 (A=5, B=4, C=3, D=2, F=1, NR=0). This allowed trends and calculations to happen on these measures of school performance. However, care was necessary to exclude NR (0) values from the calculation of averages, etc. This conversion also allowed the creation of an overall measure that averaged the 16 measures into a single value. Once these numeric translations and calculations were made, additional work was required in order to translate these calculated values back to the A-F ranking system used by the State of Ohio.
Page 1, Ohio Overall, presents a snap shot of all school performance in the State of Ohio with the ability to filter using the county map in the lower left corner. The coloring of the county map is based upon the calculated measure of overall school performance with teal being better and pink being worse. This page also displays overall information on the amount of money spent per pupil, the number of districts, schools (buildings), students, % poverty and % minority.
Page 2, District Comparison, is a tour de force in the ability to edit visual interactions. Color coding has been added to provide the user a visual cue as to what slicers affect which visuals. This report was designed to allow educators to find districts that are similar in classification and poverty level as their own and is specifically designed to identify outliers that significantly overperform based upon their classification. By identifying overperformers, it is hoped that other educators can learn what these overperformers are doing well that they are not and thereby facilitate learnings by school administrators that will ultimately benefit students in lower performing districts. The way that overperforming districts are identified is through simple calculations involving standard deviation. By translating the letter grades to numeric grades and creating the overall measure of school performance, it is possible to calculate the average (mean) of that value as well as its standard deviation. By subtracing this mean from a school district's overall score and then deviding by the standard deviation, a number is calculated that is that school district's number of standard deviations above or below the mean. School districts with higher, positive numbers are overperforming districts and those with lower, negative numbers are underperforming.
Page 3, District Buildings, provides administrators with a snap shot of individual district and school performance similar to page
Page 4, Trends and Correlations, provides interesting insights into the relationship of overall performance with respect to things like % poverty, % minority, revenue per pupil and enrollment. Some specific insights include:
- % Poverty and overall performance appear to be strongly correlated. The higher the poverty within a district, the lower the test scores. This appears to contradict Ohio's state regulators who claim that poverty is factored into school rating scores. If poverty is indeed factored into the scoring, it is perhaps not factored in enough
- Revenue per pupil does not appear to have a meaningful effect on school performance. If any correlation can be drawn, it must be concluded that greater revenue per pupil actually negatively affects student performance. This can most clearly be seen with "Urban" schools
- There is a potential correlation between % Minority and overall school performance
- There is a potential correlation between the number of students enrolled and overall school performance
- Unlike "Small Town", "Suburban" and "Urban" schools, "Rural" schools do not show a correlation between % Minority and % Poverty
- Revenue per pupil has increased steadily from 1995-2012 (latest year data was available) with an odd spike in 2011 (currently unexplained)
Finally, page 5, County Map, provides a larger view of overall school performance by county with the ability to select individual counties and see vital statistics.
EDIT 10/4/201
eyJrIjoiNmE2ZDc0NTQtY2NmYS00YmYxLWI5NzYtMTRkMzFhM2U4ZWI4IiwidCI6IjRhMDQyNzQzLTM3M2EtNDNkMi04MjdiLTAwM2Y0YzdiYTFlNSIsImMiOjN9
Did I answer your question? Mark my post as a solution!Proud to be a Datanaut!
Sadly, no dynamic ability to set the gauge color based on a measure value. At least I am not aware of that capability with the default visualizations, and I prefer to not use custom visuals if I can at all help it.
