5th Grade Science Test Scores: What Do They Tell Us?
The top 10 performing schools, starting with first place:
Sea Gate
Pelican Marsh
Veterans Memorial
Lake Park
Tommie Barfield
Laurel Oak
Vineyard
Mason Classical Academy
Corkscrew Elementary Poinciana Elementary
Do test scores measure student achievement? To some degree, they do, but can we say what causes a school, district, or state to perform better? Many factors influence student performance on tests.
Let’s look at this year’s 5th grade science test results for Collier County School District (FCAT 2.0). This test is designed to assess what students have learned, based on the Florida Next Generation Sunshine State Standards science standards. Science standards have not been revised since 2008.
In the bar chart below, the average test score of the 10 highest-scoring elementary schools is compared to the 10 lowest-scoring elementary schools. The 5th grade science scores in these 20 schools ranged from a high of 217 to a low of 187. State and district averages were the same - both 200.
There are many possible explanations for these differences. Let’s start with some characteristics of the students, as reported in the CCPS school profiles. Some of these variables are: the number and percentages of English Language Learners (ELL), Exceptional Student Education students (ESE includes students with autism spectrum disorders, for example; gifted students are also ESE, but are not included), free/reduced lunch program students (an estimate of the number of students from lower income families), and migrant students. These data are gathered to help a school qualify for additional funding, because students in these groups need more resources to support their learning.
The school profiles also contain data on race, ethnicity, and gender, but this analysis focuses on factors which affect students’ abilities to learn, and are expected to influence test performance.
Race/ethnicity is sometimes used when information about students' socioeconomic status isn't available, because unfortunately, these variables are often highly correctly. In this analysis, we will use the school data on student participation in free/reduced lunch programs.
Race, ethnicity, and gender
