2018 Student Growth

The map below shows the percentage of students who had typical or high growth in math over the past three years. Mouse-over or tap the map to view school names and percentages, or use the filters to change to ELA, grade levels, or to a specific year.

Student growth percentiles are calculated by the Georgia Department of Education by comparing a student’s current year test score to those of other students across the state who had similar test scores in previous years. Typical or high growth is defined as growth in the 35th percentile or higher. We include a three year average in the map above because student growth results often fluctuate from year to year.

The next visual shows student growth at Barack and Michelle Obama Elementary (BAMO) over the past four years. BAMO had high growth this year in 5th grade math, where 88% of students had typical or high growth. In 5th grade ELA 60% of students had typical or high growth. This is below the state average (dotted line at 65%), but we can see from the overlapping blue confidence interval that the difference from the state is not statistically significant. BAMO is a relatively small school, so the confidence intervals are larger due to the small sample size1.

Note that this visual can be filtered to different APS schools or clusters and to different school districts. Select “All” on the school filter to view the district-level result.

Elementary schools only receive student growth scores in grade 4 and 5. Our school performance vs. poverty graphs are a helpful complement because they better represent accumulated achievement from earlier grades.

The next visual is the same view, but filtered to “All” schools to show the district-level results. We can see that APS tends to have growth similar to the state average and has made noticeable progress in 8th grade math over the past two years, from 52% in 2016 to 63% in 2018. Confidence intervals are not visible because the number of students is much larger than a single school.

Although district growth is close to the state average, small differences can impact many students. For example, in 5th grade math 4,115 students recieved growth scores and 71% had typical or high growth. If APS had performed at the state average of 65% typical or high, then 2,675 (65% of 4,115) of the students would have had typical or high growth. By performing six percentage points higher than the state average, an additional 247 students had typical or high growth.

The next view shows student growth at Drew Charter for end of course exams. Students at Drew grew well above the state average in Algebra and Geometry. Note that this view also includes results for eight grade students taking advanced math (Algebra).

To recieve a notice for similar posts, please subscribe to the blog.


  1. Confidence intervals are calculated at the 95% level using the standard error of the sample mean. This is only one of two sources of uncertainty. Each student SGP score is also an estimate from a statistical model run by GADOE and has associated uncertainty, but is not available to us. Given the large sample size for the GADOE model, this second source of error is likely smaller. But regardless, the confidence intervals are an underestimate of the total uncertainty. We compensate for this by using a relatively large confidence interval- 95%.