Using Student Growth Percentiles for Measuring Student Growth - Promises & Challenges - Michigan School Testing Conference February 17, 2016

Oct 3, 2026

Using Student Growth Percentiles (SGPs) to quantify individual student growth over time is presented alongside key conceptual assumptions and evaluation implications. The material explains how SGPs compare a student’s growth to peers with equivalent prior test scores, then assign a percentile rank based on current-year ordering, without accounting for other achievement drivers. Illustrative scenarios contrast students with the same current score, and results are discussed using regression residual comparisons and correlation evidence. Challenges are highlighted, especially around defining “academic peers.”

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages50

Reading Time2 h 6 min

AccessFree to download

Student Growth Percentiles

Conceptual definition and limitations

Promises & challenges in use

Example walkthrough with matched scores

Comparisons with multiple years of test data

Growth trajectories and correlation evidence

Using Student Growth Percentiles for Measuring Student Growth – Promises & Challenges Michigan School Testing Conference February 17, 2016 Ed Roeber & Dave Treder [T]he phrase value-added models is used as an umbrella term to refer to a variety of “true” value added models, student growth percentiles, and certain growth models that are used for evaluation….[that are] are employed in an attempt to determine teachers’ and leaders’ contributions to student learning outcomes, as captured by standardized tests. Student Growth Percentiles Student Growth Percentiles 􀁽 Student Growth Percentiles (SGPs) represent one way to quantify the growth of individual students over time; 􀁽 Conceptually, SGPs describe students’ growth compared to the growth of other students who had equivalent past test scores. 3 In order to calculate SGPs 􀁽 Students are grouped with other students throughout the state who had equivalent scores on the previous test; 􀁽 Students are then ordered in their group based on their score on the current year test; 􀁽 Each student then receives a percentile rank based on their order in the group; 􀁽 By design, student growth percentiles make no attempt to consider other factors that contribute to student achievement growth. SGPs – A National…Phenomenon Add Georgia, Nevada, Virginia, Utah (and, I'm sure, others) A brief, simple walkthrough Two Students, both took the grade 4 reading test: both got a 434 One way to think about Jane's and John's growth is to look at students who scored the same as they did on the (previous year's) 3rd grade reading test: Jane hada 301 John hada 364 Or, maybe we could think about Albert, who lives in New Jersey.. Michigan's SGP presentations appear to exercise caution in the use the term academic peers – other states aren't as careful. Back to John and Jane Jane 90 SGP John 20 SGP Jane is compared to the kids who got a 301 on the 3rd grade test John is compared to the kids who got a 364 on the 3rd grade test With two years of Test Scores, most methods will give very similar results SGPs X Regression Residuals* District-Level Correlation (Number of Districts=85) r = .995 (Student Level r = .940 *Grade 7 to Grade 8, Math Fall 2013 MEAP – M-STEP 2015 (Genesee, Lapeer, Oakland, Ottawa, & Macomb ISDs, N = 28,340) With 3 (or more) years of test scores, it gets a bit trickier… | YR 1
PCTILE
RANK | YR 1 to
YR 3
SGP | | --- | --- | | 67 | 63 | | 46 | 74 | | 40 | 78 | | 34 | 80 | | 31 | 82 | | 29 | 83 | | 26 | 83 | | 11 | 89 | Student "Growth Trajectories" Different Year 1 Score, Same Year 2 & Year 3 Score 58 42 | | Yr1-Yr2 | Yr1-Y2-Yr3 | | --- | --- | --- | | Yr1-Yr2 | | 0.01 | | Yr2-Yr3 | -0.21 | 0.96 | MEAP Fall 2012 MEAP Fall 2013 M-STEP Spring 2015

FAQs

What do Student Growth Percentiles (SGPs) measure and how are they conceptually defined?

SGPs quantify how an individual student grows over time by describing growth relative to other students with equivalent past test scores. Percentile rank is assigned based on a student’s position within the ordered group for the current-year test.

How are SGPs calculated from prior and current test scores?

Students are grouped with others statewide who had equivalent scores on the previous test, then ordered by their current-year test scores. Each student receives a percentile rank based on their order in the group.

What are the stated promises and challenges of SGP use in evaluation?

The approach helps quantify growth and can yield similar results across some methods when multiple years of test scores are available. However, it makes no attempt to consider other factors affecting achievement growth and can be challenging in how “academic peers” are defined and compared.

Oct 3, 2026

Upload by Well Done