Test Score Gain Decay
Short-term test score gains from educational interventions often diminish rapidly in subsequent academic periods, complicating the evaluation of long-term program effectiveness. Evaluating program efficacy without accounting for natural decay masks structural behavioral adjustments by households and schools.
Picture this
Learning gains achieved in one year naturally fade over time as students move to higher grades and encounter new material unless reinforced by continuous, net-additive instructional inputs. It works like momentum on a bicycle: pushing once speeds up the bike temporarily, but without continuous effort or additional force, friction gradually slows it down.
What the evidence says
Test score gains decayed across years, with Year 2 treatment effects falling to 0.02 SD (0.005 SD in math, 0.035 SD in language) from Year 1 gains of 0.085 SD, rendering cumulative two-year effects statistically indistinguishable from zero (p = 0.611).
- Who
- 200 primary schools (42,710 pooled student-year test observations in grades 2-5) across rural Andhra Pradesh, India.
- How
- Multi-year panel tracking of normalized student test scores (z-scores) comparing baseline, Year 1, and Year 2 performance using OLS and non-parametric estimation.
What to do
Account for empirical decay rates when modeling multi-year returns on educational investments to avoid overestimating long-term impact from short-term baseline gains.
From the source
"The size of gains after two years (with point estimates below the point estimates after Y1) suggests that the second year of block grants did not add much to learning outcomes, while decay of earlier gains may explain why average gains (in terms of point estimates) after Y2 are smaller than achieved after Y1."
School Inputs, Household Substitution, and Test Scores