Does Writing Three Lines a Day Really Change Anything?

Robert Emmons and Michael McCullough's 2003 gratitude experiments launched a global research industry, but twenty years of replication have quietly deflated the original promise. The 2025 meta-analysis pooling 145 papers and 24,804 participants landed at Hedges g = 0.19: statistically real, practically modest. The shrinkage isn't failure; it's what happens when comparison groups improve. A 2026 mega-study across 34 countries confirmed the pattern: "list five things" was the weakest arm, and effect sizes varied by nation in ways no cultural variable could explain.

You'll learn:

- Why active comparison groups collapsed the effect to near zero

- How the Romania expectation trial isolated placebo from practice

- Which format adjustments outperform uniform daily listing