Blog More signal, less noise: using ANCOVA analysis in RCTs 23 September 2026 Difference-in-Differences is often the default choice for evaluating RCTs with baseline data, but ANCOVA is usually the sharper, more powerful tool. Drawing on a reanalysis of two large business-support trials involving 7,000 UK firms, this post shows how ANCOVA reduces statistical noise, improves precision, and uncovers hidden treatment heterogeneity without forcing an inefficient one-for-one baseline adjustment. Read more about More signal, less noise: using ANCOVA analysis in RCTs
Policy report Building evidence on AI-assisted review in public funding 22 September 2026 Read more about Building evidence on AI-assisted review in public funding
