Is the Global Consciousness Project detecting real effects?

The GCP has accumulated over two decades of data, and its proponents report a cumulative z-score that is astronomically unlikely by chance. However, the project's methodology has been criticized for its event selection process: which world events are designated as "test events" and how the analysis window is defined can substantially affect results. The Bosch (2006) meta-analysis of the broader RNG literature casts doubt on the effect class as a whole. Independent analyses (Bancel 2017) suggest the effect is fragile under different analytical choices. Holmberg (2020) adds an independent econometric test: by correlating GCP Max[Z] with daily stock market returns across 11 global indexes, he sidesteps event-selection entirely — but introduces post-hoc model selection concerns and the unresolved common-cause problem (both series may react to global news shocks). The GCP remains a unique dataset in parapsychology — large, continuous, publicly available — but interpretive consensus remains out of reach. The fundamental debate is whether the cumulative statistical signal reflects a real phenomenon or the accumulated effect of analytical degrees of freedom over 20+ years of data.