Fig. 2

Regression coefficient t-value of digital economy on technological innovation in cities in 500 regressions. It suggests that only 1% of random shocks in randomly generated experiments have a significant positive effect on urban innovation and that the probability of falsification errors in this study is extremely low.