How Big Data Applications Drive Disruptive Innovation: Evidence from China’s Manufacturing Firms

Accordingly, this study uses Chinese A-share listed manufacturing firms from 2014 to 2023 as the sample. Data on big data applications are collected from firms’ annual reports available on the official websites of the Shenzhen and Shanghai stock exchanges. To ensure the validity of the disruptive innovation measure, the disruptive innovation indicator is constructed using patent data from the China National Intellectual Property Administration (CNIPA) for 2000–2023, with the detailed procedure reported in Section 3.2. Other data are mainly obtained from the CSMAR and Wind databases.The data are processed as follows. First, firms designated as Special Treatment (ST, Firms that have exhibited financial distress for two consecutive years), particularly Special Treatment (*ST, Firms that have reported consecutive losses for three years or face the risk of trading suspension), or Particular Transfer (PT) are excluded. Second, firms with substantial missing values in key variables are removed. Third, revenue growth rate, Tobin’s Q, asset turnover, and leverage are winsorized at the 1st and 99th percentiles to mitigate the influence of outliers. Fourth, to reduce potential multicollinearity, the core explanatory variable (DE) and the moderating variable (GI) are standardized before constructing the interaction term. After these procedures, the final sample consists of 19,035 firm-year observations.