Measuring disruptive innovation is a critical and still-developing topic. Although the disruption (D) Index has been widely utilized, it ignores the structural differences between i- and j-type nodes and suffers from inconsistencies, biases related to reference lists, and little comparability across different clusters. To address these possible biases, we propose the improved disruptive Index (ID Index), using a dataset of 114,202 patents from Chinese listed firms to test its validity. The results show that the ID Index (i) provides a more precise measurement of disruptiveness, resolves inconsistencies, reduces biases related to reference lists, and enhances comparability across clusters; (ii) demonstrates better convergent validity, correlating more closely with expert evaluations and more effectively identifying determinants such as knowledge search, recombination, and coordination; (iii) shows better validity in predicting stock market reactions, renewal durations, firms’ short- and long-term performance. Finally, we separate the ID index to independently measure the extent of disrupting and consolidating existing knowledge, and the convergent and predictive validity are demonstrated.
