The current research employs two survival analysis methods: Cox regression and life tables. The first determines the effect of inventor, assignee and country for receiving the first citation by patents. Life tables concern the time-lag between the dates of granting and receiving the first citation by patents. Bradford’s method is also established as a technique for categorization of patents, inventors, assignees and countries as a prerequisite for survival analysis. The research materials consist of 2837 patents in the area of ‘purification, separation, or recovery of hydrocarbon components’ which were classified under the classes 585/800 and 585/868 by the United States Patent and Trademark Office (USPTO). The findings showed that Bradford’s method complies with the distribution of citations of patents, first inventors and assignees. It means that Bradford’s distribution is well suited for determination of key patents, inventors and assignees in an area too. Cox regression revealed that only the inventors’ variable decides for receiving the first citation in terms of frequency, degrees of their inventions and citations. Life table data revealed that one half of the first citations were received in the first 10 years. As a conclusion, survival analysis methods provide the possibility for deciding technology lifetime and for predicting the determinants for the flow of knowledge through citation analysis.
