Analysis of factors associated with patent applications (2000 - 2010): indicators of patentability success

Pablo Alvarez; Arturo Arguello; Julia Sanchez-Cantalejo

Volume : VII, Issue : IV, April - 2018

Analysis of factors associated with patent applications (2000 - 2010): indicators of patentability success

Pablo Alvarez, Arturo Arguello, Julia Sanchez Cantalejo

Abstract :

We developed a multivariate linear regression model to analyze patent applications in European countries in order to identify patentability success indicators.

Information was gathered from the Eurostat and World Intellectual Property Indicators databases (period 2000-2010). Three regression models were constructed using as response variables: the total number of applications for PCT patents in the national phase (M1), patent applications to EPO per year at national level (M2), and patent applications to EPO per year and economic activity (M3), and considering 10 variables related to R&D funding and research personnel as predictor variables

Multivariate linear regression models were estimated using the Bayesian Information Criterion (BIC).The most influential predictive variables were: the total number of R&D research personnel by sector, qualification, and sex (determination coefficient of 44.9 %) in the M1 model; the total R&D research personnel by sector, qualification, and sex as well as the business R&D expenditure by economic activity in both the M2 model (coefficient of 79.2%) and M3(coefficient of 78.8%). In conclusion, the mathematical models show that the predictors with greatest effect on patentability are qualified R&D personnel and business R&D funds, and they reveal the distribution of European countries as a function of these variables.