Selection Bias in Innovation Studies: A Simple Test
نویسندگان
چکیده
The study of the innovative output of firms often relies on a count of patents filed at one single office of reference such as the European Patent Office (EPO). Yet, not all firms file their patents at the EPO, raising the specter of a selection bias. Using a novel dataset of the whole population of patents by Belgian firms, we show that the singleoffice count results in a selection bias that affects econometric estimates of innovation production functions. We propose a methodology to evaluate whether estimates that rely on the singleoffice count are affected by a selection bias. JEL Codes: O31, C18, C52, C81
منابع مشابه
تصادفی سازی در مطالعات کارآزمایی بالینی: از تئوری تا عمل
Background & Aim: Randomized clinical trials (RCTs) are studies which are able to provide the most valid evidence to compare various interventions in health research. Biases can affect the quality of research and ultimately make the results of a study invalid. One of the most important biases is selection bias. The best way to reduce selection bias is the use of random allocation. The aim of th...
متن کاملVector Autoregressive Model Selection: Gross Domestic Product and Europe Oil Prices Data Modelling
We consider the problem of model selection in vector autoregressive model with Normal innovation. Tests such as Vuong's and Cox's tests are provided for order and model selection, i.e. for selecting the order and a suitable subset of regressors, in vector autoregressive model. We propose a test as a modified log-likelihood ratio test for selecting subsets of regressors. The Europe oil prices, ...
متن کاملOn the Link Between R&D, Innovation and Productivity: Panel Evidence for Dutch and French Manufacturing
This paper studies the relationship between R&D, innovation and productivity in Dutch and French manufacturing using an unbalanced panel of enterprise data from three waves of the Community Innovation Survey pertaining to the period 1994-2004. We estimate by FIML a simultaneous-equation model of R&D, innovation and productivity accounting for individual effects and correcting for sample selecti...
متن کاملA simple criterion for controlling selection bias
Controlling selection bias, a statistical error caused by preferential sampling of data, is a fundamental problem in machine learning and statistical inference. This paper presents a simple criterion for controlling selection bias in the odds ratio, a widely used measure for association between variables, that connects the nature of selection bias with the graph modeling the selection mechanism...
متن کاملOvercoming Innovation Resistance beyond Status Quo Bias - A Decision Support System Approach (Research-in-Progress)
When innovative products and services are launched to the market, many consumers initially resist adopting them, even if the innovation is likely to enhance their life quality. Explanations for this behavior can also be found in specific personality traits and in general pitfalls of human decisionmaking. We believe that decision support systems (DSS) can help alleviate such innovation resistanc...
متن کامل