Multivariate analysis to research innovation complementarities
It is widely recognized that orthodox economics is obsessed with econometrics tools. However, econometrics techniques have a limited capacity to deal with qualitative variables coming from surveys. This paper presents a defence of the use of statistical methods, in particular multivariate analysis,...
| Autores: | , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2017 |
| País: | Argentina |
| Institución: | Consejo Nacional de Investigaciones Científicas y Técnicas |
| Repositorio: | CONICET Digital (CONICET) |
| Idioma: | inglés |
| OAI Identifier: | oai:ri.conicet.gov.ar:11336/58286 |
| Acceso en línea: | http://hdl.handle.net/11336/58286 |
| Access Level: | acceso abierto |
| Palabra clave: | Innovation Complementarities Multivariate Analysis Plurality Software Sector Supermodularity https://purl.org/becyt/ford/5.2 https://purl.org/becyt/ford/5 |
| Sumario: | It is widely recognized that orthodox economics is obsessed with econometrics tools. However, econometrics techniques have a limited capacity to deal with qualitative variables coming from surveys. This paper presents a defence of the use of statistical methods, in particular multivariate analysis, which is the overall objective of the paper. Multivariate analysis is a set of methods that can be used when the problem that arises implies multiple dependent or interdependent variables of a qualitative nature. We considered an issue in the literature to probe multivariate analysis in a particular topic, namely: the question of innovation complementarities. We analyzed the presence of complementarities between internal and external innovation activities in 257 software firms from Argentina during the period 2008–2010, comparing the consideration of the problem of complementarities with the more modern complementarity econometrical tests, super and sub modularity tests arising from diverse firm-innovation function estimations (OProbit, Tobit and Probit), with the engagement of the same issue with multiple factor analysis and cluster techniques. The results show not only that the same results obtained by the econometrical tools can be reached by multivariate analysis techniques, but also that multiple factor analysis and cluster techniques allow for better exploitation of the richness of qualitative data. |
|---|