Innovation in developing country firms is thriving, but this is being overlooked because the methods used to study private sector ingenuity are inappropriate, SciDev.Net reported last month. The story was based on a pilot study carried out in Ghana by Xiaolan Fu of the University of Oxford, United Kingdom.
I called one of Fu's doctoral students, Claudia Contreras Rojas, to find out more about the research methods she is using to better understand how companies innovate in resource-poor countries.
Contreras is originally from Chile and has been investigating how firms there collaborate with universities to create innovations.
She tells me that one big problem she faces is that developing world companies often don't understand what counts as an innovation.
Academics have been trying to rectify this problem with definitions, she says. For example, UNESCO (the UN Educational, Scientific and Cultural Organization) has released guidelines, known as the Bogota manual, on how to identify innovation in Latin America. 
These build on an earlier guide for identifying innovation, published by the OECD (the Organisation for Economic Co-operation and Development) and known as the Oslo manual, which researchers eventually realised couldn't be applied in poor settings.  Contreras explains that this is because the processes of technological change are radically different in developed and developing nations.
"For example, as part of my research in Chile, I was talking with people that work in the mining industry," she says. "Often they will get their machinery from abroad. But they have to adapt everything to the local conditions: to the capacity of the local people, to the type of local rock and to the density of copper found there."
Clearly, this is innovative, says Contreras, yet local mining engineers don't see it as such. For this reason, asking them questions via a survey - without properly understanding their wider perspective - may not be enough to extract a true picture of the ingenuity involved in the work.
Yet even if survey questions are phrased more carefully, these may not be enough to get an accurate picture of their innovation, says Contreras.
"In my experience, it's better to use mixed methods," she says. "Surveys give you a general image of the situation. But you cannot stop there. It's very important that you then go to specific agents - perhaps public policymakers or companies' employees - and then you do case studies, you chat to them."
Those case studies can reveal more details and spur further surveys or research. "The process is kind of iterative," says Contreras. "But in order to create a complete story, you have to talk to people."
Joshua Howgego is SciDev.Net's deputy news and opinions editor.
 Hernán Jaramillo and others Bogota manual (UNESCO, March 2001)
 Oslo manual (OECD, 2005)