Traditionally, most agriculturists conducting agricultural technology adoption research use a very “loose definition” for a technology “adopter”. Generally, they define a technology adopter as anyone who reports that he is practicing a technology of the researchers’ interest at any level or scale. There are many problems that arise from defining a technology adopter this loosely. In this article, however, I will only highlight one.
As you may already be aware from this post or from elsewhere, agricultural technology adoption is a process with definite stages (see Everett Rogers). Experience has shown that one of the most important stages of this process is experimentation. In this stage, farmers try out the technology they have been exposed to on a very small-scale so as to make an independent assessment of whether or not the technology will meet their specific needs before they can make a decision to keep using it (in other words adopt it). Being rational beings, if they find that the technology does not meet a criterion which they use to score a technology as meeting their needs, the farmers simply do not adopt the technology (or some very experimental farmers try to modify it so that it suits their specific needs). Therefore, if a researcher using the loose definition for adoption interviews a farmer during the experimentation phase, they are more likely to record the farmer as an “adopter” based on the fact that the farmer is “practicing” the technology albeit at an experimental level.
However, what the researcher actually gets is not a “full picture” of the state of adoption of the technology in question [as he has an inflated number of adopters] such that he may overestimate determinants of adoption of the technology as well as its impacts on farm households because he is using a wrong sample. Consequently, any report or policy recommendation that may be made based on data collected using the loose definition may actually be quite misleading and harmful to the farmers as well as costly to government, donors and development organizations. Unfortunately, no amount of statistical/econometric techniques can correct this error.
From the foregoing then, I suggest that when designing agricultural technology adoption research, it is important to keep in mind that adoption is a process such that when you are conducting your survey you are likely to find farmers at different stages of adoption. Therefore, it is advisable for researchers to clearly define who they consider an adopter based on a concise criterion like scale of adoption.
P/S: I’m writing a series of essays on agricultural technology adoption so i thought i should share with you some of my opinions on this topic. This blog post is just one of several that i will publish on this topic. I hope you will find to give me your thoughts on this topical issue.in the next post on this series on agricultural technology adoption, I will focus on how researchers and policy analysts lose a lot of valuable information that may help improve agriculture by working on the assumption that technology adoption is mostly binary; adopt or not.