Researchers sometimes use ‘dummy variables’ to capture qualitative information that is difficult to measure when conducting academic research. The use of dummy variables enable researchers to investigate issues that could have been overlooked due to lack of data availability. However, it seems dummy variables contain limited information when compared to continuous variables. Unlike dummy variables, continuous variables may establish the effect of X variable on Y variable, at least in a loose sense. Is a dummy variable sufficient to claim contribution in academic research?
Yes.Dummy variable can form part of contributions to knowledge.It depends on the phenomenon you are assessing.For instance the role of institutional quality as driver of financing choice of firms.The quality of say legal institutions determines contract enforcement and transaction cost for firms.This may be represented by dummy say 1 where there is high contract enforcement or 0 otherwise.
There is nothing "loose" about the effect of a dummy variable -- it as a well defined algebraic meaning in terms of its effect on the intercept in the regression equation. So, there is no issue with regard to the potential importance of dummy variables in a statistical sense.
In a substantive sense, a dummy variable might be a poor substitute for something that could have been measured with a continuous variable, but many things are inherently categorical, such as sex.
- Badges
- Science topic
- Similar topics
- Microbiology
- Scientific Research