Nowadays
more information become available and more accessible, big data analysis
becomes very popular. Sometimes people can spot a phenomenon first through data
analysis, then try to create a theory to explain this phenomenon. However,
sometimes people use empirical study and/or experiments to test whether a newly
developed theory works in the real world or not. Therefore, we can say that
there is an undirected link between theory work and empirical work (and
experiment work).
Theory is
very important for us to understand the world. It helps us to understand the
world. Theory and model often simplify the real world problem. More
importantly, theory can help us to open the black box. When we observe a
relationship between two objects, it is hard for us to draw a conclusion about
the causality behind the relationship; then here comes the help from theory.
From the mathematics and logic inside theory, people can understand the
phenomenon better.
However,
it does not mean theory is always correct. Especially in social science, for
one thing, there are multiple theories to explain; all the theories seem legit.
Sometimes theories are different because they are looking at the problem from
different angles; sometimes they are different because they include various
factors. When we face such differences, we do not need to worry too much, we
just need to apply different theories based on the circumstances we face.
However, when the logics behind them are different, then we should be very
careful, because the two theories are possible to contradict each other.
Machine
learning and other methods may be able to help us to solve problems, but theory
will help us to look into the black box and gain better understanding.
No comments:
Post a Comment