Literature Review

Literature review is the most boring part of research, but it is also the most important part. Without a properly done literature review, your research question and methodology are built on sand. As a rule of thumb (but not necessary), a journal article with about 3000 words should have about 50 references. During this stage, a common experience for many of us is that we can only remember the last two or three papers as we keep reading new literature. How to make the literature review more effective and efficient? Let me show you a useful technique. 

As an analogy, this technique is like estimating the "moments" (such as mean, standard deviation, skewness and kurtosis) from the complicated "data". Now that you have hundreds of papers (the "data") to be reviewed, and you want to compare the following features (the "moments") of these papers (including but not limited to): the sample used in the paper, the model features, the estimation techniques, the conclusions and the policy implications. A very powerful tool for this purpose is actually MS Excel, given its flexibility of matrix-like organisation. Usually, I use rows as papers to be reviewed, and columns as features of interest. You can add or remove papers and features in a very convenient and intuitive way. The attached table below is an illustration of literature review on Unified Growth Theory using this technique.

After filling in the spreadsheet, you have only done the first half of literature review. You also need to write it up! Note that a literature review is NOT a literature compilation. You should NOT write your literature review like this: "A argues XXX, B claims YYY, and C finds ZZZ." It is not informative at all. Rather, the literature review should discuss the similarities and differences in terms of method, technique, conclusion, implication, etc. In this sense, the spreadsheet done earlier can nicely translate into the write-up stage, because you can group different papers together in terms of different features/criteria. Moreover, remember to critically review the existing literature, so that you can justify your own hypotheses, methods and techniques--that is the ultimate purpose of the literature review.

This technique turns out to be a very useful way of conducting literature review. It does not make literature review less boring, but it does make everything nice and clean, effective and efficient. Please try to use it in your dissertation and thesis!