In thinking about my own experience with writing literature reviews and re-reading my earlier post Those old papers are worth something…, I’ve realized I have more to say about this topic. I forgot an important detail about how to actually do the writing of literature reviews. It’s probably worth a deep dive so let me give a little history on this and then walk you thought how to use the modern tools.
I completed my graduate degrees between 2003 to 2008. I’m confident to say that I was probably the last group of graduate students who needed to physically go to the library to find the older books and scan in articles. After that time internet search engines improved, and libraries offered services to scan articles. In recent history, I’ve only gone to the library to retrieve some books that have too many pages to offer a digital copy or old enough that the digital copies that exist online are poor quality such that I need the figures in real life.
Ok, so here we are in the present with a need to do a good literature review and great tools to accomplish it but relatively underdeveloped skills in using these tools. Maybe this is not a problem for you, so please offer your tips but here are my own. First, use internet search tools or library database tools to find your favorite reference using a keyword search. I teach in a metallurgical engineering field so I’m going to pick out something interesting to me. In fact, I’m going to give you a pickup line to use, “Are you made of copper and tellurium? Because you are CuTe.” Corny I know, but what temperature does this alloy melt at? I want to know how hot CuTe really is.
Step 1 go ahead and do a keyword search for “CuTe phase diagrams.” Realize the first 10 articles are generic resources on phase diagrams because those are way more popular than copper tellurium alloys. If you scroll down far enough or revise your search for “copper tellurium phase diagrams” you will find articles related to what you were looking for.
Step 2 after the keyword search, we have everything we need right? Why would anyone really need to know this info besides the pickup line requirement? Cu2Te has semiconductor properties in a liquid state so maybe that’s cool for you. This is about literature reviews so let’s work through this. Maybe you are interested in liquid semiconductors, so we could do one or two more searches using these new keywords. However, in our case I’m going to recommend a different plan. Who wrote the article? The search engines for searching papers are much better at finding authors than looking for every combination of keywords. Typically, a graduate student or advisor develops a theme on certain topics. Therefore, you can see the evolution of the research area if you follow a certain research group or set of people. Also, you can track the topic in time both backwards and forward from the article that you start with. If the individual only wrote one paper, then you are complete. If there are more, you can collect all the articles that author has written. Then you can also look at co-authors as well as the references for that work.
Step 3 I need to emphasize this again, using keyword searches on internet search engines will never reveal all the reference articles. You must find the papers the same way that they are written, by authors! Eventually authors will write other papers or wander off into other areas. So, what we end up with are small groupings of papers that are relevant to your research topic. I also generally advise to add papers until the reference set is “complete” for me complete is that there are no other authors or groups that have anything else to say on the topic. If you have too many papers then you don’t have enough specifics and if you don’t have enough, then you need to widen the search. I use a rule of thumb that about 100 papers is usually enough to define the edges of my research topic.
Step 4 don’t stop looking for papers. Eventually using keywords, author searching, and references gathering you will find a complete set of papers. Then you may have reviewers that reveal complete new sets of literature related to your topic.
Dr Monroe Thoughts 