Litrature Review

A literature review begins as a collection of material and sources (usually peer-reviewed journal articles) that are related to your chosen topic/argument.

When you begin researching your topic, you will be acquiring different journal articles that are related to your discipline, thesis, and topic/argument. For example, if you are a student of computer science and your thesis is focused on machine learning (training computers to interpret information), and your topic/argument is focused on expanding the collection and usage of data within machine learning to achieve better results, then you will be searching for journal articles involving keywords such as data collection, machine learning, data in machine learning, training machine learning models to interpret data, etc. Download or save all of these journal articles in one place for easy access.

Once you have a significant amount of journal articles, you'll need to read through them and understand their arguments and results.

You'll want to read through your sources and only really keep the ones that are relevant and useful for your paper. You might have a few whose titles sound like they'll work with your paper, but when you read them, you realize that they actually do not. It's okay to delete them or omit them. You don't have to use every source you read. For the ones that are related to your topic, try and group them a little bit based on their theme or their content. For example, if five of your sources are all concerned with the method used to collect data for machine learning, group them together; if six others are focused on expanding the amount of data (rather than the method used to collect data), group those together. Find commonalities between articles and make note of them, as you can group these into useful paragraphs for readers.

And perhaps the most important thing about a literature review:

it is not a summary of journal articles related to your topic. This is perhaps the number one  problem with most graduate students' literature reviews: they simply spend paragraph after paragraph devoted to summarizing journal articles and that's it. The problem here is that the reader doesn't want to simply know what those articles were about. No, the reader wants to know (i) what arguments did each article make, (ii) what results did each article discover, and (iii) how is that article related to your topic/argument. You essentially need to be answering all three of these things. Number three (iii) is usually answered at the beginning of each paragraph because you want to tell the reader why each article in that paragraph is grouped together (i.e. - the reason you are even mentioning them in your paper). Then you want to introduce the article, provide the reason the article exists (the argument/topic), and then give the results of the article.

There is a shortcut.

I spoke with several professors during one semester where I was particularly hurting for time (because I took too many courses), and they both offered me some advice: there is a shortcut, if you need it, to completing your literature review, and that shortcut involves quickly processing the content of a journal article. Basically, all you need to do is read: (i) the Abstract (which should always tell you the topic of the article, as well as a brief mention of the final results); (ii) the Introduction (which will provide you with the thesis statement and some background information); and (iii) the Conclusion (which will provide you greater detail of the final results). Using this shortcut, you can get through journal articles quicker and be able to summarize them much more efficiently.

[1]