Analyzing and browsing your literature with ATLAS.ti

April 7, 2020

This blog article was written by one of our Certified ATLAS.ti Professional Trainers, Jorge Alberto Mahecha Rodriguez (Boston College, USA)

There is no question about how useful ATLAS.ti is as a tool for supporting qualitative research. However, every research project, even those with research questions that need to be addressed by quantitative research methods, needs a thorough review of the relevant literature. My experience with ATLAS.ti is framed in this context. My research project has research questions that require quantitative methods: my doctoral dissertation is about comparing Randomized Controlled Trials (RCTs) to Quasi-Experimental Designs (QEDs) based on propensity scores. However, as quantitative as my research questions are, I need to do what every other scientific researcher has to do: I have to thoroughly review the relevant literature and get a sense of what has been said and done in my field of research.

Studying highly specialized topics like that of my dissertation, or any other doctoral dissertation topic for that matter, requires a deep dive into literature that, besides being arcane and intimidating, is in most cases quite abundant. For example, the ATLAS.ti project I have built to support my research includes so far 59 papers out of some 140 that I have collected during the last year. As time passes, this number is likely to only grow. And while it is easy to remember a handful of seminal papers key to your research, at some point the amount of papers and details relevant to any given project grows beyond what is practical or feasible to recall. Remembering who said what, in what context, and more importantly, what has been said about a particular topic, is perhaps one of the most important capabilities needed for rigorous research. In this blog article, I will describe how I have been using several ATLAS.ti tools to deeply explore the literature relevant to my research.

Although not by any means sophisticated, the one thing I like to do the most in ATLAS.ti is reading papers. And I do not mean skimming through a paper, which you can always do using the PDF reader of your choice. When I want to commit to serious, research-level reading, I use ATLAS.ti. The reason why I read in ATLAS.ti is that I can go way beyond simply highlighting a quotation that could be interesting for potentially many reasons. Of course, highlighted quotations are important, but why? It’s key to know exactly why they are important. In ATLAS.ti, I can take any quotation, code it, and comment on what it is that makes this quotation relevant to the problem at hand. ATLAS.ti quotations have an additional interesting feature that makes copying and pasting from ATLAS.ti quite convenient: they are stored as plain text exactly as you see them in the PDF file. There are no weird paragraph marks between text lines, as it is common when copying from PDF readers, and no accidental selection of both columns in a paper when you mean to select just one. The process of selecting quotations and linking codes to quotations is greatly facilitated by the ATLAS.ti for Mac keyboard shortcuts: command + H (for quotations, ⌘+H) and command + J (for linking a code to that quotation, ⌘+J).

In addition to the advantages of having quotations linked to codes, there’s the fact that as you read and further generate quotations and codes, you can relate them to previous codes and quotations you have identified. In this process, code networks are built that help you analyze your research. For example, here is a network showing everything I have identified about something key to my research project: what has been done in regard to simulating outcomes of experiments or quasi-experiments:

Simple strategies, like following a consistent approach when naming codes, allow for quickly examining everything you have managed to identify on a relevant topic. For example, because of their names, which all share the same stem, my codes are adjacent and easily visualized in the code manager:

This organization allows me to easily review everything related to simulations in my project using the code manager. The code manager in ATLAS.ti is not only a great tool for having an organized repository of whatever is important to your research, but also a great system for exploring all of your quotations and documents. By clicking on each code, all of the quotations that are linked to this code are available in the lower panel, where I can browse through them all, regardless of the documents where they come from. If I need to explore the context of the quotations in further detail, all I have to do is double-click the quotations, and it opens the document, at the exact place where the code and quotations are. This feature makes searching for specific pieces of information in a large number of documents a very easy process.

Another simple but powerful feature of ATLAS.ti is the ability to conduct simple text searches across a multitude of documents using the “Find in Project feature”. In my field of research, for example, it is key to understand the process and characteristics by which participants are included in quasi-experimental studies, which is known as the “selection process” or “the selection model”. If I just type in “selection process” in the Find in Project search box in the left-side panel of ATLAS.ti, I get a list, organized by document, of every instance where the expression is found. It is hard not to overstate how useful this feature is, not only for browsing, as seen in the figure below but also for coding:

Auto coding allows me to code either the exact words, the sentence, or the paragraph that contains the search terms I am interested in. I can do this for one document or all documents in my project. For my research, I am rarely interested in words or sentences; what I need to know is the full context where selection processes are mentioned. To do this, all that is required is to open the auto coding dialogue, and in the “Extending” dialog box, select the level I am interested in (paragraph) and then the documents where I want to find this. I also added the “ignore case” option, as case is not relevant to my search:


I could do the coding by revising each specific instance where this phrase appears, or I could click “code all”, which is what I did, because it is what serves my purpose of finding every instance of that particular expression. In my project, this resulted in 198 quotations, which I can now easily browse through in the code manager and decide which ones I’d like to keep. In this way, this searchable list of quotations is an invaluable resource for me to learn about a particular aspect of key importance to my research.

In sum, I’d like to restate that no matter your field of research, rigorously reviewing the relevant literature requires a tool that allows you to conduct very specific searches in an efficient and easy manner. ATLAS.ti offers unparalleled flexibility to do this, because of the manual and automatic options for coding that are available and the use of the code manager as a sort of research browser for your project.






Share this Article