Hyperlinking in ATLAS.ti: Expanding the Boundaries of Data Exploration

July 6, 2015

Author: Ricardo B. Contreras

Introduction

Data segmentation in ATLAS.ti is the process of selecting fragments from the sources of information (i.e., documents) and transforming them into quotations. A unique characteristic of these quotations is that they are independent objects. This means that they can be created without having to immediately link them to codes. This opens the doors to freely explore the data in a rather open-ended way, identifying passages that may call the researcher’s attention, and storing those passages so that they can be revisited later.

Besides being connected to codes, quotations can be renamed, commented, linked to memos, and linked to each other through hyperlinks. This flexibility means that much can be done in terms of description and analysis without having to resort exclusively to coding. Following Saldaña (2009:3), coding is a process whereby summative labels are assigned to the data, and by doing that, data are reduced, complexity is somehow simplified. But what if we do not want to rush too soon into data reduction, what if we prefer to first describe and explore openly in an effort to develop preliminary understandings of the data in all of their complexity? The independent nature of quotations allows for that. In this short article I will describe the role of hyperlinking in data analysis with ATLAS.ti.

Hyperlinking: Building Webs of Meanings Between Units of the Discourse

Hyperlinking is the process of connecting quotations to quotations through meanings in an effort to describe how study participants build their arguments (see Friese 2014:256). Through hyperlinking, it is possible to identify strong examples of contradictions in the discourse, show how one argument may follow or complement another, demonstrate how a picture may illustrate what its oral or written description cannot fully express, describe how the movement represented in a video frame may help to support what was found in the interview text, among other things. Quotations can be hyperlinked within a document or across documents, allowing the researcher to build webs of relationships among units of the discourse that can span across as many sources of information as desired. Silver and Lewins (2014:53) stated that through hyperlinking in ATLAS.ti “it is possible to analyse qualitative data without using coding tools at all.”

ATLAS.ti comes with a set of standard hyperlink relations but the researcher can create new ones as needed, preferably guided by a theoretical framework. The standard relations are the following:

  • Continued by
  • Contradicts
  • Criticizes
  • Discusses
  • Expands
  • Explains
  • Justifies
  • Supports

These relations can be used to describe and illustrate how units of the discourse (i.e., quotations) relate to each other as arguments are being constructed. Through hyperlinking it is possible to describe how a single individual builds an argument and how arguments are built across study participants.   An example of a hyperlink across documents is given by the following example, in which contradicting opinions regarding the role of social relations in shaping one’s own happiness are given in two sources of information:

Table 1. An example of a hyperlink using the "Contradicts" relation.

Table 1. An example of a hyperlink using the “Contradicts” relation.

Another example of hyperlinking across documents is given by the following two quotations, both of which support the same idea: a person’s happiness cannot be shaped by external factors.

Table 2. An example of a hyperlink using the "Supports" relation.

Table 2. An example of a hyperlink using the “Supports” relation.

Of the hyperlink relations that come standard with ATLAS.ti 7, the “Continued by” one is particularly useful as a complement to coding. Let’s say that close to each other, but not in the same paragraph, there are two mentions to the same idea in the same document. Given this, the researcher has three options if a decision to code has been made: to code the entire segment with same code, to create two quotations and code them with the same code, or to create two quotations, but code only one of them. Following, the one quotation would be hyperlinked to the other using the “Continued by” relation. The benefit of this approach is that the central idea is captured and the quantitative significance of the code is not artificially inflated.   Using the same hermeneutic unit, see below an example of a “Continued by” hyperlink.

In the text, this hyperlink looks as follows (the arrowed line and the text highlight were added while editing the screenshot):

Figure 1. Screenshot showing the hyperlinked quotations on the margin.

Figure 1. Screenshot showing the hyperlinked quotations on the margin.

In figure 1, the second quotation consists only of the name of the person, while the first one includes a short biography of that person. Thus only the quotation with the biography was coded while the other was not, and the second one was linked to the first one through the “Continued by” relation.

This hyperlink can be shown in a table format as follows:

Table 3. Example of hyperlink using the "Continued by" relation.

Table 3. Example of hyperlink using the “Continued by” relation.

The written output, after editing it a little bit, may look like this:

Figure 2. Report on a code showing hyperlinked quotations.

Figure 2. Report on a code showing hyperlinked quotations.

Reporting Hyperlinks Through Summary Tables and Network Views

There is the option of reporting hyperlinks using summary tables, network views or written reports. Here I will say a few words about summary tables and network views. Although ATLAS.ti will not create the summary tables for you, it is easy to make them using your favorite word processor. It is a good idea to summarize the key hyperlinks around specific topics in this way. For example, if we explore the topic “Arguments that participants are making in relation to happiness” we find that there are arguments that support each other and some that contradict each other. Thus, the summary tables would look as follows:

Arguments About Happiness: Mutually Supportive Arguments Across Sources of Information

Table4

Table 4. Summary table showing mutually supportive arguments.

Arguments About Happiness: Contradictory Arguments Across Sources of Information

Table 5. Summary table showing contradictory arguments.

Table 5. Summary table showing contradictory arguments.

It might also be useful to include network views of hyperlinks in your study report, particularly when there are several quotations hyperlinked to each other, or when there is one quotation that is rather central in relation to other quotations. For instance, see the network view below:

Figure 3. Network view showing web of hyperlinks.

Figure 3. Network view showing web of hyperlinks.

Conclusion

Hyperlinking in ATLAS.ti expands the boundaries of qualitative data analysis by providing an opportunity to describe, analyze, and interpret the data without having to resort exclusively to coding and to data reduction techniques that may too soon limit the potential for discovery. Hyperlinking allows for the exploration of relationships between units of the discourse rather than between conceptual constructs. From that point of view, it helps to keep the researcher close to the data, close to the voice of the participant. Importantly, the researcher can create her own set of hyperlink relations and in that way approach the analysis of the data from the theoretical framework of her choice.

References Cited

Friese, S. (2014). Qualitative Data Analysis with ATLAS.ti (2nd ed.). London: SAGE Publications Ltd.

Silver, C., & Lewins, A. (2014). Using Software in Qualitative Research: A Step-by-Step Guide (2nd ed.). London: SAGE Publications Ltd.

Saldaña, J. (2009). The Coding Manual for Qualitative Researchers. 2009: SAGE Publications Ltd.

About the Author

Ricardo B. Contreras directs the Training & Partnership Development division of ATLAS.ti Scientific Software Development GmbH. His undergraduate training is in sociocultural anthropology (Universidad de Chile) and his graduate degrees are in applied anthropology (University of South Florida). Ricardo is also a sociocultural research consultant.

Share this Article