After reading Susanne Friese's book, I knew what functions of the ATLAS.ti program I most wanted to put to work: memos, queries, and network views. I have tried unsuccessfully in the past to use these features, but they require a certain amount of finesse and a certain amount of insight into their underlying logic that I just did not possess. My usual strategy for learning a new technology (pushing a lot of buttons, clicking a lot of links, and generally just fooling around with it until it works) just wasn't cutting it.
So, I turned to Friese, whose book outlines the strategic use of memos for running data queries and integrating findings. Memos are instrumental to the entire analytical process but play an especially important role during writing and production of the final report (read: DISSERTATION).
Friese encourages: "In order to see the benefits of it, you have to try it." She says some people may call it "magic," but it is really just "the added value of approaching analysis in a systematic way utilizing the options available" (p. 142).
Stop expecting "magic." Start applying a "systematic" approach. That is what I needed to do.
Memos
So, I started small, with one primary document (Watt's 2007 article on researcher reflexivity), one memo, and one writing task. I wanted to use the memo tool to develop some discursive text -- a blog post -- based on Watt's article. Before I began reading and coding the article in ATLAS, I opened the Memo Manager, created a memo, gave it a "proper" title, and categorized it by type.
Meanwhile, I also created, titled, and categorized a free memo to record my process of using memos in ATLAS. Yes. A memo about memo'ing -- a "meta memo," which, in turn, became this blog post.
According to Friese, memos can be linked to quotations, to codes, and to other memos. I tried this and linked the Watt memo to the PDoc itself, to specific quotations within the PDoc, and to significant codes. Now what?
Queries & Network Views
Friese says memos serve an analytic function (a "container of ideas") and a technical function. After coding the data (in my case, the Watt literature), the user may probe more deeply by running queries. Queries serve a data retrieval function by probing connections between PDocs, quotes, codes, and memos in depth. The memo serves as a place to record descriptions, interpretations, and ideas based on these probes. Friese calls it "second-level conceptual analysis" (p. 7).
However, for the task at hand, I simply had one PDoc, one memo, a few codes, and a handful of quotes. Upon further consideration, I decided to save the query function for another day. For purposes of developing my blog post, I turned my attention to the network view and memo output functions. I was especially intrigued by the potential of network views for displaying data, creating concept maps, and developing writing heuristics.
Projects created within ATLAS.ti are referred to as hermeneutic units (HUs), based on the science of text interpretation (I know because I finally took the time to look it up.). The HU consists of links that the user creates between all sorts of object nodes: PDocs, quotes, codes, and memos. Thus, the HU is really a network, and network views are detailed perspectives on different aspects of the network.
The thing is, there truly is nothing "magical" about network views, which is why I never had any success with them in the past! I possessed a fundamental misconception of this tool: I thought when I clicked the display network option, the network would magically populate the screen like a flow chart, with all the nodes logically mapped out and connected for me.
On the contrary, when the user displays the network for an item and all its nodes, the user must manipulate the objects, linking and labeling them in a manner that makes sense to the user. This is why Friese has an entire chapter (Chapter 7) devoted to the creation and manipulation of network views.
My Process
I took some time to explore the quotes and codes associated with the Watt PDoc. How could I use the network view to glean new insights from Watt (2007) in relation to my current project, the dissertation? This exploration was messy and recursive. I spent a lot of time clicking on the quote nodes, which allowed me to read them in full. I found that it was more fruitful to right-click on the quotations and Display in Context. This resulted in more reading and, in some cases, more coding. Finally, under the Display menu, I opted for Quotation Verbosity, Full Text, so I could work with the quotations in full as I arranged them in the network view.
Arranging the codes and experimenting with the Code-to-Code Relations Editor helped build my understanding of exactly what I wanted to say in my Watt reflection. Naming the semantic relationships between the code nodes led me to see how most of the codes had one thing in common: they each connected somehow to the "trustworthiness" code.
I discovered I could double-click on the memo object to display its text in full, and I could double-click on the memo text to edit it. So, as I arranged the network view, I revised and wrote whole new portions of the memo. To illustrate points in the memo, I copied and pasted selected quotes from the context of the article. Toggling in this manner, from network view to memo and back again, became tedious, so I broke my paperless rule and exported the (mostly) finalized network view as a graphic file to my desktop and printed it. It sat on my physical desktop as I finalized the blog post. I would not have done this if I had a larger computer screen or dual monitors.
As can be seen in the "before" and "after" screenshots, I only scratched the surface of what is possible with network views. Using the network view did not make my writing process more efficient -- it took as much, or possibly more time to compose the blog post -- but it did help me conceptualize something new about an article I had already read multiple times over the last three years.
 |
Network View BEFORE |
 |
Network View AFTER |
A final word about Memo Output
So, in the end, I did not utilize the memo output function. With the memo output function, the user selects the memo, right-clicks, selects Output and Selected Memo with Quotations, and chooses Editor as the destination. An editable document is generated that "includes everything you need to write the results chapter of your research report or paper" (p. 146). However, I can see that if I was working with multiple memos and needed to string them together into one coherent piece of text, such as a chapter of findings, I might want to use this option to pull everything into a more robust word processing program.
I am registered for the special topics webinar on ATLAS.ti Networks, scheduled for April 9 at 11 a.m.
ReplyDeleteAwesome - I love it!
ReplyDelete