I’m no masochist so I use CAQDAS software to make coding qualitative data easier. But which to use? I decided to try a few to find out.
Back on my masters course, I was taught how to use ATLAS.ti. I also downloaded trial versions of NVivo 8 and MAXQDA 10, both proprietary CAQDAS packages. In addition, I tried WEFT QDA and Digital Replay System, both open source (and therefore free) software. I installed and tried each for varying amounts of time.
All of the programmes, with the exception of Digital Replay System, come as a self-installing .exe file, making installation easy. Digital Replay System can either be downloaded as source code or a java package. I strongly recommend the pre-packaged version, as the source code has a number of dependencies, and involves setting environmental variables. NVivo is also dependent on Microsoft’s .NET framework, but it does install this for you.
I found that none of the software packages were so easy to use that I didn’t have to delve into the manual or online help at least a few times to get myself going. Having said that, MAXQDA emerged as a clear winner for me in terms of ease of use, and in the intelligence of the interface, which means that you can code quicker. The interface is filled with useless buttons whose function I could not case less about, or only guess at their possible function, but the bulk of the interface is taken up by the document view (in this case, showing one of my transcribed interviews which has been imported as a .pdf file). The grey margin serves to show which sections of text have had which code or codes applied to them (indicated by the large green brace). Finally, on the very left, is the list of all current codes using in this project:
To code a piece of text you simply select it in the document view, then, if the code is one you’ve already created, just drag and drop the code from the code view on to the text! This sounds ridiculous but it’s so simple and no other CAQDAS software uses this system. For example, when using ATLAS.ti, to code a section of text, you have to select text to be coded, right click, select to apply a code, then select your code from a (small) drop down list – a much more painful process.
MAXQDA is the clear winner for me with regards to usability. ATLAS.ti, NVivo, and WEFT QDA had similar interfaces and inefficient coding systems, although neither were as bad as Digital Replay System.
File Sizes (and, therefore, Portability)
I found that file sizes for MAXQDA, ATLAS.ti and WEFT QDA were similar, and less than about 10MB for a typical project. However, I only referenced my documents (rather than imported them into the project), keeping file sizes down. Therefore, to make your project portable, you also need to remember to copy your documents around with you.
I found NVivo, on the other hand, was extremely greedy with hard drive space, using nearly 100MB for a typical project. When you think that each document I used was less than 1MB, this is excessive.
Technical support for MAXQDA 10 has been exceptionally good, but then they’re the only ones I’ve had to email for help. Their license requirements are just clandestine, and don’t work without some tinkering under the hood. You also can’t easily migrate your license from one computer to another (i.e. if you replace your computer) without contacting their help desk. However, they have resolved all my queries to date promptly and politely, which is reassuring should I encounter any more serious problems.
All the proprietary packages come with a hefty price tag, starting at £800 for standard, single-user licenses, up to about £1,100. Student licenses start at about £55 for twelve months (NVivo), although ATLAS.ti and MAXQDA 10 will work indefinitely but cost more (from £68). As a student, it therefore makes little difference to me in terms of cost, but at a difference of £300 between packages, it’s certainly worth considering if you’re going for a standard license.
Your mileage may vary, but the only software I could stand was MAXQDA, although it certainly isn’t without its own problems. If any computer programmers read this, you could make a killing programming a simple coding tool that allowed people to code textual data. No fancy bells and whistles, I just want to easily and quickly apply codes to my data.
But, out of the available systems, I would recommend you plump for MAXQDA 10. All the software packages I tried have their own unique problems and issues, but MAXQDA is far and away the most usable interface and has the quickest coding system. It also has support to open PDF and .docx files (Word 2007 or 2010) which many of the packages don’t support, even the proprietary ones.
If you’re forced to shell out upwards of £800, you want want to give some thought to the free packages, particularly WEFT QDA. However, as a student I would definitely recommend shelling out the small price for MAXQDA, as it’s much quicker to get the hang of. Just make sure you check if you need to use a specific piece of software for your assessed work (ATLAS.ti is more common). I would still argue it’s worth spending the money if your primary business relies on qualitative data analysis as the difference is significant.
This verdict comes with a caveat: I’ve only tried the basic coding systems for each software package. I have no need for many of the fancy bells and whistles that, to me, seem like feature creep, rather than useful additions. But, you never know, these might be useful to some people. All come with trial versions, so don’t be afraid to give them a try.