Is anyone well-versed in QSR such as atlas.ti, NVivo, and TAMS Analyser? I’m wondering to what extent these programmes duplicate or complement the functionality of DEVONthink, as well as how they compare with each other. Has anyone come across any good recent comparative reviews of the products?
Thanks for any info or ideas.
Follow the links at caqdas.soc.surrey.ac.uk/ for a variety of resources relating to qualitative analysis software.
On the Windows platform the main programs, used by sociologists, anthropologists, journalists and the like are, in no particular order, Atlas.ti; Maxqda, and Nvivo. These programs are most useful where one wants to work with passages of text rather than with the individual words in a text or the document as a whole, although all of the programs mentioned above do allow you to generate concordances and indexes, and to add metadata to documents. The basic paradigm is one that has become known as ‘code and retrieve’. That is, you identify passages of text in a document that are of interest to you and attach to them a ‘code’. This is the preferred term in the field but ‘codes’ in this context are also referred to as tags, keywords or index terms. The point is that, whatever they are called, the software makes it possible for you to retrieve the codes, and more importantly, the text associated with them, often through the use of complex boolean searches (although other kinds of retrieval based, for example, on proximity searches, are also possible).
The market for qualitative software is a very Windows-centric one. For the Mac, one is looking at TAMS Analyzer. This is free software that has been getting better over the years but depends on a system of text markup that very quickly becomes cumbersome. HyperResearch is now showing its age and has also always had issues about its coding and retrieval functions. Transana is oriented to the analysis of video rather than textual data.
The Windows programs mentioned above tap into what has now become a very substantial market indeed. The continued howls of pain on the Devon forums about tagging and annotation point to the relevance of such software tools for very many users. I’ve never quite understood why Mac developers have been so slow to tap into this market, especially with the advent of Leopard.
That was most helpful, marcellus. I went right ahead and downloaded the Lewins & Silver PDF and ordered their book. The lessons I took from your message were that, while DT might perhaps be twisted into doing something like the software in question does, their purposes are very different, and also that I’d probably be best served by one of the “Big 3” commercial packages.
I do find it interesting how different platforms seem to attract different kinds of software and how pricing works. I haven’t seen any evidence that qualitative research software is any more advanced or difficult to write than DevonThink, but its Windows-only, specialized nature seems to be a licence to sell it at a much higher price (not owning any of the commercial packages, of course I might be totally wrong about this). On the other hand, DT/DN has become associated to some extent with various “snippet keepers”, so my impression is that the developers have to keep something available at around $40 or less to keep the affections of the Mac community at large.
I guess I’ll pick this sort of thing up by going through the PDF and the book, but do you have any perspectives you’d care to share on the relative merits of the Big 3 commercial packages you mention?
caqdas.soc.surrey.ac.uk/Choosing … July06.pdf
In my eyes, the one most convincing advantage of DT is the see also function and the fuzzy search. A problem with ocr’ed Pdfs is that it is hard to find specific passages because of ocr errors. Fuzzy search helps a lot.
Having said that, my first thought after browsing through the software you mentioned was: why why why, why can’t I do this on a mac. It makes me realize that DT currently caters for the needs of “snippet keepers” instead of tapping the academic market (cf. the skim thread). I still hope DTPO can become a tool for making sense of huge amounts of data.
My guess is that early in the OS X era market share was too low to make it worthwhile (and people were scared of Cocoa); and now market share is up (though still not very high), but we can run Windows on a Mac so people are prepared to buy Parallels and Windows just to run a key program.
I think you’ll find the Lewins and Silver book a big help. I also think you should download the demos for the Big 3 and play around with them. To some extent, competition between the developers has meant that in terms of features the major programs pretty much allow you to do the same kinds of things, albeit sometimes in different ways.
One key point, I think, is to think about your own style of analysis and to see which of the programs fits your style best, rather than adapting your style to a particular program. The other thing to keep in mind is that each of the major programs does have fairly substantial learning curve, so playing around with each package in turn is a useful thing to do.
If you do want to stay with the Mac and can’t tweak DT, then TAMS Analyzer is perhaps a look. I’m not entirely comfortable with it but perhaps other forum members might like to register an opinion.
Hope this helps.
My fault for talking about other types of software on the DEVON forums, but I’d just like to clarify that my original question wasn’t intended to imply any criticism of DT (I’m a generally happy user of DTPO). I figured that there’s usually someone on the DEVON forums who knows about pretty much anything of interest to academics.
Having said that, I am interested in opinions such as Maak’s, who seems to be suggesting that DTPO ought to be able to expand its functionality to cover the CAQDAS field. In the PDF mentioned in previous messages, I was particularly intrigued by mention of “the new generation of qualitative software, crossing methodological boundaries and mixing methods” (p.34), in connection with QDA Miner from http://www.provalisresearch.com. I’m wondering whether perhaps that’s something like what DT is aiming to become in future versions?
The following post on the Scrivener forum might be of interest:
literatureandlatte.com/forum … =19&t=3723
I notice that MAXQDA now has a fairly nice series of tutorials (though still early days yet) at
A quick update on MAXQDA’s tutorials: There are still only 4 of them. They’re very nice, but it seems a little unfortunate that they retired the old help system in favor of these before they’re finished.
One rather old method for the manual analysis of qualitative data involved making multiple copies of transcripts, cutting up relevant passages (i.e. explicitly producing ‘snippets’ of text) and pasting them onto cards which could then be sorted for themes. You can emulate something like this in DT by selecting some text, taking a note and tagging it, and then saving into a ‘snippets’ folder. You can then retrieve relevant snippets from the tags view or via search. Not the most elegant or efficient method and I haven’t ever tried to scale it up to a full-size data set, but it’s worked reasonably well for me on a small project.
Like many others, I would like to see greater qda functionality in DT.