Image0.jpgAn analyser to assist in categorising and managing textual research data


Are you doing post graduate research?
Have you undertaken 17 interviews and exhausted yourself transcribing them?
It's difficult to know what to do next, isn't it, but this package will help you.

These notes concern a Microsoft Access database system that is intended to assist with the analysis of qualitative data in research. That means interview transcripts, documents, web pages and other sources of text.

There are two versions of the Analyser:

Here are some features of the "Starter" version:

  • The "front end" and "back end" are maintained as a single package, so you just make a fresh copy for each project and data set (the simple approach).
  • Table and field names have been generalised so as to deal with the content of documents and other sources, not just interviews.
  • You can set up your own labels for the metadata.
  • Navigation between chunking, coding and browsing is now easier.
  • Your last-edited chunk and category is remembered.
  • All words in your data can be indexed with a single click.
  • You can choose to use Pivot Chart controls to refine the frequency analysis according to your needs.

Here are some additional features of the "Full" version:

  • The front and back ends are separate so that you can easily work with multiple data sets (helps to manage your complexities).
  • A record of your work sessions is maintained, making it easier to work with multiple data sets and you can immediately go back to any point in your history of work.
  • Bulk metadata from RIS dumps (such as from Zotero) can now be imported automatically, including annotations and comments from your PDF files.
  • Category structures can now be organised as trees or networks (ontologies) and can be easily organised using drag-and-drop.
  • You can now browse through category structures using a "Tree view"* or "Hi-Lo view".
  • Chunking of text and coding of categories is made easier with an auto-chunking and auto-copy feature (takes practice, can save lots of time!).
  • Category allocations (coding) can be reviewed individually or globally and category structures can be imported and exported.
  • The analyser offers a "pairs analysis" capability that corresponds (I think!) to (statistical) factor analysis of quantitative data, and contributes to the development and testing of ontologies.
  • Ad hoc subsets of your chunks can be selected, and output as material for thematic development through the clip board (for pasting into your work, with attribution, date and place already assembled and inserted).
  • There is a diagnostic facility that scans the data and provides pointers to possible problems and some simple statistics to guide you.

[* The "tree view" requires a particular Microsoft "ActiveX" component file to be loaded and registered with Windows, which is sometimes not installed - it is only needed for the "full" version of the analyser. There are instructions for checking and installing it here.]

Bad news: This analysis package has been developed using Microsoft Office Professional 2010. Functionality may be limited in earlier releases of the office suite. We are working on ideas for scalable, distributable versions of the package.
Good news: The STARTER version of the Analyser now works with the free-to-download Access "Runtime" Package, available from Microsoft's web here (or just Google for "Microsoft Access 2010 runtime version download".
Tentative news: One or two researchers have successfully operated the QCA package on their MACs, using Windows emulation. Let us know if you try this, we are feeling our way here.

Community

The analyser has been conceived and implemented by Andy Bytheway. It is a work-in-progress, and therefore the details of what can and should be done is certain to change. In order to make it better, we need to build a community that will take this package forwards. If you are interested, or if you have any comments or queries, please initiate a discussion on this wiki, or contact me directly.