An Introduction

Maybe at first one should explain the name of the program - or should one first mention that the roots of the program are in Australia - since that by itself might already count as a good enough explanation J ? More seriously, QSR stands for Qualitative Solutions and Research and NUD*IST stands for Non-numerical, Unstructured Data * Indexing, Structuring, Theorizing.

The orginal developers of QSR Nud*ist are Tom and Lyn Richards. Tom is a computer scientist and Lyn a social scientist. And as so often with QDA software, before it existed, there was a project that needed to be analyzed. This is also how Nud*ist came into being. Lyn was working on a large qualitative project and Tom wrote the appropriate computer code to facilitate the anaylsis of the data (at least this is what the rumors tell traveling from all the way down under to this part of the world). Nud*ist, like all of the other QDA programs, also shows the influence of its creators. If you are interested in the ideas that underly the development of Nud*ist, have a look at the following references.

Nud*ist 4 features

UP

Coding

Coding can be done interactively while you read the text on screen or, if you prefer, on a hard copy of your data. Coding done on paper can be transferred to the computer by using line numbers as reference points. All coded segments will be stored in a node. Nodes in Nud*ist are presented as little boxes carrying the code word name and a number. Nodes can either be a free or are part of the index tree. The index tree can be used as a way to represent your coding scheme in a hierarchical form.

The smallest segement you can code is a text unit. Text units are something you define for yourself - text units are as long as the distance to the next hard return. In other words, if there is a hard return at the end of every line, the smallest chunk of text you can code is one line. If there is a hard return at the end of every paragraph, then the the smallest codeable text unit is one paragraph. Thus a decision about text unit size should best be made when transcribing your data. A fairly common procedure is to choose lines as text units since then you can simply save the transcriptions as "text with line breaks", an option you have in any word processing package.

It is possible to make your coding visible by specifing the code words you want to see (up to 26 at a time). You then see the codes displayed against the text in the margins.

Another possiblity is to index off-line documents. This means that you can reference documents that cannot be put into a computer readable format. For example, you code the second paragraph on page 15 of a particular book with codeword X. When retrieving coded segments from such documents in a search, Nud*ist will then tell you where to look for in the original text.

UP

Memos and Annotations

In Nud*ist you can write one memo for each document and for each node. Memos can be date and time stamped, they can be as long as you like them to be and are editedable. They are displayed in their own separate windows, can be saved in ASCII text format, and its content can be cut and pasted somewhere else. They are directly attached to the object they belong to. Annotations can be directly inserted into the data at any time in the analyis process.

UP

The Index Tree

The Index Tree is a graphical representation of your coding scheme. It needs to be understood as an upside down tree. The top of the index tree is the root and from there all the branches ramify. The Index Tree offers you a hierarchical way of organizing your data by categorizing them into higher and lower order categories, into parent and child nodes. Nodes can be freely moved around the tree. You can cut them at one place and paste them to other parts of the tree. You can also copy them and paste them somewhere else or merge them with other nodes. Another option is to delete an entire nodes or only the coding that is contained within the node or the memo that you have written for it.

It is possible to create the tree before you start coding your data. In fact, it is very common to set up certain parts of your tree right when you start a Nud*ist project. This part of the tree is generally referred to as the base data tree. It can be used for coding factual information that you have about your data like gender, age, race or occupation. If you have time series data or have collected data at different locations, you can create nodes that reflect this structure and code your data accordingly. Since this is a fairly routine process, it can be automated by using command files. If the factual information is available in form of a spread sheet, you can use the Importing Table function a) for generating the base data tree and b) for coding your data accordingly.

UP

The Node Explorer

The Node Explorer offers a sideways representation of the coding scheme (or index system) similar to the representations of data files in the Windows 95 Explorer. Within the Node Explorer you can cut, delete and merge nodes, you can browse the text that is coded at a node, you can change node addresses, write and review memos or modify the coding at a node. In addition, the node explorer is a very handy device when it comes to searching. From there you have direct access to the results of the search operations.

UP

Retrieval

Text Search and Auto Coding

You can search your data by using simple string searches or the more powerful pattern-based text search facility. While searching your data, you can restrict the search to particular documents, exclude certain documents or confine the search to text passages with certain chosen characteristics. Every find (or hit) can be coded automatically, but one has also the option to control this process. This means that you can first review a hit and decide for yourself whether it is worthwhile to code it or not. One benefits of such a feature is that it allows you a quick entry into your data (see also Tips & Tricks: Auto Coding).

The Search Engine

The Nud*ist search engine is a very powerful tool. It offers you 17 different ways of asking questions about your data including Boolean, context, sequence, proximity, matrix and vector searches. The results of a search are saved on the node clipboard. In addition, a node containing the search results is automatically created and put at a designated place of the Node Explorer. Findings can be reviewed and browsed by accessing this node or via the node clipboard.

Having the results saved at a node (in the Nud*ist language this is called "system closure") offers you one big advantage - you can ask further questions about specific parts of your data. This way it becomes possible to build up more and more layers of analysis and to go deeper and deeper into your data.

UP

Importing and Exporting Tables

This feature of the program allows you to link qualitative with quantitative data. If you have some quantitative information about your data in form of a spreadsheet, maybe from a survey or some other statistical publication, you can import it into Nud*ist. Nud*ist will automatically create the index tree (most likely this will be your base data tree) and code your data accordingly.

Exporting Tables is more or less the opposite process. You export outcomes of the analysis for further analysis to statistical packages. This could be for example your base data coding or those parts of your coding where you have defined typologies or categories. The results of matrix and vector searches are also appropriate candidates for exporting.

UP

Command Files

Many processes in Nud*ist can be automated by using command files. Even those people who don't think of themselves as being computer wizards will probably also make use of some of the simpler command files, like for example the one for importing data files. Command files are especially helpful if you have to perform certain tasks more frequently.

Nud*ist makes the use of command files fairly simple by providing a number of templates. From there you can copy and paste the commands you need for the tasks you want to perform. It is also possible to save search queries as command files. This way you can keep a record of the search operations you have performed and you can make use of the command files for latter searches.

UP

Model Building - Exporting and Importing the Index Tree

You can export the Index Tree to either Decision Explorer or Inspiration. This allows you to break away from the hierarchical structure and to freely play around with the nodes, thus to enter the phase of model building. While you loose the direct access to your raw data, in Inspiration you can still view and edit the nodes' memos. The advantage of exporting the Nud*ist Index System to either Decision Explorer or Inspiration is that you can use programs specialized in model-building and diagramming in a complimentary way to the facilities offered by Nud*ist. Inspiration might be a good choice if you want to display and built very flexible conceptual networks. Decision Explorer offers good facilities for mind mapping like for example the graphical display of cognitive maps and conceptual links. A new function offered by Decision Explorer is an interface that allows you to import hierarchical models created in Decision Explorer to Nud*ist.

UP

Other Features

• Adding data and Editing Documents: New data can be appended to existing documents and already imported documents can be edited. This includes inserting or removing text units, sub headers, and annotations.
• Support for case based analyses
• Merging Projects: This feature facilitates team work. It allows to compare and to contrast two different projects or to build a central project from multiple sites. It is also possible to use the merge tool across different platforms (i.e. Macintosh and PC platforms).
• Videotapes: Data from videotapes can be handled by using the CVideoä software for Macintosh.

UP

References

Richards, T. and Richards, L. (1995). Using Hierarchical Categories in Qualitative Data Analysis'. In: U. Kelle (ed.) Computer-Aided Qualitative Data Analysis, London: Sage, 80-95.

Richards, L. and Richards, T. (1994). From Filing Cabinet to Computer. In: A. Bryman and R.G. Burgess (eds.) Analyzing Qualitative Data. London: Routledge, 146-172.

More references can be found under: http://qsr.com.au/OtherInfo/OtherInfo.htm#References

 DISCLAIMER: The descriptions reflect my personal experiences with the programs and should not be (mis)understood as an objective report