Weft QDA User's Manual

Most textbooks and articles on qualitative data analysis stress the importance of choosing a software package that fitting your own particular method and methodology. Weft QDA offers a generic set of facilities for working with text documents, and doesn't make any particular assumptions about how to think about and generalise from data. At the same time, there may be a fair number of other CAQDAS packages available to you, especially if you use Microsoft Windows, and especially if you have money. So the following list summarises some situations where Weft QDA may be particularly suitable, and some where you might be advised to use an alternative software tool.

There are a number of different ways to install Weft QDA. They're listed below, in descending order of ease. You only need to use one of these methods, and unless you're interested in the technical aspects of Weft, you should probably choose an installer or compiled binary.

Weft QDA is being actively developed, and versions are periodically released with new features and fixes for bugs. Upgrading Weft should be straightforward, whether you use an installer or are using a source package and have already installed the prerequisites yourself.

There may be more than one 'current' version of Weft available at any time. The stable version, with a version number 1.x.y, where x is an even number, is the most tested and reliable release. This is the one recommended for general use. An experimental version has a version number 1.x.y, where x is an odd number. Experimental versions have the newest features, but may have more bugs, and less documentation. They are intended for people who want to test out the latest features.

The Weft Project File Format (*.qdp) can be read by any version 1.x.x of Weft QDA. Very old pre-release versions of the software - 0.9.5 and below - may not be able read projects created in newer versions of Weft QDA.

Once you have made changes to a project by importing new documents, creating categories or marking text, you can save these changes to the project file. Choose Save from the File menu. If you are working on a new project Weft QDA will prompt you for a name and location to save the project to.

If you want to discard changes to a saved project and return it to its last-saved state, choose Revert. Weft will warn you that all your unsaved changes will be lost.

Documents are added to a project by importing them from files saved on your computer. Weft can import documents in the plain text and PDF formats. The process of of importing a file copies the text in the file into a new document. It also indexes the text in the file so that it can be quickly searched later.

Once a document has been imported from a file, it can be read and marked for the categories it contains. To open a document for reading, either double-click its name in list in the Document and Categories window, or highlight its name and press the View button.

Figure 4.1. Using the document list to open a document

A window will open within the workarea, showing the document's text. Use the scroll bar on the right of the window to move through the document. The next chapter explains how to use the coding buttons at the bottom of the screen.

Figure 4.2. A document opened for viewing

Within the document's window, click on the 'Details' tab. Within the text boxes you can enter a new title and a memo for the document. Before closing the document, save any changes by clicking the Apply.

Figure 4.3. The document details pane

To delete a document, simply highlight its name in the document list in the Documents and Categories window, and press the Delete key. Alternately, highlight its name and choose Delete Document from the Project menu. The document, and any associated memo and coding will be deleted.

If you are working with interviews where you have ask different respondents similar questions, one place to start is to create categories that mark text where a particular questions was asked in different interviews. This will allow you to come back and compare people's replies side-by-side. First, it's a good idea to create a single cateogry in the tree under which to store the different questions. This will help organise your categories, as you create more as you develop and refine your analysis.

Figure 5.1. Creating a new category to organise 'interview questions'

To then add categories to represent the questions, click the New button with the newly-created 'Interview questions' category selected in the Category tree. Type in the name - a short summary of the question is good, such as 'Where living before?', if the respond was asked 'Where were you living before you moved ?' then click OK

As you read documents that have imported, you can mark passages of text as relating to one of the categories you have created. First, to mark a document, open it by double-clicking its name in the Documents & Categories window.

There are several ways to choose the category that you want to mark the open document with. The first of the methods below is simpler, the others may be faster for more experienced users, or when working with a large number of categories.

Once you have chosen the category, select the part of the document that you want to code by highlighting the relevant text in the document window. To highlight a passage, click the mouse at its start, hold the mouse down down, and drag the cursor to the end of the passage.

Figure 5.2. Marking a text passage from a document with a category

Click the Mark button at the bottom of the document window, and the text is marked as relating to that category. As an alternative, press CTRL+mto mark the selected text passage.

Note how, Once you have done some marking, as you select a category, passages already marked by that category are shown in a highlighted colour (Note: Windows only).

To remove marking by the currently selected category from document text, select the relevant passage and click the Unmark. The keyboard shortcut CTRL+, may also be used to unmark text.

After marking some of your data with categories, you can then review all the document sections coded by a category side-by-side. This allows you to compare them, seeing what they have in common, and how they differ.

To view all the text marked by a category, either double-click its name in the tree in the Documents and Categories window, or select its name and press the View button below. The category opens in a new window in the workarea.

Figure 5.3. Viewing a category's text

All the text from different documents marked by the category is shown in sequence. Each passage of text has a heading before it indicating which document the text comes from. The passages are shown in alphabetical order by document title. The numbers indicate whereabouts in the document the text is taken from, in numbers of letters.

Memos allow you to record further information about a category, such as the specific types of things it refers to, or notes about how it was developed.

To create or edit a category's memo, open the category, then click the details tab in the category's window. Type the memo for the category in the large text box provided. You can also edit the category's name by typing in the small name box at the top.

Figure 5.4. Changing a category's name and creating a memo

To save your changes, click the Apply button in the bottom right. To discard any changes, simply close the category window without applying the changes.

To delete a category, select its name in the category tree in the Documents and Categories window, and press the 'Delete' key on your keyboard. Alternately, choose Delete Category from the Project . The category, categories attached to it, and any memos and coding associated with them will be deleted from the project.

As the analysis develops and the number of categories increases, you may find that they need to be re-arranged to keep the category tree manageable.

To move a category, go the category tree and drag the category from its current position to the category under which you would like it to be attached. The procedure is similar to moving a file in Windows Explorer. The category will be moved to the new position, along with all the categories that are attached underneath it.

As well as recording observations about passages within the text of a document, sometimes you may also wish to record facts or observations about a document as whole. Commonly, where documents are interviews, the attributes one might wish to record include known facts about the individual interviewed, such as their sex, age range or membership of a designated part of a sample. You may also wish to apply categories developed in the course of your analysis to documents. Document attributes can be used in conjunction with the search tools to examine the incidence of text marked by certain categories in particular types of documents.

This can be done by marking all the text of a document with a single category. To record an attribute of a document, create a category to represent the attribute. For example, if you wanted to be able to relate the words of interviewees with their sex, you might create a category "Sex", with two sub-categories "Male" and "Female" beneath it. To note that an interview was with a man,select the relevant category in the category tree, open the corresponding document, select all the text in the document - the shortcut CTRL+a is usually available - and click the Mark button.

Weft QDA can search through a set of documents and return all the occurrences of a given word or phrase. In Weft QDA, the text results of a text search are presented as a list of extracts, similar to text coded by a category. The results are also automatically saved for later reference.

Queries allow you to ask more sophisticated questions about the categoriesin your project. For example, you might be interested in how often text marked by one category is also marked by another category. Or you might wish to do a text search only within text marked by a certain category. Queries allow you to explore a project in this way.

To start a query, choose Query from the Search menu. A dialog appears to allow you to create the query:

Figure 6.2. A blank query

An example of using queries

Say we have a set of interviews in which parents talk about how comfortable and adequate they find their homes for living with children. Some say they find they have a particular need for more space. So we have a category called "Space needs", which we have used to mark all the passages where the interviewees talk about this topic. You've also recorded the age of the children in the house, using categories for different age groups. (see the section called “Using categories to record document attributes”). We particularly want to look at what the families with older, school-age children say.

To turn this is into a query, we want text passages that are from families with school age children: document text CODED BY the category 'Primary school age' OR CODED BY the category 'Secondary school age'. Then within those we want to look at passages CODED BY the category 'Space needs'. So we create the following query:

Figure 6.3. An example query

To start a code review, choose Code Review from the Search menu. A new window opens in the work area. Choose the categories you wish to compare in the grid, either by selecting them from the main category tree in the Documents and Categories window, or by typing the first few letters of their name in the dropdown and pressing Enter. Then click the Add as row or Add as column button. To remove a category that has been previously added to the grid, click its name in the row or column label and press the Remove button.

Once there are categories in both the columns and the rows, the grid cells show the number of documents in which text is coded by both the row category and the column category. Each cell within the grid is tinted more or less strongly based on the number of documents in which the coding of the relevant categories coincides.

Figure 6.4. A code review

In the example above, the project documents are interviews with people bringing up children in a neighbourhood. Some of these families already had ties with the area when they moved into their house, such as having relatives nearby or having lived there before. The whole text of these documents has been coded by a category called 'Is Local'. Other families had moved in without existing ties to the neighbourhood, and these are coded as 'Not Local'. Each interviewee was asked how likely they were to move out of the neighbourhood in the coming five years, and the section of the document where they talk about whether and why they might leave has been coded by a category 'Likely to move away' or 'Unlikely to move away'. The code review indicates that a large proportion of "local" families think it is unlikely they will stay in the area, even though the interviews had suggested that these families were generally much happier with the neighbourhood as a place to raise children.

The number of documents coded is usually the most suitable measure when comparing the incidence of coding between documents that are cases, as in this example. In other situations, we may be more interested in the coincidence of particular coded analytic themes within documents. Using the dropdown menu at the bottom of the code review window, it is possible to change the display in the coding grid to show the number of separate passages which are coded by both the row and column categories, or to show the number of letters in those passages.