Top of Page
What do you want to do?
Before you can start working with framework matrices, you must import your source materials, code source content to case nodes and set up thematic node hierarchies. The following table describes these steps in more detail:
| Step | Actions | Comments |
|
1 |
Import sources and code each person's responses to a case node that represents them |
If each person has a separate interview transcript (for example, a Word document or audio file), you can create case nodes from the sources—refer to Create nodes automatically (Create case nodes from sources) for more information. If you have focus group transcripts that contain the responses of more than one person, then you can import the transcript and create your case nodes manually, and then code each person's responses to their case node. |
|
2 |
Set attribute values (such as age and gender) for each case node you created |
Now that you have your source materials inside NVivo and each person's responses are coded to a case node that represents them, you can record information like age and gender by setting the classification and attribute values of the case nodes. You can set attribute values in a number of ways:
Refer to Classify nodes (Set attribute values to record information) for detailed instructions. |
|
3 |
Create a thematic node hierarchy |
You may have set up a node hierarchy for your themes before you imported your source materials, if not, then you should set it up now. You can create the node hierarchy manually, or you can import it from another NVivo project. Refer to About nodes for more information. |
|
4 |
Thematic coding (optional) |
You do not need to code your source materials to theme nodes in order to work with framework matrices, however you may want to do some thematic coding to test the validity of your node hierarchy. Alternatively, you can fully code your source materials before you start summarizing the data. If you have done thematic coding, you can choose to use NVivo's auto summarize feature to speed up the summarization process. Refer to About Coding for more information about the different ways you can code your source materials. |
|
5 |
Create framework matrices |
Now, that you have your source materials coded at case nodes, and you have a node hierarchy that you are confident will support your analysis, you are ready to start summarizing your source materials. You need to create framework matrices in order to summarize your data. |
|
6 |
Summarize your source materials using framework matrices |
The framework matrix contains case nodes (as rows) and theme nodes (as columns). The intersection of each case and thematic node is a matrix cell, and this is where you summarize the source content that relates to the case and theme. For example, you might summarize a respondent's comments about their family environment. By default, the associated view (on the right) shows all the source materials that have been coded to the case—for example, an interview transcript. As you read through the interview transcript, you can fill in the matrix cells in the row, summarizing what the respondent said in relation to each of the themes. You can add summary links to connect text in the summaries to the supporting source materials you see in the associated view. This means that you can always drill down to the primary data that supports your summary. NOTE If you have done thematic (as well as case coding), you can use the associated view to show content coded to both the case and theme. Refer to Work with framework matrices for more information. |
|
7 |
Share your summaries with others by printing or exporting your framework matrices |
Once you have summarized your source materials, you might want to share your framework matrices with colleagues or people outside your organization. You can print the matrix or export it to a spreadsheet file or text file. Refer to Print and export framework matrices for more information. |
A common approach is to read through each interview transcript chronologically, summarizing the content as you go. Taking this approach, you fill in the matrix row-by-row as you read through each transcript. This allows you to gain a thorough understanding of each case, and identify relationships between the different themes in your source materials.
You can do this as follows:
Code each interview transcript at the case node for the person you interviewed—a quick way to do this is to create case nodes from the transcripts.
Create a framework matrix that includes the case nodes and the theme nodes that are relevant to the questions you asked in the interview.
Click in the first cell in the matrix—the associated view on the right shows the complete interview transcript (because you coded it to the case node).
Read through the transcript chronologically from start to finish, summarizing the content into the cells in the row.
When you have completed the first row, click in the next row—the content in the associated view changes and you will see the interview transcript for the next respondent.
Continue filling in the matrix row-by-row.
Alternatively, you can focus on one theme at a time—this can be useful when the subject matter is very complex. If you have fully coded your source materials to thematic nodes, you can show cell coding in the associated view on the right—this makes it easier to find the source material that relates to the theme. Taking this approach, you fill in the matrix column-by-column, summarizing the responses of each person in relation to the theme.
If you have a focus group interview, then you might have the responses of five or more people in a single transcript. If you create a separate case node for each participant and code each person's responses to the case node that represents them, you can fill in the matrix row-by-row—focusing on each case and summarizing each persons responses against the themes in the matrix.
You can do this as follows:
Create a case node for each of the people who participated in the focus group.
Code each person's responses to their own case node—you may be able to use autocoding to speed this up, refer to About coding (Autocoding structured content) for more information.
Create a framework matrix that includes the case nodes for each participant and the theme nodes relevant to the questions you asked the group.
Click in the first cell in the matrix—the associated view on the right shows the responses of the first participant (rather than the entire interview transcript).
Read the participant's responses and summarize the content in the cells in the row. Because you are not seeing the entire interview transcript, you may need to see their responses in context—you can use the node view options to show the surrounding text.
Continue filling in the matrix row-by-row.
If you have fully coded your source materials to thematic nodes as well as coding to case nodes, you can choose to create summaries automatically from your coding.
Textual content coded at both row and column nodes is copied into the corresponding matrix cell. This gives you a starting point for your summaries—you can edit this text to reduce and synthesize the material.
It is a good idea—especially in a shared project—to agree on any abbreviations you want to use in summaries before you start work in your framework matrices.
If you want to use text formatting (for example, color or italics) to identify verbatim quotes or comments within summaries, you might also want to agree on the formatting you will use before you start.
You can record these decisions in a memo that everyone can access. For example, you might choose to always abbreviate certain words to save space, or use green italics to indicate verbatim quotes.
When you are filling in the matrix, you may want to keep the summaries roughly the same length—this can make it easier to read and review the completed matrix.
When you have finished summarizing, it is a good idea to check for empty cells. Empty cells may mean that you have missed some of the source materials, or that a particular theme was not relevant to the respondent (you may want to enter some explanatory text to indicate why the cell is empty).
If the matrix contains many empty cells, because the themes were not relevant to some respondents, you may want to consider changing the rows and columns in the matrix, so that the majority of the cells contain a summary. Removing rows or columns from the matrix does not delete the summaries you created—refer to Understand how summaries are stored in your NVivo project for more information.
When you are reviewing the matrix, you can show summary links in the matrix cells and highlight (or show only) the linked content in the associated view. This can be useful, if you want to drill down to the primary data and see the content that supports the summaries. Refer to Work with framework matrices for more information about showing links, highlighting source content and changing the content shown in the associated view.