Analytic Functions in Networks
The Network tool is an analytical tool in itself. You will learn a lot about your data by building conceptual networks. The two options that we would like to present here are:
- Add Neighbors
- Add Co-occurring Codes
These two options are the visual equivalent of the Code-Document Table and the Code Co-occurrence Table.
For both options it is often useful to set a global filter first as otherwise too many items are imported. This results in a very crowded network that might look nice but is of little analytical value. See Applying Global Filters.
The add neighbors option allows you to construct a connected network step-by-step. It imports all direct neighbors of the selected nodes into the network.
When applying it to document nodes, it is the visual equivalent to the Code Document Table. ATLAS.ti draws a blue lines between the codes, and the documents / document groups in which they occur.
When applying it to hyperlinks, it allows you to bring into a network all connected quotations.
When applying it to codes, you can successively bring in connected codes to support data modelling and theory building.
How to add neighbors
Select one or more nodes, right-click and select the option Add Neighbors from the context menu and then the type of entity that you want to import.
You also find the 'Add Neighbors' option in the Nodes ribbon of the network editor.
For documents when adding codes, make sure that Show Code-Document Connections is selected under the View tab in the ribbon.
Depending on the entity type, possible neighbors are:
The option all common imports all those that are applicable.
Application: Comparing Cases
Adding code neighbors to a document node, or a document group node allows creating case-based network. You can ask questions like: Which of the codes have been applied in which document or document groups. ATLAS.ti draws a blue lines between the codes, and the documents / document groups in which they occur. Make sure that Show Code-Document Connections is selected in the View options.
When you have applied lots of codes to a document, it is useful to set a global filter when adding codes. Otherwise, your network will be too crowded, and it will be difficult to interpret it or to see any meaningful pattern. See Applying global Filters.
Add Co-occurring Codes
The add co-occurring codes option is the visual equivalent of the Code Co-occurrence Tools. It allows you to explore code co-occurrence spatially, which is different from looking at numbers in a table, and different from exploring code co-occurrence in a Sankey diagram. The recommendation is to use these functions in combination to look at data from different perspectives in the sense of a triangulation.
Co-occurring codes are those that code the same or overlapping quotations. When importing co-occurring codes, ATLAS.ti checks all five possible ways that a code can co-occur:
quotations coded with a code A overlap quotations coded with a code B
quotations coded with a code A are overlapped by quotations coded with a code B
quotations coded with a code A occur within quotations coded with a code B
quotations coded with a code A enclose quotations coded with a code B
quotations coded with a code A code the same quotations coded with a code B
See Available Operators > Proximity Operators.
Select one or more codes in a network editor, right-click and select Add Co-occurring Codes. This option is also available from the Nodes ribbon.
In order to achieve meaningful results, it is often useful to set a code group as global filter first. Think about which concepts you want to relate to each other and add those to a code group. See Applying Global Filters.
Below you see a network with global filter setting. Only the full colored nodes pass the filter. It helps you to focus on what is relevant in your network:
If you have no idea which codes co-occur, using the Code Co-occurrence Explorer is a good starting point. If you have a more specific idea which codes do co-occur, the Code Co-occurrence Table, and the "Add Co-occurring Codes" option in networks become more useful.