Home | Intro to SNA | Start-Up | First Example - MyEgonet | MyNames: Collecting egonet, network data (section two)

MyNames: Collecting egonet, network data (section two)

This section assumes you have your MyNames data organised into VNA format.

For good housekeeping [see … above] check that you have a Windows folder labelled MyNames (or similar) and it is the UCINET default folder.

Task One: Read the (VNA) tie data into UCINET (DL Editor ->Edgearray1)

There are multiple ways to enter tie data into UCINET. NetDraw can read the tie data list but it is best to learn about the DL Editor in UCINET.

Copy your tie data list onto your clipboard.

In UCINET, open Data ->Data editors ->DL Editor.

Paste in your tie data list (use Edit ->Paste or Ctrl-V) and go the format drop-down list to select “Edgearray1 (ego alter rel1 rel2 ..)”. Note the Edgearray format includes the Edgelist1 format further up the list.

  1. UCINET sets its defaults for the edgearray1 format.
  2. It automatically designates the first row as column headers. (Make sure you included them in your clipboard copy.
  3. The output options include the option to force symmetry. (The image of ‘symmetry’ comes from matrix formats.) This option will treat each tie record as two di-ties, A to B, and B to A. This doubles the number of di-ties in the data. For this exercise (MyNames) leave this unchecked.
  4. To create a UCINET dataset use File ->Save UCINET dataset or the fourth ‘save active sheet to disc’ icon.
  5. The tie data will save into the default folder. I advise calling it MyNames-net as you will also have node data files in the folder.
  6. UCINET displays a new logfile with the data in matrix format.
  7. To see the dataset in your MyNames folder use UCINET’s Data ->Display menu, the D icon or simply Ctrl-d.

Open NetDraw from within UCINET then open the (tie) dataset in NetDraw.

  1. NetDraw has automatically applied the graph theoretic layout. As the ego node has the most ties it is in the centre.
  2. The di-tie arrows only go from ego to each alter since we did not symmetrise the data.
  3. There is no system to the arrangement of the alter nodes.

I recommend that you usually load tie data into UCINET directly using the DL editor (Edgearray1).

Task Two: Load node data using NetDraw

SNA software packages have fairly good ways of transferring tie datasets among themselves. The real problems arise when dealing with node data. This is problematic even within the UCINET/NetDraw interface. I recommend that you have separate node data input files for NetDraw and UCINET program as NetDraw will accept character data but UCINET requires numerically coded data.

The key element in loading node data is that the nodeIDs in the two lists are exactly, exactly the same. I once spent five hours on a student dataset before finding that one nodeID had a white space after it and, hence was not matching correctly!

Loading node data into NetDraw is that you can retain and see actual value labels for each variable rather than numeric codes. The one difficulty is that you cannot have any white spaces in the variable names, or in any of the data cells. Here are my procedures for dealing with this.

Enter the node data through NetDraw as a VNA node attribute file.

  1. Set up variables in columns. The variable names (Col headers) must be cleansed of white spaces.
  2. There should be no blank cells. Put in a ? or NA or other character.
  3. Check that you have eliminated any white spaces and special characters in the actual data.
  4. Insert a row and type “node data” (with the white space) in the first cell.

Save the worksheet as a text (Tab delimited) file *.txt.

If you open the .txt file it will look like this.

The node data is now ready to read in but first you need to read in the tie dataset.

Open NetDraw and read in the UCINET tie dataset.

  1. Once the tie data is read in, you will Open this .txt node data file as NetDraw (VNA Attributes). If the node IDs in the tie data and node data do not match you will get an error message. (No message, it is ok.)
  1. Ego node placed at centre.
  2. Node attributes (Column headers) appear in a drop-down list in the box at the top of the Node tab.
  3. To cross-check: Go to the last variable and check that the values you expect are all there.

If you want to quickly view node data when in NetDraw use Transform ->Node attribute editor. You should however make all corrections to your original Excel worksheet. Then Save as *.txt.

Leave a Reply