Ties between departments

Post date: Nov 12, 2015 9:53:7 PM

If you have a network and an attribute of each person (such as which department they are in), you can get a table that shows the number of ties within each department and between departments. You can then compare this with what would be optimal (as judged by someone in the organization).

In the UCINET main menu, go to Networks|Cohesion|Density|Density by Groups. Fill out the form like this:

The output will include a table like this:

Frequency of ties

1 2 3 4 5 6 7 8 9 10

---- ---- ---- ---- ---- ---- ---- ---- ---- ----

1 3624 1720 439 38 746 247 225 916 339 172

2 1720 2280 386 109 679 559 505 418 398 84

3 439 386 866 68 253 277 834 123 211 20

4 38 109 68 506 151 389 1175 12 86 5

5 746 679 253 151 1672 1037 1305 322 251 36

6 247 559 277 389 1037 4584 1907 91 298 10

7 225 505 834 1175 1305 1907 6804 102 422 5

8 916 418 123 12 322 91 102 232 89 70

9 339 398 211 86 251 298 422 89 116 21

10 172 84 20 5 36 10 5 70 21 18

This gives the number of ties from the row department to the column department. However, it does not take into account the fact that some departments are larger than others, making it possible to have more ties to and from them. So a better table is the next table in the output, which is the density table:

Density table

1 2 3 4 5 6 7 8 9 10

----- ----- ----- ----- ----- ----- ----- ----- ----- -----

1 0.573 0.291 0.152 0.021 0.126 0.037 0.026 0.881 0.530 0.538

2 0.291 0.422 0.145 0.064 0.124 0.090 0.063 0.435 0.672 0.284

3 0.152 0.145 0.687 0.082 0.095 0.092 0.215 0.263 0.733 0.139

4 0.021 0.064 0.082 1 0.089 0.201 0.473 0.040 0.467 0.054

5 0.126 0.124 0.095 0.089 0.310 0.167 0.163 0.335 0.424 0.122

6 0.037 0.090 0.092 0.201 0.167 0.657 0.210 0.083 0.443 0.030

7 0.026 0.063 0.215 0.473 0.163 0.210 0.589 0.073 0.488 0.012

8 0.881 0.435 0.263 0.040 0.335 0.083 0.073 1.487 0.856 1.346

9 0.530 0.672 0.733 0.467 0.424 0.443 0.488 0.856 2.071 0.656

10 0.538 0.284 0.139 0.054 0.122 0.030 0.012 1.346 0.656 1.500

The 0.291 in the first row second column says that 29.1% of possible ties between dept 1 and dept 2 are actually present.