Contents
- Index

**NETWORK>COHESION>HOMOPHILY**

**PURPOSE **Calculate measures of homophily on the whole network associated with a given partition..

**DESCRIPTION **Given a partition of a network into a number of mutually exclusive groups then the E-I index is the number of ties external to the groups minus the number of ties that are internal to the group divided by the total number of ties. This value can range from 1 to -1 and can be seen as a measure of the extent a group chooses themselves a value of -1 showing homophily and a value of +1 showing heterophily. For valued data it is the sum of the tie strengths instead of the number of ties.The procedure also gives two other correlation based homophily measures by correlating the dichotomous variable "is in the same group as" with the corresponding entry in the data matrix. Point biserial correlation is specifically designed for dealing with one dichotomous variable and this is calculated together with the standard correlation. Note that if the data matrix entries are dichotomous these two measures are the same. The E-I index treats the edges as binary and ignores any values on the edges.

**PARAMETERS**

** Network Dataset**

Name of UCINET dataset to analyzed. Data type: Valued Graph.

**Dataset containing Row Partition (defining groups of nodes)**

** **Name of a UCINET dataset that defines the row partition (an attribute dataset) either in its columns or in its rows. All entries in the identified column (or row) with the same value will be put in the same partition. The user must select whether it is the rows or columns of the dataset that define the groups. Finally the user must select which column (or row) contains the requited partition data.

**Output Homophily Scores**

Name of UCINET file that contains the whole network homophily scores the default is <filename>-homophily.

**Utilize diagonal (reflexive ties)**

** **Check if diagonal values are to be included.

**Valid:**

** **A list of the groups by number, all checked groups are included but unchecking a box removes a group and their ties are ignored.

** **

**LOG FILE **Recoding of the attribute vector used to partition the dataset with a list of the size and members of each group.

A list of the number of edges (not taking account of any edge weights) within and between each group.

The E-I index, point biserial correlation and correlation for each relation in the dataset.

**TIMING** O(N^2)

**COMMENTS **For a full E-I index run NETWORK>COHESION>E-I Index

**REFERENCES **Krackhardt, David and Robert N. Stern (1988). Informal networks and organizational crises: an experimental simulation. Social Psychology Quarterly 51(2), 123-140.