ERGM-Practice

download the Practice.RData first.

SNA-assignment-3-by-Yiqiu-Wang.pdf

1 Explain why the mechanism of gender homophily might create networks in which there is a substantial degree of reciprocity and transitivity, even if no explicit mechanisms of reciprocation and transitive closure operate.

2 Explain why the opposite is not generally true, i.e., why the mechanisms of reciprocation or transitive closure do not generally create networks which are gender segregated.

3 Identify three R-functions that can evaluate a given network in terms of (i) density, (ii) reciprocity, and (iii) transitivity. Identify a fourth function that evaluates a network and a group indicator variable in terms of (iv) homophily on that variable.

4 Evaluate your personalised data set on these four dimensions by applying the functions from Exercise 3.

5 Please fit three exponential random graph models to the data.

6 Report the results in one table and interpret them in a brief text. What can a comparison of the coefficients from both analyses tell us?

7.1 Make four network visualisations: one example graph from each set of simulations, and the empirical data set, with nodes coloured by gender.

7.2 Make four plots, one for each index, each of which contains the three simulated distributions and the empirical value of the index as reference (i.e., the value you calculated in Exercise 4).

8 Try to substantiate your reasoning in Exercises 1) and 2) above based on these results.