Hi all<br>
<div><br></div><div>My name is Jenny Long, I'm a PhD student working on a social networks project in schools.</div><div><br></div><div>I'm currently revising a social network survey after running an initial pilot. My results suggest that networks based on the top 5 friends (from a list of ten) are as useful for predicting behaviour as networks based on the full list of 10 friends based econometrics spatial filtering, spatial lag etc analyses. I'm tempted to shrink my friend question down to the top 5 friends in order to limit the burden on students</div>
<div><br></div><div>Before I do that, I want to check whether using 5 friends to build a matrix (over about 1000 students) would be an issue for using SIENA at a later stage. Does anyone have any comment based on their own experience or their knowledge of the mathematics that underly these models?</div>
<div><br></div><div>Are there any other critical features that I need to think about if I want to apply SIENA to my longitudinal results? (I am familiar with those outlined in the latest SIENA manual)</div><div><br></div>
<div>Thank you in advance for your help</div><div><br></div><div><br></div><div><br></div><div>Jenny Long</div><div>PhD Student</div><div>University of Auckland</div>