Dear Elena,

it depends what data you actually start from. But if I understand correctly you should have just an array of vectors of same length that you should bind to generate the Adjacency matrix.

In more technical but hopefully still simple terms, for each person-ego (i), you should have the connections of ego. This is a vector of length equal to the number of your population, say n. Each of these vectors, that should be ordered in the same way, is a row of the full network matrix N (the Adjacency matrix) you need. Then to create it, you simply need to bind (by column) your vectors. Mind, you need to follow the same order that you use to create the ego network. As a matter of fact, the Adjacency matrix is a nxn matrix where the order of the columns is the same of the order of the rows.

if you want to be a bit nerdy, you can use any programming tool, for instance in R you can first import each single 'ego-centered' network (or each simple vector), and then you use the command cbind. 

Or if you are ok with more easy to work tools, Excel for example, you just need to copy and paste each row one below the other until you have your matrix.

best of luck with your research, it sounds interesting

claudio

Elena, hello. I recommend exploring the theory of social representations of Moscovici Sergio.

• Moscovici, S. (1961). La psychoanalyse, son image et son public. Paris: P.U.F. (1976) segunda edición revisada (trad. Castellana de Huemul, Buenos Aires, 1979).

• Moscovici, S. (1981). On social representation. En J.P. Forgas (Ed.), Social Cognition . London: Academic Press (trad. castellana en G. Serrano y J. Sobra, Lecturas de psicologia social. Santiago de Compostela: Torculo, 1992)

• Moscovici, Serge y MILES Hewstone (1988). De la ciencia al sentido común, en Psicología Social, I, II. Barcelona: Paidós.

• Moscovici, Serge y Vigraux, Georges. (2001). The concept of Themata. En Gerard Durveen (editores) Social Representations Exploration in Social Psychology, New York University Press.

Dear Juan, dear Claudio,

Thank you both so much for your help! I really appreciate it.

I will write again as soon as I have the time to read through the material and look into R. But this might exactly be what I was looking for!

Thanks again!!

elena

Claudio hi,

I tried your solution. An IT-Nerd I know wrote an excel macro for me where I could easily combine ego-centered vectors and create an adjacency matrix.

But the next question follows: I want to collect network data at three time points (T1, retrospective at T1 for an earlier time period, T2 a year or two later) in order to analyze dynamic network parameters. I just read that it is only possible to calculate dynamic network analysis with RSiena if one has a whole/closed network like a school class. As you remember I want to combine my ego-centered networks because I want to create a closed or whole network. But my network might change much more from T1 to T2 than a normal closed network (like a school class). What is more my participants work at a big university and it is only in theory possible that all participants know each other (in contrast to the school class example). Therefore do you know if it is still possible to calculate a dynamic network analysis with this kind of data? Or do you have an idea who might know this??

Thank you very much in advance!!

Dear elena, I have a paper published on plos One just just last year with a dynamic network whose participants (thousands of scientists in my case) change in number, position in the network etc. All you need is to be creative and understand what you really want to understand from your data.

Once you understand your research question, then you can process the data and analyze it with more than one tool.

Dar Claudio,

thank you for the reference to your Paper. I will read it :)

Hi, 

for everybody, who is interested in this topic: I just got an email from Prof. Molina. He created EgoNet, which also can connect Ego-Networks to a whole network. And there might be no limit how many ego-networks you want to combine. It maps the networks using the names, but there is no possibility to map them using the attributes of the nodes, which would be really cool... 

hi Elena,

I think you need to learn how to program a little bit. To do 'cool stuff' with network, rudimentary notions of any programming language are ideal

Hi Claudio, 

 I hope you didn´t misinterpret my statement, I just wanted to inform people who read this conversation and are interested in the topic that there is a program that can combine ego-networks. I wanted to describe the features of the software, without judgement. It was really helpful for me to get the info about EgoNet and I just wanted to share this information.