Personal networks

June 23, 2010

Migrant network role graphUnderstanding the composition of an ensemble of personal networks
Ulrik Brandes, Jürgen Lerner, Miranda J. Lubbers, Christopher McCarthy,  José Luis Molina, and Uwe  Nagel
Ulrik.Brandes@uni-konstanz.de


3 Responses to “Personal networks”

  1. admin Says:

    PEER REVIEW COMMENT No. 1 – This network role graph offers a unique representation of the distribution of immigrant ego networks that nicely crosses levels of analysis. At a macro level, the visualization is organized along role types, where some ego networks are characterized by many ties to fellow immigrants while other ego networks are characterized by assimilation, where there are many ties to people of the immigration country. At the micro level, ego networks are placed in a multidimensional space, showing how close they are to ideal role types. The role graph display is a highly creative layout for an important type of data (ego networks). The image is very complex, and thus rewards multiple views and continued study, this represents a balancing connundrum between a single summary point and detail. Nice work.

  2. admin Says:

    PEER REVIEW COMMENT No. 2 – The image portrays a multidimensional space of immigrant ego networks combining two crucial features: ethnic composition and density. The placement of each ego net on the display is a according to its similarity to the other ego nets surrounding it and its similarity to four prototypical ego net configurations. The use of color to indicate ethnic composition is effective at a distance however the detailed darkness around the circles and edge color denoting density gets lost (perhaps a scale/resolution issue?). Although this image conveys much detailed information, substantively would be nice to think through what is gained over a simpler representation, boiled down to the prototypes and the proportion of ego nets that are clustered with that prototype, for example.

  3. admin Says:

    PEER REVIEW COMMENT No. 3 – This creatively re-imagined scatterplot effectively provides information at multiple levels. At the macro-level, it gives us an intuition about the distribution of ego network structures across typologies, while at the micro-level it provides case-level insight into the structures behind the patterns. I would be interested to see some sort of independent variable mapped into this visualization- what predisposes these migrants toward certain network patterns?


Leave a reply to admin Cancel reply