16 Matching Annotations
  1. Mar 2017
    1. What definition(s) of the neighborhood will make sense for your research projects?

      I would prefer to define the "neighborhood" for my ego-centric analysis to one-step, because I am most interested in who that person is directly connected to, instead of looking at the larger network that the ego is nearly connected to.

    2. How ego-centric networks could be applied to your research projects?

      If I were to apply ego-centric networks to my project, I would focus in on studying members of the game development community that have gone through a public education program to see if they have become more well-connected in the community since going through the program.

    1. Table 7.1

      The data presented in these tables is often kind of confusing, if it is generated from the program they are using.

    2. In addition, most analyses of ego networks use binary data: Two actors are either connected or they are not. This makes the analytical task of defining an ego's “neighborhood” much easier. However, if the relational data between ego and alter is valued—that is, the strength of the tie has been measured—you have to decide the point at which a tie exists.

      This doesn't seem specific to ego-level analyses, since this type of work needs to be done with network-level analyses as well.

    3. Whereas the previous chapter focused on concepts and measures most appropriate for complete network analysis, this chapter shifts the perspective to the analytical level of a sole focal actor—ego.

      Is it common to perform both types of analyses on a dataset?

  2. Feb 2017
    1. Clustering

      I am very interested in clustering measures, because I plan to analyze data from a Slack group that I am a part of, where I suspect there are many subgroups who only interact with each other.

      After looking around for some different clustering algorithms, I found the "cluster_label_prop" function in the igraph package, which seems to do what I would like to do. To summarize, this function automatically detects groups within a network by initially labeling every node with a unique label and at every step each node adopts the label that most of its neighbors currently have. In this iterative process densely connected groups of nodes form a consensus on a unique label to form communities.

      There seem to be many different ways to define clustering though, so I am sure that I will need to do more research on the topic of clustering as I move forward with my research project.

    1. Managing Relational Data

      Your summary of this area was very helpful Bodong! Your visuals were very useful to understand the structure of data necessary for SNA.

    1. Table 3.4

      Is it possible to indicate other types of relationships between actors, other than a 0/1? Or do you just need multiple tables identifying different relationships between the different actors?

    2. Table 3.2

      So in this example, if both actors have a 1 in relation to each other they share an edge (friends), but if one or the other actors does not have a 1, they have a unilateral relationship, and if they both have 0's then there is no relationship between the actors?

    3. Table 3.1

      Very useful visualization of what data looks like.

    1. the tendency to “reduce” individual actors to a collection of attributes removed from context

      The social factors are at least as important as the individual attributes when studying a human phenomenon.

    1. Fig. 3.

      This figure is very helpful for identifying the types of data necessary for SNA

    2. Another key contribution was the influential strength of weak ties (SWT) theory developed by Granovetter

      I find this theory very interesting, especially in careers that are fairly exclusive, such as the video game industry.