46 Matching Annotations
  1. Apr 2017
    1. Appendix. Technical demonstration of the SOMprocedure

      This is a great example of Kohonen's Self-organizing Maps and the use of the U-Matrix. The authors were very thorough in explaining how it can be used.

    2. The SOM consisted of 360 neurons on a 24by 15 map grid, with hexagonal lattice andGaussian neighborhood function.

      Data Structure

    3. (Kohonen’s Self-Organizing Maps

      This is the novel innovation approach to network analysis that I will be discussing in my article review presentation

    1. analyses using these ego-level measures can be done using statistical packages like SPSS

      I am familiar with SPSS, but I had not considered using it for my final project until now. I am a little confused about how I would have to code the data (adjacency matrix) so that it works well in SPSS, but I think I'm going to try it just to see if I can get the same results that I obtain from using R.

    2. UCINET is NetDraw, a visualization tool that has advanced graphic properties.

      I am interested in learning more about the NetDraw software in UCINET. While I like statistics, nothing really says SNA like having clear visuals that display network ties

    3. visualizing your network in the form of a graph provides some hints as to whether your motivating research questions or hypotheses are worth pursuing or in need of revision.Figure

      I need to more practice with creating cleaner graphs that show my network. So far, I've been sticking with the orange and blue colors but I plan to get a little more creative for the final project

    4. Now that your network's boundary has been specified, you are ready to collect your data.

      My project deviated from the author's process because I actually obtained the dataset before identifying research questions. I am using a public dataset so I had to develop me project around what I had available

    5. First, start with an initial topic of interest and then turn this topic into a question.

      Throughout the semester, I have struggled with turning my research topic into a researchable question. I think I've come up with some decent questions, but I'm constantly rewording them.

    1. This hypothesis can be tested by comparing the average out-degree (number of dichotomized collaboration ties sent) of district- and school-level leaders.

      I have a background in quantitative statistics. The use of a t-test assumes normality of data. I know that normality does not matter for SNA, but is this particular section referring to a nonparametric test that works like a t-test or is it actually referring to the use of a t-test and just ignoring the assumption of normality?

    2. This section shifts the analytical lens to predict an individual actor's outcome, whether it is an attribute variable (e.g., a student's test score) or a structural variable (e.g., a teacher's betweenness centrality score), using relational data. For example: does a teacher's gender predict his or her influence (as measured by degree centrality)?

      Ok... So now this is making a little more sense for my own data. In my case, I could analyze the countries geographic loaction as an attribute varibale that predicts the probability of certain industries having multiple companies in a given country. Does this make sense?

    3. Similarly, the basic idea of the p* is to understand these same relations but to include actor-level and network-level attributes in the model

      I really wish there was an explanation of how to construct/create/execute a p* model using R.

    4. the focal variable is the actor's group membership expressed as an unobserved, latent variable whose value is the result of the observed ties among actors

      Ahhhh... This makes sense. I asked about this in the previous chapter.

    1. The average of the simulated distribution of reciprocal ties is calculated and then compared to the value in the empirical (observed) network

      Can ERGMs be used for nonreciprocal networks?

    2. except, of course, ego-level network studies from which egos have been randomly drawn from some target population)

      I am using ego-level networks for my final project, but how would I go about ensuring that the ego in question was chosen randomly?

    3. artifacts

      Maybe I missed something in our previous readings/videos, but can someone explain to me what is meant by the term "artifacts"?

    4. Even the tools of predictive modeling are commonly applied to network data (e.g. correlation and regression)

      Would running such tests require a need for assessing latent variables that emerge from network analysis? I will keep reading, but from what I know of SNA, it seems like you are only analyzing observable variables and it would be difficult to obtain a correlation from such unique variables. Am I way off here?

    5. factor analysis,

      I would actually be interested in conducting a factor analysis on my data. Though I am a bit confused about how such a test would be useful for SNA

    6. the sample in these instances is the same as the population.

      For my current data set, this is true. However, it would be nice to be able to make generalizations to a larger population.

    7. limiting the types of questions that could be asked,

      I have been dealing with this issue all semester. I am really struggling to determine which questions are appropriate and which can be answered using SNA techniques. Perhaps this is due to the type of data I am working with, but I am still working on developing sound research questions.

    8. focusing on a conceptual understandin

      I am really looking forward to better understanding the concepts of SNA... beyond the explanation of just looking at relationships.

  2. Mar 2017
    1. Egocentric analysis shifts the analytical lens onto a sole ego actor and concentrates on the local pattern of relations in which that ego is embedded as well as the types of resources to which those relations provide access.

      Given the nature of my data (Forbes top companies), I think it would be appropriate to look at specific countries as the Ego and the job categories as the alters. Am I correct in assuming that the local pattern of relations would be how my selected county (the ego) is connected to other countries through job category?

    2. it is possible to examine directed relations in egocentric network studies, or what are referred to as out- and in-neighborhoods: ties sent or ties received

      Since I am working with countries and job categories, I think it would be best for me to work with non directional neighborhoods. I would like to see some directed ego centric data though.

    3. gocentric analysis is primarily concerned with describing how individuals are embedded in local social structures and, ultimately, how these individual indices of social structure relate to varied outcomes

      My research doesn't actually focus on individuals so this is interesting.

    1. Equivalence, in general, refers to actors who occupy the same position.

      Nice definition

    2. for two high school teachers to occupy a structurally equivalent position, both teachers must teach the same set of students

      This reminds me of my middle school. Rather than each class period having different students, we had one classroom with the same students and we would just switch teachers each period. They all taught the same students. It's nice to have a personal example of structural equivalence

    3. equivalence

      I'm interested to learn more about this.

  3. Feb 2017
    1. structural holes (one actor connected to two others who, in turn, are not connected to each other)

      From my understanding of this, structural holes differ from triads since triads generally indicate that all three actors are connected in some way. Is this correct?

    2. effective density,

      I like the explanation of effective density. I think it will be useful for my own "potential" research

    3. UCINET

      This is the software that was used by the authors of the article that I chose for our Week 3 assignment. Apparently, it's a fairly common choice for SNA research

    4. network's topography
    1. If an individual opts out, this should mean that their name appears nowhere on the social network diagram (even if they are identified by another individual as being part of their social network). For instance, in the sample map, you can see that the map would be very disjointed if John and Holly opted out of the SNA.

      Are we allowed to include nodes for John and Holly if they are identified by others, but without using their name? For example, we would refer to John as Anonymous 1 and Holly as Anonymous 2. Or would we have to exclude any data that involves them, regardless of anonymity?

    2. it may be possible to guess the names of individuals by virtue of their location in the network

      While I know this is possible in some situations, I wonder how often it actually occurs.

    3. the purpose of the network analysis may be to identify areas of the firm that just aren’t critical to its mission, vision, and strategy. As social network researchers Steve Borgatti and Jose-Luis Molina note, “This introduces dangers for the respondents because management may make job or personnel changes (e.g., firing non-central workers) based on the network analysis. In fact, in the case of a consulting engagement, this may be the explicit purpose of the research, at least from the point of view of management (Borghatti & Molina, 2005).”

      This seems similar to conducting a program evaluation. I guess the only difference would be the application of network analysis.

    1. snowball sampling

      Interesting. I am unfamiliar with this type of sampling. I have heard of it before, but I have not seen many studies that utilize this sampling approach.

    2. For example, using the public Facebook profiles of a set of students in the same school, it is possible to construct a network of who bullies whom, so long as you can precisely define how bullying is measured.

      I think this is a really good example of SNA research. It would take a lot of thinking to operationally define bullying as identified on facebook, but I bet the results would be interesting. I wonder if something of this nature has already been studied.

    1. Cultural capital theory

      This is my first time learning about cultural capital. It makes sense so I am surprised that I have not come across it before.

    2. human capital theory

      Human Capital is a big deal in the field of Human Resource Development. Supporters of the human capital theory argue that humans are the best asset of any organization. It's nice to see the theory mentioned in literature outside of business and HR fields.

    3. social closure vs. structural holes

      While I can make an educated guess as to what these terms mean, I am not very familiar with them.

  4. Jan 2017
    1. Multiplex data, discussed later in this chapter, are those network data that measure more than one kind of relation, which most contemporary network studies incorporate.
    2. core-periphery structure:


    3. arcs or edges

      Here is a better explanation of the SNA terms Arcs and Edges. Arcs represent those relations that are directed from one student to another, meaning that the friendship nomination has not necessarily been reciprocated. Edges, on the other hand, are those lines that do not have arrowheads (since friendships are directed, there are no edges in Figure 3.1), which are appropriate when the relation is by definition reciprocated (e.g., “studies with”).

    4. arcs.

      I am unfamiliar with the term "arc" in this context since I am unsure what is meant by "directed lines".

    5. Pittinsky asked each student (ego) to rate their friendship on six-point scale (1 = best friend; 2 = friend, 3 = know-like; 4 = know, 5 = know-dislike; 6 = strongly dislike) with every other student (alter) in the class. In addition, the teacher was asked [Page 45]to do the same.

      Applied quantitative SNA methodology

    1. Examining 36 classes in two schools over the course of 1 year, McFarland's analysis, which includes a number of student-level social network measures, reveals that students’ background characteristics only partly influence students’ decision to defy

      Great example of SNA at work.

    2. relational thinking

      I think this will occur a lot in SNA

    3. nductive modeling strategies

      This is a really great SNA term/phrase. Generating big ideas from small observations is a nice description of SNA. It reminds me of the grounded theory approach in qualitative research.