50 Matching Annotations
  1. Nov 2017
    1. The degree distribution of the giant component seems tofollow a power-law distribution

      Yes, by eyeball test it looks like a power law - wonder why they did not fit the distribution with the power law equation?

    2. Group co-membership: two bloggers belong to the samegroup (blogring). This is an undirected relationship withan integer weight (based on the number of groupsshared by the two bloggers

      This sounds like a two-mode or affiliation network.

    3. assuming the photos are reallythose of the bloggers themselves

      A big assumption.

    4. It is shown that a random networkusually has a small average path length and is moreefficient because an arbitrary node can reach any othernode in a few steps. Small-world networks usually havesignificantly high clustering coefficients than their randomnetwork counterparts. Scale-free networks are categorizedby a power-law degree distribution, which is different fromthe Poisson degree distribution presented in randomnetworks

      Given the time, I would like to read a review of each of these types of networks. Sort of like the Wasserman book, but focused on examples first.

    5. Recent years have seen theemergence of hate groups in blogs, where high-narrativemessages are the norm. This has made blogs an idealmedium for spreading hatred. Blogs have also made itpossible for individuals to find others with similar beliefand ideology much more easily. As a result, hate groupshave emerged in blogs

      The Luddite in me points out that the advent of radio (and newspaper before that) also led to spreading messages of hate. For example Nazi control of radio in the years leading to WWII was critical to getting the public to embrace the Nazi policies.

    6. To study these online hate groups in blogs

      Does anyone else find these studies depressing?

    7. Because of such interactions among bloggers,these communities are less similar to thecyber communitiesas discussed inKumar et al. (1999)but more resembling tothevirtual communities


  2. Oct 2017
    1. Especially in societies likeTurkey where the state is dominant in the business life,organizations and managers prefer to be included in reli-gious networks to make close contacts with the state.Another significant finding is that efforts of the members ofreligious networks—in spite of their relatively closedcharacteristics—in terms of being at the center of anetwork and taking the brokerage role, are highly devel-oped on the contrary to the literatur

      Does anyone else find this difficult to understand? I am not sure if these statements are the authors interpretation of their results or if this is their hypothesis. Also, what are the authors saying about the place/role of members of religious networks?

    1. 6K-core which includes 11 participants connected to at least six other participants

      From Wasserman and Faust: K-cores are used to identify sub-groups (or cliques?) within a network. The clique is a complete subgraph of a whole network. I think this means that all the nodes in this small group have all possible ties. So (again this is my understanding), a K-core of 6 means that all the nodes are linked to (at least) the same 6 nodes. Nodes in one clique can be in other cliques.

    2. Two coders collectivelycoded the 3,319 tweets. Inter-coder reliability was tested based on a subset of 250 tweets(8 % of the sample), and expressed as percentage of agreement (Lombard et al. 2002).

      This is am impressive amount of work.

    3. We call our readers’ attention that the study is not designed to incorporate all LINKScomponents. The study is bounded by the nature of Twitter data and healthcare knowledge

      Network boundaries.

    1. We posit that strate-gies to increase the number of alters who do not engage in substanceuse may be an effective intervention strategy

      This is clearly the goal of VCU's The Well and their "publication", the Stallsheet.

    2. Multiple times a week

      30% use alcohol multiple times a week. Very different from the results posted on VCU's stallsheets!

    3. 281 undergraduates (189 females) enrolled in lower-level psycholo-gy courses at a large university participated in the study.

      The population of students is undergraduate students at a somewhat selective university.

  3. Sep 2017
    1. tacitknowledge,metis,situatedpractices,orskill

      similar to social capital, but not quite the same?

    2. new knowledge did not necessarily dis-place local ecological knowledge

      Could this knowledge be represented as a network? Crops are nodes, people who know about the crops are links. Or, crops are nodes, similarities and/or differences between crops link them - these edges could be positive or negative or have some other attribute.

    3. The bivariate screening for the reduced sample (N=32)found that three socioeconomic variables,‘education,’‘occupation,’and‘money spent on processed food’showasignificantrelationshiptoIEK-MSagreementatthe90 % confidence level (Table3)

      Rather than go with a lower confidence limit, I would think a higher confidence limit is warrented when looking at the combination of three variables.

    4. 90 % confidence

      Yikes, why not 95%? Is this standard?

    5. The resulting mea-sures served as four additional independent variables and wereadded to the variables used in the analysis of the larger sub-sample and evaluated employing a method similar to that not-ed above

      The authors appear to use network analysis as a descriptive tool to give a clearer picture to their linear regression model.

    6. ocial network analysis to ascertainwhether the perpetuation and transmission of knowledge isshaped by positions within social and expert networks

      Examine the role of network structure on individual attributes.

    7. raft of recent research has focused on the quan-titative measurement of individual variation in knowledge(Atran and Medin2008; Reyes-Garcíaet al.2007).

      This seems critical to understanding how students learn and teachers should teach! It also related to widespread misunderstanding and misconceptions around, for example, climate change, vaccinations, GMO foods, ....

    8. nowledge was understood as a norma-tive, socially transmitted trait shared across particular culturesor societies, a perspective that has a long intellectual pedigreein the social sciences

      a theory - possible untested?

    9. most regions of the world nowexperiencing rapid social and ecological change

      Not only third world - these ideas apply to, for example, West Virginia and the Mountain West of the U.S. (mining).

    1. especially for female athletes.

      Actually, I would conclude that the result of this paper (that efficacy and trust matter more than friendship) makes sense for females just as it does for males. To me the results of this paper contradict the conclusions described for the two earlier papers cited.

    2. Example of Longitudinal In-Degree Centrality for Efficacy Network (Team A)

      I think these data could be presented very effectively by, for example, looking at average values of the in-degree centrality, and then looking at each individual's deviation from that average. This would then move from the whole network measure to the individual's attribute.

    3. Without the multifaceted approach afforded through the use of this new analytic tool, this level of revelatory insight would not be possible.

      These conclusions point toward the idea that it is the nature of the network structure that matters - by placing the right individuals or organizations at the right place in the network, more "stuff" happens.

    4. Cohesion Measures

      Density of ties?

    5. eam B)

      Team A has many more connections than Team B.

    6. I consider this a person a friend), trust (i.e., I trust this person

      Presumably each player, coach, and support staff was asked to rate every other player, coach, and support staff.

    7. 58Figure 1 — Example of Efficacy Network, Off-Season and Post-Season (Team

      Use of color would help the reader to interpret these more effectively. Also, I believe the nodes could be sized by centrality measures.

    8. Although there is not a particular criterion for analysis in the inspection process, conclusions are drawn through cross-referencing the graphical output with both the generated indices and the researcher’s intuition

      Hmm. I wonder if a systematic approach to interpreting the visual result would be more valid - that is, having a specific question in mind.

    9. dyadic

      Other papers have discussed the importance of triads.

    10. Due to the nature of SNA research, one group of individuals, or in this case one team, would be con-sidered a sufficient sample to conduct an SNA inquiry (Kilduff & Tsai, 2007).

      Need to look at this paper to see how sample size matters and is factored into the analysis.

    11. participants (n = 43) w

      A fairly small sample size. N of two for the teams, N of 4 for males, N or 7 for coaches. It will be interesting to see how sample size affects interpretation of the network.

    12. To help define, categorize, and measure levels of cohesion, the model developed by Carron et al. divides cohesion into four dimensions: group integration-task (GI-T), group integration-social (GI-S), individual attractions to the group-task (ATG-T), and individual attractions to the group-social (ATG-S). Cohesion within this framework has been evaluated based on the widely-used Group Environment Questionnaire (GEQ), an 18-item self-report inventory anchored on a 9-point Likert-type scale.

      I use a team-based learning approach in my classes, so this framework is very intriguing. Combining this framework with SNA in a team-based classroom vs. a traditional lecture-based classroom may uncover the source of difference in how students learn in each type of class.

    1. state agencies (such as the Ministry of Education and Culture that provideincentives for universities to collaborate with private sector);

      Here is another possible topic for study: what is the funding source for research? Who are the funders, and who gets funded? The nodes could be the agencies doing the funding and the people requesting funding (submitting proposals).

    2. . The latter is ageographically and institutionally uneven process that is based on practicesand networks that integrate higher education institutions and market actorsacross nation-state borders (

      Here the authors actually attribute the expansion of academic capitalism to "networks". I wonder what is meant by network here. What/who would be the nodes, and what would link the nodes? And how would the nodes and links be "measured"?

    3. intermediating organisations (such as SHOKs);corporations (as components within multi-scalar R&D networks withuniversities or donors);state agencies (such as the Ministry of Education and Culture that provideincentives for universities to collaborate with private sector);international organisations (such as WTO, World Bank and OECD thatproduce discourses and narratives that emphasise the importance of theintegration of higher education and markets for national innovationsystems); andsupranational organisations (such as the EU that tends to subordinatehigher education policy to economic and innovation policy through variousstrategies such as the 2000 Lisbon strategy)

      are there people in the universities who bring these ideas to their campuses? Who are they and what are their roles and attributes?

    4. actors who have contributed to ashift towards, and/or reproduced, academic capitalism

      Would these be the nodes in a social network?

    5. trends in the university–industry linkages (Kauppinen, 2012), treatinghigher education as a commodity (Naidoo, 2003) and the establishment oftechnology transfer offices (Kauppinen, 2013), as well as various legislativechanges, show that higher education institutions also in many other countrieshave become more and more directly involved with those networks,organisations, practices and circuits of knowledge that characterize academiccapitalism

      How do these practices spread? Individual action or social network or both

    6. e establishment of technology transfer offices,legislative changes, US research universities’ intellectual property rightspolicies, increased managerial capacity, board interlocks between universitiesand corporations and increased tuition fees

      items that have linked universities to the market; are there people in the universities to share the knowledge or who have the managerial capacity?

    7. new circuits of knowledge linking universities and private companies

      nodes = universities, private companies knowledge circuits - the relations

    8. material and symbolic profits

      attributes of universities

    9. universities and colleges have nowadays integrated moretightly with the new economy

      how is the integration of the academy with the marketplace measured? Integration implies relationships of some sort - connections.

    10. institutional and organisational transformations

      social network

    11. ifferent organisational levels and between different socialsphere

      organization implies social network

    12. different disciplines

      Nodes - different disciplines? or faculty in different disciplines?

    13. Academic Capitalism

      I am learning social network analysis in part to help describe and understand the "corporatization" of the academy in the U.S. I chose this article as it is concerned with the spread of "academic capitalism" in a smaller system. I hope that the article helps define terms and gives insight into how to decide the appropriate nodes and relationships.

    14. activities and practices related to academic capitalism remain, how-ever, unevenly distributed among different discipline

      How does the practice of academic capitalism spread through the university system? Is the spread related to a "network"? If so, how do we define and describe that network?

    15. integrated Finnish universities

      How are universities related?