社会网络分析: 理解网络

政务大数据应用与分析 (80700673)

胡悦

清华大学社会科学学院

概要

理解网络 (本次课)

目的:描述网络结构,识别网络特征

  • 网络方法论
  • 网络要素
  • 自我中心网络

分析网络 (下次课)

目的:揭示网络构成因素和原因

  • 邻居分析
  • 扩散分析
  • 全网分析

1 社会网络·一种方法论

1.1 方法论 vs. 方法

你学过什么方法
你的方法论是什么?
你持有什么样的认识论

Source: Generated by FLUX

Source: Generated by FLUX

“定性” vs. “定量”

Please don’t let me start…

1.2 期望

Although the methods today could be a little complicated
you’ll get there! But probably not today…

Kick-in software

UCNet

UCNet

Good starting point

Prell (2011)

Prell (2011)

1.3 方法论的迁越

The foundation of econometric methods

For a sequence of random variables \(X_1, X_2, ···, X_n\),

\[ \begin{aligned} P(X_i \cap X_j) =& P(X_i) \cdot P(X_j), \\ F_{X_i}(x) =& F_{X_j}(x), \quad \forall i, j. \end{aligned} \]

For the last thirty years, empirical social research has been dominated by the sample survey. But as usually practiced…the survey is a sociological meat grinder, tearing the individual from his social context and guaranteeing that nobody in the study interacts with anyone else in it (Barton 1968, 1)

Interdependence → Relationships → Structure

1.4 跃迁带来的认知改变

关系的特殊性 (Parkinson 2013)

  • 相似性(Similarity)
  • 关联性(Social relations)
  • 互动性(Interactions)
  • 流动性(Flows)

关系的测量方式 (Mucha et al. 2010)

  • 距离(Distance)
  • 通道(Degree)
  • 聚类(Modularity)
  • 预测(Relation prediction)

Y

  • 同质性 → 关联的紧密程度
  • 特质性 → 谁是核心
  • 组织结构 → 系统密度

X

  • 关系 → 行为
  • 位置 → 推广
  • 距离 → 效率

1.5 小结

  • 网络分析代表另一种方法论视角
    • Crux: 打破IID幻象,模拟更真实人类社会
  • 认知进步
    • 关系居于研究核心
      • 关系的属性
      • 关系的实现
    • 结果(Y)的变化
    • 干预(X)的变化

2 网络要素

2.1 网络组成要素

Node/Actor/Vertex
Edge/Relation/Link/Tie

Subgraphs: Dyads, triads, k-ads, isolate/island

2.2 Walks/Paths

Seven Bridges of Königsberg

Seven Bridges of Königsberg

2.3 网络分型

  • “全网”数据 (Complete network)
    • 二模网络(Two-mode/bipartite)
  • 样本网数据 (“Sample” network)
    • 自我核心网络(Ego-net)
    • 滚雪球网络(Respondent-drive sampled data)
    • Somewhere in between
  • 随机网络 (Random graph)
    • 代表性?
    • The large, the better?

2.3.1 全网数据

2.3.2 样本网数据

2.3.3 多模网络I

2.3.4 多模网络II

2.4 给网络抽样

2.4.1 方案I: 随机抽样

2.4.2 方案II:自我中心网络样本

2.4.3 方案III:痕迹网络样本

2.4.4 方案IV:边际网络样本

2.4.5 方案V:滚雪球样本

2.5 记录网络

  • 有向(directed)vs. 无向(Undirected)
  • 有无(binary)vs. 多少(valued)
  • 单线(single)vs. 多线(multiplex)
  • DGP是否会产生影响?
  • Missing Data

属性

  • Nodal attribute data
  • Edge weights

2.6 网络数据

Adjacency Matrix

Adjacency Matrix

Incidence Matrix

Incidence Matrix

2.7 Incidence to Adjacency

2.8 Adjacency to Edge List

2.9 小结

  • 网络要素
    • Node + edge
    • Subgraphs
  • 网络分型
    • Complete ~
    • Sampled ~
    • Multi-mode ~
  • 网络抽样
    • Random(👎)
    • Ego
    • Trace
    • Boundary
    • Snow-ball
  • 网络数据
    • Incidence
    • Adjacency
    • Edge

3 自我中心网络

3.1 数据生成

  • 全网抽样
  • Name generators

3.2 关注焦点

  • Ego-alter connections
    • 方向、强度
    • 关联频率
  • Alter-alter connections
    • 他他关联否?
    • 强度、频率?
  • Alter characteristics

3.3 网络测量

规模

  • 量级(Size)
  • 密度(Density)

相似性

  • 同质性(Homophily)

(未)连结性

  • 结构洞(Structural Holes)
  • 掮客(Brokerage)

聚合性

  • 中心性(Centrality)
  • 小团体(Subgroups)

3.4 量级

Size = Count(Alters).

e.g., nHY = 6.

3.5 密度

密度公式: \(d_i = \frac{L_{-i}}{n(n-1)/2}\)

L-i: Alters间的实际连结数;
n: 量级(# Alters)

\(d_{YH} = \frac{5}{6(6 - 1)/2} \approx 0.33.\)

如果网络是有向的,怎么办?

3.5.1 应用实例

🌰 Security Egonets (Maoz 2010)

  • 目标:在国际网络中,国家何时动武
  • 假设:国家的Strategic Reference Group (SRG)越大,其越有可能在外交政策中施用争端性政策工具(如介入MID或战争)
  • 数据源:新独立国家独立五年内和已独立国家发生MID五年内的国际环境。
  • 测量: SRG,以国家为Ego,测量与其产生敌对关系、冲突的国家及国家联盟

🌰 Power of weak ties (Roch, Scholz, and McGraw 2000)

  • 目标: SNA对纳税人税收政策态度影响
  • 理论:动机和能力 → weak ties (同事、熟人等)
  • 假设:人们从weak ties比从strong ties获取更多对税务的知识 → 与strong ties 一起决定对照章纳税的态度

3.6 行为体间异同

Moody (2001)

Moody (2001)

3.6.1 为何关注同质性

Important

Similarity breeds connections.

什么创造了同质性?

  • 人口学因素
  • 状态
  • 价值
  • 行为

3.6.2 测量同质性

Same Different
Tie (a) 5 (b) 6
No tie (c) 4 (d) 6

Homophily:

  • Yule’s Q = (ad - bc)/(ad + bc)s
  • EI = (b - a)/(b + a)

Heterogeneity

  • Blau’s H: H = 1 - ∑pi2
  • Index of Qualitative Variation (IQV): H/(1-1/r)1

🌰 某甲有九亲、四友、十二同僚,其路野乎?

Alters = 9 + 4 + 12 = 25

p = 9/25;
p = 4/25;
p = 12/25.

H = 1 - [(4/25)2 + (12/25)2 + (9/25)2] = 0.6144

IQV = 0.6144/(1 - 1/3) = 0.9216

3.6.3 应用实例

🌰 Smith, McPherson, and Smith-Lovin (2014, fig. 1)

🌰 Smith, McPherson, and Smith-Lovin (2014, fig. 1)
  • 目标:美国民众核心圈子(core discussion networks)在人口学维度上的同质性
  • 数据:GSS 1985–2004
  • 发现:年龄同质性总体保持稳定,但年轻人变得更加孤立 (性别同质性下降,教育同质性上升,种族和宗教同质性保持稳定)

思考:

  1. 同质就是关联吗?关联能代表同质吗?
  2. 同质性关联的时间维度

3.7 网络测量(Have a break)

规模

  • 量级(Size)
  • 密度(Density)

相似性

  • 同质性(Homophily)

(未)连结性

  • 结构洞(Structural Holes)
  • 掮客(Brokerage)

聚合性

  • 中心性(Centrality)
  • 小团体(Subgroups)

3.8 结构洞 (Structural Holes, Burt 2009)

  1. Ego的冗余性: Effective Sizebinary = TiesEgo - TiesAlters
  2. Alters对于Ego的限制: Constraintij = (pij + ∑qpiqpjq)2, q ≠ i,j
  3. 不考虑Ego情况下Alters的连接情况, betweenness: \[C_B(i) = \sum_{i\neq j\neq k}(\frac{\sigma_{kij}}{\sigma{kj}})\]

3.9 掮客 (Brokerage)

Coordinator, consultant, representative, gatekeeper, liaison

Coordinator, consultant, representative, gatekeeper, liaison

Liu (2021)

Liu (2021)

Chinese Migrants in Central-Eastern Europe

Bonding/bridging social network

Political incorporation

Tip

Survey-based non-random-network analysis

3.10 聚合性

  • 小团体:局部聚合性
  • 中心性:总体聚合性

  • Degree*^1
  • K-step reach*^
  • Eigenvector*^
  • Alpha/Katz*^
  • Power*^
  • Beta*^
  • PageRank*^
  • Hubs and Authorities*^
  • Closeness*^
  • Betweenness*
  • Flow Betweenness*
  • Random Walk Betweenness*
    ……

3.11 基础测量

Degree ~

\[ C_D(i) = \sum_{j=1}^{n}n_{ji}, j \neq i. \]

  • 关注点:local network, ego connect
  • 目标:Activity,与别人的连接度

Eigenvector ~1

\[ C_E(i) = k^{-1}\sum_{j=1}^{n}A_{ji}C_E(j). \]

  • 关注点:近邻关系
    • 可以看作degree centrality 改进版

3.12 Degree vs. Eigenvector

3.13 专门测量

Betweenness

\[ C_B(i) = \sum_{j,k}\frac{\delta_{jik}}{\delta_{jk}}, i\neq j\neq k. \]

目标:Potential controls

Closeness

目标:Independence, 与每个结点的关系

3.13.1 应用实例

3.14 如何挑选

3.15 小结

Ego-network

  • 生成
    • Ego
    • Path
    • Boundary
    • Snowball
  • 描述
    • Size
    • Density
    • Similarity
    • Centrality

下一节

分析网络

参考文献

Barton, Allen H. 1968. “Survey Research and Macro-Methodology.” American Behavioral Scientist 12 (2): 1–9. https://doi.org/10.1177/000276426801200201.
Box-Steffensmeier, Janet M., Dino P. Christenson, and Matthew P. Hitt. 2013. “Quality over Quantity: Amici Influence and Judicial Decision Making.” American Political Science Review 107 (03): 446–60.
Burt, Ronald S. 2009. Structural Holes: The Social Structure of Competition. Harvard University Press.
Klein, Katherine J., Beng-Chong Lim, Jessica L. Saltz, and David M. Mayer. 2004. “How Do They Get There? An Examination of the Antecedents of Centrality in Team Networks.” Academy of Management Journal 47 (6): 952–63. https://doi.org/10.5465/20159634.
Liu, Amy. 2021. The Language of Political Incorporation: Chinese Migrants in Europe. 1st edition. Philadelphia: Temple University Press.
Maoz, Zeev. 2010. Networks of Nations: The Evolution, Structure, and Impact of International Networks, 1816–2001. Cambridge University Press.
Moody, James. 2001. “Race, School Integration, and Friendship Segregation in America.” American Journal of Sociology 107 (3): 679–716. https://doi.org/10.1086/338954.
Mucha, Peter J., Thomas Richardson, Kevin Macon, Mason A. Porter, and Jukka-Pekka Onnela. 2010. “Community Structure in Time-Dependent, Multiscale, and Multiplex Networks.” Science 328 (5980): 876–78. https://doi.org/10.1126/science.1184819.
Parkinson, Sarah Elizabeth. 2013. “Organizing Rebellion: Rethinking High-Risk Mobilization and Social Networks in War.” American Political Science Review 107 (3): 418–32. https://www.jstor.org/stable/43654915.
Prell, Christina. 2011. Social Network Analysis: History, Theory and Methodology. Sage.
Roch, Christine H., John T. Scholz, and Kathleen M. McGraw. 2000. “Social Networks and Citizen Response to Legal Change.” American Journal of Political Science 44 (4): 777–91. https://doi.org/10.2307/2669281.
Rossman, Gabriel, Nicole Esparza, and Phillip Bonacich. 2010. “I’d Like to Thank the Academy, Team Spillovers, and Network Centrality.” American Sociological Review 75 (1): 31–51. https://doi.org/10.1177/0003122409359164.
Smith, Jeffrey A., Miller McPherson, and Lynn Smith-Lovin. 2014. “Social Distance in the United States: Sex, Race, Religion, Age, and Education Homophily Among Confidants, 1985 to 2004.” American Sociological Review 79 (3): 432–56. https://doi.org/10.1177/0003122414531776.
Wellman, Barry, Renita Yuk-lin Wong, David Tindall, and Nancy Nazer. 1997. “A Decade of Network Change: Turnover, Persistence and Stability in Personal Communities.” Social Networks 19 (1): 27–50. https://doi.org/10.1016/S0378-8733(96)00289-4.
胡悦, and 欧阳睿. 2023. “软实力与硬实力关系研究——基于网络分析的实证研究.” 世界经济与政治, no. 7: 27-50+156-157.