越览（151）——精读博士论文的5.3研究设计与变量定义

摘要：This issue of tweets will introduce 5.3 Study design and variable definition of the doctoral thesis "Research on Industrial Chain

分享兴趣，传播快乐，

增长见闻，留下美好。

亲爱的您，这里是LearningYard学苑！

今天小编为大家带来

“越览（151）——精读博士论文

《供应链网络结构视角下的产业链韧性研究》

的5.3 研究设计与变量定义

欢迎您的访问！

Share interests, spread happiness,

increase knowledge, and leave beautiful memories.

Dear you, this is LearningYard Academy!

Today, the editor brings you

5.3 Study design and variable definition

of "Yuelan (151)——Intensive reading of

the doctoral dissertation

‘Research on the resilience of

the supply chain from the perspective of

supply chain network structure’".

Welcome to visit!

一、内容摘要（Summary of content）

本期推文将从思维导图、精读内容、知识补充三个方面介绍博士论文《供应链网络结构视角下的产业链韧性研究》的5.3 研究设计与变量定义。

This issue of tweets will introduce 5.3 Study design and variable definition of the doctoral thesis "Research on Industrial Chain Resilience from the Perspective of Supply Chain Network Structure" from three aspects: mind mapping, intensive reading content, and knowledge supplement.

二、思维导图（Mind mapping）

三、精读内容（Intensive reading content）

（一）数据来源与供应链网络构建（Data sources and supply chain network construction）

本研究以彭博和国泰安为数据源，收集汽车产业供应链网络数据。其中，彭博提供完整且连续10年以上的供应链关系数据，国泰安则为国内最常用且最完整的数据源，包括公司信息、股票、银行关系等。

This study uses Bloomberg and Guotai An as data sources to collect automotive industry supply chain network data. Bloomberg provides complete and continuous supply chain relationship data for more than 10 years, while Guotai An is the most commonly used and most complete data source in China, including company information, stocks, bank relationships, etc.

具体过程如下：

The specific process is as follows:

1. 数据收集：从国泰安中依证监会2021版行业分类确定出的155家汽车生产企業出发，通过彭博获取每家公司的主要供应商和客户信息。这一过程若遇到非上市公司、外国公司或者分公司时，则采取中断、不收录或者用母公司替代。

1. Data collection: Starting from the 155 automobile manufacturers identified by Guotai An according to the 2021 version of the industry classification of the China Securities Regulatory Commission, we obtain the main supplier and customer information of each company through Bloomberg. If this process encounters non-listed companies, foreign companies or branches, it will be interrupted, not included or replaced by the parent company.

2. 数据整理：整理出的供应链关系数据以“公司-供应商（或者客户）”为表进行编码。若公司已在三级供应链中有编码则重复利用，若无则赋予新编码。

2. Data collation: The collated supply chain relationship data is coded in the table of "company-supplier (or customer)". If the company already has a code in the three-level supply chain, it will be reused; if not, a new code will be assigned.

3. 网络建模：将整理出的数据导入Gephi进行网络建模，形成由16024个节点（公司）和39239条边（供应链联系）组成的汽车产业供应链网络，同时为每个节点计算网络结构指标。

3. Network modeling: The sorted data was imported into Gephi for network modeling, forming an automotive industry supply chain network consisting of 16,024 nodes (companies) and 39,239 edges (supply chain connections), and the network structure indicators were calculated for each node.

4. 样本清理：为了与组织韧性数据进行对接，移除三家退市公司，以及2021年才上市且缺少2020年股票数据的10家公司，最终形成142家核心汽车生产企业样本。

4. Sample cleaning: In order to connect with the organizational resilience data, three delisted companies and 10 companies that were listed in 2021 and lacked 2020 stock data were removed, eventually forming a sample of 142 core automobile manufacturers.

（二）模型构建与变量定义（Model construction and variable definition）

为了检验供应链网络结构对企业韧性的作用，本研究从网络结构、企業韧性及控制变量三大方面进行量化衡量。具体来说，本文借鉴史金艳等（2019）、赵珊等（2020）的方法，搭建回归模型进行实证分析。

In order to examine the role of supply chain network structure on enterprise resilience, this study quantitatively measures network structure, enterprise resilience, and control variables. Specifically, this paper draws on the methods of Shi Jinyan et al. (2019) and Zhao Shan et al. (2020) to build a regression model for empirical analysis.

1.被解释变量（Explained variable）

被解释变量为企业韧性，从绩效和运营两个维度进行衡量。其中，股价下跌幅度衡量公司面对外部冲击时的抵抗力，数值越大则表明风险损失越小，韧性越强；存货周转率下跌幅度衡量公司生产运营稳健性，下跌幅度越大则稳健性越差。

The explained variable is corporate resilience, which is measured from two dimensions: performance and operation. The decline in stock price measures the company's resistance to external shocks. The larger the value, the smaller the risk loss and the stronger the resilience. The decline in inventory turnover rate measures the robustness of the company's production and operation. The larger the decline, the worse the robustness.

2. 解释变量（Explanatory variables）

解释变量为供应链网络结构，包括点度中心性（衡量联系数量）、介数中心性（衡量中介作用）、离心度（衡量下游供应链长度）、接近中心性（衡量获取资源时的独立性）、特征向量中心性（衡量网络中邻近主体的重要性）、聚类系数（衡量供应链网络联系的紧密程度），上述网络结构数据均由Gephi软件计算所得。

The explanatory variables are the supply chain network structure, including degree centrality (measures the number of connections), betweenness centrality (measures the intermediary role), eccentricity (measures the length of the downstream supply chain), closeness centrality (measures the independence in obtaining resources), eigenvector centrality (measures the importance of neighboring entities in the network), and clustering coefficient (measures the closeness of supply chain network connections). The above network structure data are all calculated by Gephi software.

3.控制变量（Control variables）

控制变量则包括企业经营年限、盈利能力、财务杠杆、资本强度、运营效率、无形资产、固定资产、第一大股东持股比例、前十大股东持股比例、两权分离率等。此外，为保证模型稳健性，同时加入供应商集中度、客户集中度为控制变量，同时对区域（东部/非东部）和股权性质（国有/非国有）进行异质性分析。

Control variables include the company's operating years, profitability, financial leverage, capital intensity, operating efficiency, intangible assets, fixed assets, the largest shareholder's shareholding ratio, the top ten shareholders' shareholding ratio, the separation rate of the two rights, etc. In addition, to ensure the robustness of the model, supplier concentration and customer concentration are added as control variables, and heterogeneity analysis is conducted on regions (Eastern/non-Eastern) and equity nature (state-owned/non-state-owned).

通过上述模型设置，该研究旨在揭示供应链网络结构对汽车生产企业韧性的作用机理，为供应链风险管理提供数据基础。

Through the above model setting, this study aims to reveal the mechanism of the impact of supply chain network structure on the resilience of automobile manufacturers and provide a data basis for supply chain risk management.

四、知识补充（Knowledge supplement）

在量化实证分析中，解释变量、被解释变量与控制变量是回归模型中三种最为基础且关键的变量。

In quantitative empirical analysis, explanatory variables, explained variables and control variables are the three most basic and key variables in the regression model.

被解释变量（又称因变量），是模型中需要解释或者预测的主体变量，通常用以衡量所关注现象发生的结果或者水平。换句话说，被解释变量表征因果关系中所受到解释变量作用发生的具体变化。

The explained variable (also called the dependent variable) is the main variable that needs to be explained or predicted in the model, and is usually used to measure the result or level of the phenomenon of interest. In other words, the explained variable represents the specific changes in the causal relationship caused by the explanatory variable.

解释变量（又称自变量），是对被解释变量发生作用、导致其发生变异的主要原因或者决定因素。控制变量是为了消除其他无关因素对模型估计结果产生干扰，从而保证解释变量对被解释变量作用效应的稳健性。

The explanatory variable (also known as the independent variable) is the main reason or determining factor that affects the explained variable and causes it to vary. The control variable is to eliminate the interference of other irrelevant factors on the model estimation results, so as to ensure the robustness of the effect of the explanatory variable on the explained variable.

控制变量是为了消除其他无关因素对模型估计结果产生干扰，从而保证解释变量对被解释变量作用效应的稳健性。

The purpose of control variables is to eliminate the interference of other irrelevant factors on the model estimation results, so as to ensure the robustness of the effect of the explanatory variables on the explained variables.

今天的分享就到这里了。

如果您对文章有独特的想法，

欢迎给我们留言，让我们相约明天。

祝您今天过得开心快乐！

That's all for today's sharing.

If you have a unique idea about the article,

please leave us a message,

and let us meet tomorrow.

I wish you a nice day!

文案|yyz

排版|yyz

审核|hzy