周胜

• 1:武汉学院艺术与传媒学院

摘要(Abstract)：

当前传播学的因果推断面临统计原理和检验手段的双重挑战。在因果关系判定上，存在相关性分析替代因果推断的问题。基于观测数据展开的因果关系分析会出现混杂变量带来的内生性问题。本文从传播学量化研究的实际出发，依据观测数据识别因果关系；以框架效应作为研究实例，从因果模型构建、因果关系检验、因果机制解释和因果推断稳健性方面给出了系统的分析流程。研究结果表明：结合先验知识的反事实推理，能够较为有效地识别出观测数据的因果关系；使用协变量匹配，对于排除混杂变量的干扰和解释因果机制具备可行性。同时，本文对于框架效应给出因果关系证据，并检验了推断结果的稳健性。

关键词(KeyWords)： 因果推断;混杂变量;反事实推理;协变量匹配

Abstract:

Keywords:

基金项目(Foundation): 湖北省教育厅科学技术研究计划项目“基于多层感知网络的受众社会认知机制研究”(项目批准号 B2022404)的阶段性研究成果

作者(Author): 周胜

参考文献(References)：

• 胡安宁(2020):《应用统计因果推论》,上海：复旦大学出版社。 加里·金、罗伯特·基欧汉、悉尼·维巴(2014):《社会科学中的研究设计》(陈硕译),上海：格致出版社。(原书King,G,Keohane,R O & Verba S.[1994].Designing social inquiry:Scientific inference in qualitative research.Princeton:Princeton University Press.) 况琨、李廉、耿直、徐雷、张坤、廖备水、黄华新、丁鹏、苗旺、蒋智超(2020):因果推理，《Engineering》,第6卷第3期，107-130页。 郎玫、简楠韬(2019):公共空间盲区：网络舆情治理绩效的因果链条——基于CNA方法的实证研究，《中国社会公共安全研究报告》,第1期，41-58页。 李雪莲、刘德寰(2023):预测与解释：走向因果表征的传播学，《国际新闻界》,第45卷第7期，157-176页。 陆屹洲、马得勇(2020):媒体框架效应及其持续性——以中美经贸摩擦为议题的实验研究，《新闻大学》,第11期，50-65,126页。 马得勇、陆屹洲(2020):复杂舆论议题中的媒体框架效应——以中美贸易争端为案例的实验研究，《国际新闻界》,第42卷第5期，99-120页。 苗旺、刘春辰、耿直(2018):因果推断的统计方法，《中国科学：数学》,第48卷第12期，1753-1778页。 邱嘉平(2020):《因果推断实用计量方法》,上海：上海财经大学出版社。 王刚(2022):基于观察数据的传播学研究设计，《新闻与传播研究》,第29卷第12期，14-31,126页。 王刚、吴星漫(2021):从统计推断到因果推断：传播学定量研究中的内生性问题，《新闻与传播研究》,第28卷第4期，19-37,126页。 吴俊、欧阳书凡、李晓华(2020):基于STM和格兰杰因果分析的网络新闻媒体倾向研究，《系统工程学报》,第35卷第4期，446-458页。 于会会、徐富明、黄宝珍、文桂婵、王岚(2012):框架效应中的个体差异，《心理科学进展》,第20卷第6期，894-901页。 张云亮、冯珺、季芳芳、柳建坤(2020):新媒体接触对社会治理参与的影响研究——基于中国社会状况综合调查2013—2017年数据的实证分析，《新闻与传播研究》,第27卷第7期，77-95,127-128页。 Angrist,J.D.,Imbens,G.W.& Rubin,D.B.(1996).Identification of causal effects using instrumental variables.Journal of the American Statistical Association,91(434),444-455. Basu,D.(1980).Randomization analysis of experimental data:The Fisher randomization test.Journal of the American Statistical Association,75(371),575-582.doi:10.2307/2287648. Card,D.& Krueger,A.B.(1994).Minimum wages and employment:A case study of the fast-food industry in New Jersey and Pennsylvania.The American Economic Review,84(4):772-793. Chan,M.,Hu,P.F.& Mak,M.K.F.(2020).Mediation analysis and warranted inferences in media and communication research:Examining research design in communication journals from 1996 to 2017.Journalism & Mass Communication Quarterly,99(2),463-486.doi:10.1177/1077699020961519. Chong,D.& Druckman,J.N.(2007).A theory of framing and opinion formation in competitive elite environments.Journal of Communication,57(1),99-118.doi:10.1111/j.1460-2466.2006.00331.x. Coppock,A.(2023).Persuasion in parallel:How information changes minds about politics.Chicago:University of Chicago Press. Dos Reis,V.L.& Culotta,A.(2015).Using matched samples to estimate the effects of exercise on mental health via twitter.In Proceedings of the AAAI Conference on Artificial Intelligence (pp.182-188).Austin,Texas USA:AAAI.doi:10.1609/aaai.v29i1.9205. Entman,R.M.(1993).Framing:Toward clarification of a fractured paradigm.Journal of Communication,43(4),51-58.doi:10.1111/j.1460-2466.1993.tb01304.x. Goffman,E.(1974).Frame analysis:An essay on the organization of experience.Cambridge:Harvard University Press. Greenland,S.& Pearl,J.(2011).Adjustments and their consequences—collapsibility analysis using graphical models.International Statistical Review,79(3),401-426.doi:10.1111/j.1751-5823.2011.00158.x. Guyon,I.,Aliferis,C.,Cooper,G.,Elisseeff,A.,Pellet,J.P.,Spirtes,P.& Statnikov,A.(2008).Design and analysis of the causation and prediction challenge.In Proceedings of the Workshop on the Causation and Prediction Challenge (pp.1-33).Hong Kong,China:JMLR. Hahn,J.,Todd,P.& van der Klaauw,W.(2001).Identification and estimation of treatment effects with a regression-discontinuity design.Econometrica,69(1),201-209.doi:10.1111/1468-0262.00183. Heckman,J.J.(2008).Econometric causality.International Statistical Review,76(1),1-27.doi:10.1111/j.1751-5823.2007.00024.x. Imai,K.,Keele,L.& Yamamoto,T.(2010).Identification,inference and sensitivity analysis for causal mediation effects.Statistical Science,25(1),51-71.doi:10.1214/10-STS321.Imbens,G.W.& Rubin,D.B.(2015).Causal inference for statistics,social,and biomedical sciences.Cambridge:Cambridge University Press. Iyengar,S.& Kinder,D.R.(2010).News that matters:Television and American opinion.Chicago:University of Chicago Press. Jou,J.,Shanteau,J.& Harris,R.J.(1996).An information processing view of framing effects:The role of causal schemas in decision making.Memory & Cognition,24(1),1-15.doi:10.3758/BF03197268. Knudsen,E.& Johannesson,M.P.(2019).Beyond the limits of survey experiments:How conjoint designs advance causal inference in political communication research.Political Communication,36(2),259-271.doi:10.1080/10584609.2018.1493009. Morgan,S.L.& Winship,C.(2015).Counterfactuals and causal inference (2nd ed.).Cambridge:Cambridge University Press.doi:10.1017/CBO9781107587991. Pearl,J.(2003).Statistics and causal inference:A review.Test,12(2),281-345.doi:10.1007/BF02595718. Pearl,J.(2009).Causality:Models,reasoning and inference (pp.401-428).New York:Cambridge University Press. Petty,R.E.,Wegener,D.T.& Fabrigar,L.R.(1997).Attitudes and attitude change.Annual Review of Psychology,48,609-647.doi:10.1146/annurev.psych.48.1.609. Potter,W.J.& Riddle,K.(2007).A content analysis of the media effects literature.Journalism & Mass Communication Quarterly,84(1),90-104.doi:10.1177/107769900708400107. Rosenbaum,P.R.(1987).Sensitivity analysis for certain permutation inferences in matched observational studies.Biometrika,74(1),13-26.doi:10.1093/biomet/74.1.13. Rosenbaum,P.R.(2005).Sensitivity analysis in observational studies.Encyclopedia of Statistics in Behavioral Science,4,1809-1814.doi:10.1002/0470013192.bsa606. Saleem,M.,Prot,S.,Anderson,C.A.& Lemieux,A.F.(2017).Exposure to Muslims in media and support for public policies harming Muslims.Communication Research,44(6),841-869.doi:1177/0093650215619214. Slater,M.D.& Gleason,L.S.(2012).Contributing to theory and knowledge in quantitative communication science.Communication Methods and Measures,6(4),215-236.doi:10.1080/19312458.2012.732626. Stone-Romero,E.F.& Rosopa,P.J.(2010).Research design options for testing mediation models and their implications for facets of validity.Journal of Managerial Psychology,25(7),697-712.doi:10.1108/02683941011075256. Stuart,E.A.(2010).Matching methods for causal inference:A review and a look forward.Statistical Science,25(1),1-21.doi:10.1214/09-STS313. Tsur,O.,Calacci,D.& Lazer,D.(2015).A frame of mind:Using statistical models for detection of framing and agenda setting campaigns.In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (pp.1629-1638).Beijing,China:Association for Computational Linguistics.doi:10.3115/v1/P15-1157. Tversky,A.& Kahneman,D.(1981).The framing of decisions and the psychology of choice.Science,211(4481),453-458.doi:10.1126/science.7455683. Valkenburg,P.M.,Peter,J.& Walther,J.B.(2016).Media effects:Theory and research.Annual Review of Psychology,67,315-338.doi:10.1146/annurev-psych-122414-033608. Verma,T.S.& Pearl,J.(2022).Equivalence and synthesis of causal models.In Probabilistic and causal inference:The works of Judea Pearl (pp.221-236).New York,NY,USA:Association for Computing Machinery.doi:10.1145/3501714.3501732. Wysocki,A.C.,Lawson,K.M.& Rhemtulla,M.(2022).Statistical control requires causal justification.Advances in Methods and Practices in Psychological Science,5(2),25152459221095823.doi:10.1177/25152459221095823. Yanovitzky,I.& Bennett,C.(1999).Media attention,institutional response,and health behavior change:The case of drunk driving,1978—1996.Communication Research,26(4),429-453.doi:10.1177/009365099026004004. Yanovitzky,I.& Greene,K.G.(2009).Quantitative methods and causal inference in media effects research.In Nabi,R.& Oliver M.(Eds.),Sage Handbook of Mass Media Effects (pp.35-52).Thousand Oaks:Sage Publications.
• ① 两组调查数据来自中国学术调查数据资料库(http://www.cnsda.org/index.php),由中国人民大学马德勇教授收集并提供。 ② 可忽略性假设：假定观测的协变量在属性的每一层做了随机化分配，这种情况下的平均因果作用是可识别的。 ③ 后门准则(backdoor criterion):因果网络图中，通过d-分离阻断原因变量X和结果Y之间所有的后门路径，从而识别出从X到Y之间的因果关系。 ④ 主成分分析法是通过选取较少的因子，尽可能解释更多的变量元素。保留大于1的特征值(eigenvalue),并进行旋转，将旋转后的载荷系数作为主成分因素。