Correlation has become a fundamental method for information science. However, correlations are limited in making concrete decisions. In this article, we detail how causal inference could be utilised in the field of information science. There are six main steps of implementing matching for causal inference, namely, selecting candidate control variables, determining control variables, calculating similarities among all samples, forming control group, examining the performance of control group and estimating causal effects. As an example, this article applies causal inference to investigate whether Nobel Physics award increases the after-award citations. The method is presented in a step-by-step manner so that researchers can reproduce our analysis in the future.
GodleeFSmithJMarcovitchH.Wakefield’s article linking MMR vaccine and autism was fraudulent. BMJ2011; 342: c7452.
2.
BuYWaltmanLHuangY. A multidimensional perspective on the citation impact of scientific publications. In: Proceedings of the 17th international conference on scientometrics and informetrics (ISSI 2019), Rome, 2–5 September 2019, p. 561–572.
3.
MeadowbrookeCCVeinotTCLoveluckJet al. Information behavior and HIV testing intentions among young men at risk for HIV/AIDS. J Am Soc Inform Ence Tech2014; 65(3): 609–620.
4.
TahamtanIBornmannL.What do citation counts measure? An updated review of studies on citations in scientific documents published between 2006 and 2018. Entometrics2019; 121(3): 1635–1684.
5.
GoodmanS. Commentary: the multifaceted nature of maternal depression as a risk factor for child psychopathology – reflections on Sellers et al. (2014). J Child Psychol Psychiat Allied Discip2014; 55(2): 121–123.
6.
WittenburgP.From persistent identifiers to digital objects to make data science more efficient. Data Intell2019; 1(1): 6–21.
7.
BeggC.Improving the quality of reporting of randomized controlled trials: the CONSORT statement. JAMA1996; 122(8): 637–639.
8.
PetoRPikeMCArmitagePEet al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples. British J Cancer1977; 35(6): 1–39.
9.
WoodAMWhiteIRThompsonSG.Are missing outcome data adequately handled? A review of published randomized controlled trials in major medical journals. Clin Trial2004; 1(4): 368.
10.
AstinJABecknerWSoekenKet al. Psychological interventions for rheumatoid arthritis: a meta-analysis of randomized controlled trials. Arthrit Care Res2002; 47(3): 291–302.
11.
RidnerSLFrostKLajoieA.Health information and risk behaviors among lesbian, gay, and bisexual college students. J Am Acad Nurse Pract2006; 18(8): 374–378.
12.
PearlJ.Causality: models, reasoning, and inference. Cambridge: Cambridge University Press, 2000, p.384.
13.
MorganSLWinshipC.Counterfactuals and causal inference: methods and principles for social research. Cambridge: Cambridge University Press, 2014.
14.
HahnJToddPVan Der KlaauwW.Identification and estimation of treatment effects with a regression-discontinuity design. Econometrica2001; 69(1): 201–209.
15.
LechnerM.The estimation of causal effects by difference-in-difference methods. Found Trend Econo2010; 4(3): 165–224.
16.
SanderG.Relation of probability of causation to relative risk and doubling dose: a methodology error that has become a social problem. Am J Public Health1999; 10: 37–48.
17.
ByarDPSimonRMFriedewaldWTet al. Randomized clinical trials. Perspectives on some recent ideas. N Engl J Med1976; 295(2): 74–80.
18.
FalavarjaniSAMZarrinkalamFJovanovicJet al. The reflection of offline activities on users’ online social behavior: an observational study. Inform Proces Manag2019; 56(6): 102070.
19.
SunJZhangDHuHet al. Predicting human discretion to adjust algorithmic prescription: a large-scale field experiment in warehouse operations. Rochester, NY: SSRN, 2019.
20.
Blumberg JoyceC. Causal inference for statistics, social, and biomedical sciences. Int Stat Rev2016; 84(1): 159.
21.
CardD.Using geographic variation in college proximity to estimate the return to schooling. NBER Work Papers1993; 3386: 4483.
22.
PsacharopoulosG.Returns to education: a further international update and implications. J Human Res1985; 20: 583–604.
23.
WoldH.Causality and econometrics. Econometrica1954; 54: 162–177.
24.
AngristJDPischkeJSPischkeJS.Mostly harmless econometrics: an Empiricist companion. Princeton, NJ: Princeton University Press, 2009.
25.
SanderG. An introduction to instrumental variables for epidemiologists. Int J Epidemiol2000(4): 722–729.
26.
HuangCYueXChenJet al. The effect of ‘open access’ on journal impact factors: a causal analysis of medical journals. Physica A Stat Mech Appl2019; 533: 122043.
27.
WaldingerF.Peer effects in science: evidence from the dismissal of scientists in Nazi Germany. Rev Econ Stud2011; 79(2): 838–861.
28.
DufloEKremerM.Use of randomization in the evaluation of development effectiveness. Conf Eval Develop Effect2005; 7: 205–231.
29.
SoveyAJGreenDP.Instrumental variables estimation in political science: a readers’ guide. Am J Political Ence2011; 55(1): 188–200.
30.
AgrawalAK.The impact of investor protection law on corporate policy and performance: evidence from the blue sky laws. J Financ Econ2013; 107(2): 417–435.
31.
BoundJBakerJRM. Problems with instrumental variables estimation when the correlation between the instruments and the endogenous explanatory variable is weak. J Am Stat Assoc1995; 90(430): 443–450.
32.
StuartEA.Matching methods for causal inference: a review and a look forward. Stat Sci2010; 25(1): 1–21.
33.
RubinDB.Using propensity scores to help design observational studies: application to the tobacco litigation. Health Serv Outcom Res Methodol2001; 2(3–4): 169–188.
34.
FrölichM.Propensity score matching without conditional independence assumption – with an application to the gender wage gap in the United Kingdom. Econometric J2007; 10(2): 359–407.
35.
GitongaZMDe GrooteHKassieMet al. Impact of metal silos on households’ maize storage, storage losses and food security: an application of a propensity score matching. Food Policy2013; 43: 44–55.
36.
LiHGrahamDJMajumdarA.The impacts of speed cameras on road accidents: an application of propensity score matching methods. Accident Anal Prevent2013; 60: 148–157.
37.
AzoulayPGraff ZivinJWangJ.Superstar extinction. Social Ence Electron Publish2010; 125(2): 549–589.
38.
HuangJGatesAJSinatraRet al. Historical comparison of gender inequality in scientific careers across countries and disciplines. Proc Natl Acad Sci USA2019; 117(9): 4609–4616.
39.
LarivièreVGingrasY.The impact factor’s Matthew effect: a natural experiment in bibliometrics. J Assoc Inform Ence Tech2010; 61(2): 424–427.
40.
WaySFMorganACLarremoreDBet al. Productivity, prominence, and the effects of academic environment. Proc Natl Acad Sci2019; 116(22): 10729–10733.
41.
ZhaoZBuYKangLet al. An investigation of the relationship between scientists’ mobility to/from China and their research performance. J Inform2020; 14(2): 101037.
42.
AlshebliBKRahwanTWoonWL.The preeminence of ethnic diversity in scientific collaboration. Nature Commun2018; 9(1): 1–10.
43.
LiWAsteTCaccioliFet al. Early coauthorship with top scientists predicts success in academic careers. Nature Commun2019; 10(1): 5170.
44.
KingGNielsenR.Why propensity scores should not be used for matching. Political Anal2019; 27: 4.
WangBBuYXuY.A quantitative exploration on reasons for citing articles from the perspective of cited authors. Scientometric2018; 116(2): 675–687.
47.
WildgaardLSchneiderJWLarsenB.A review of the characteristics of 108 author-level bibliometric indicators. Scientometrics2014; 101(1): 125–158.
48.
WaltmanL.A review of the literature on citation impact indicators. J Inform2016; 10(2): 365–391.
49.
CochranWGRubinDB.Controlling bias in observational studies: a review. Sankhyā: Indian Journal of Statistics1973; Series A: 417–446.
50.
RubinRDB. The central role of the propensity score in observational studies for causal effects. Biometrika1983; 70(1): 41–55.
51.
OakesJMJohnsonPJ.Propensity score matching for social epidemiology. Method Social Epidemiol2006; 1: 370–393.
52.
ThoemmesFJKimES.A systematic review of propensity score methods in the social sciences. Multivariate Behav Res2011; 46(1): 90–118.
53.
RubinDB.Matched sampling for causal effects. Cambridge: Cambridge University Press, 2006.
54.
BeckerSOIchinoA.Estimation of average treatment effects based on propensity scores. Stata J2002; 2(4): 358–377.
55.
RassenJASchneeweissS.Using high-dimensional propensity scores to automate confounding control in a distributed medical product safety surveillance system. Pharmacoepidem Drug Safe2012; 21: 41–49.
56.
AustinPC.A comparison of 12 algorithms for matching on the propensity score. Stat Med2014; 33(6): 1057–1069.
57.
AustinPC.Some methods of propensity-score matching had superior performance to others: results of an empirical investigation and Monte Carlo simulations. Biom J2009; 51(1): 171–184.
58.
HavilandANaginDSRosenbaumPR.Combining propensity score matching and group-based trajectory analysis in an observational study. Psychol Method2007; 12(3): 247–267.
59.
YanEChenZLiK.Authors’ status and the perceived quality of their work: measuring citation sentiment change in Nobel articles. J Assoc Inform Ence Tech2020; 71(3): 314–324.
60.
BorJMoscoeEMutevedziPet al. Regression discontinuity designs in epidemiology: causal inference without randomized trials. Epidemiology2014; 25(5): 729–737.
61.
SinatraRDevillePSzellMet al. A century of physics. Nature Phys2015; 11(10): 791.
62.
MertonRK.The Matthew effect in science: the reward and communication systems of science are considered. Science1968; 159(3810): 56–63.
63.
JeongHNédaZBarabásiAL.Measuring preferential attachment in evolving networks. Europhys Lett2003; 61(4): 567.
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
