We establish a zero-inflated (random-effects) logistic-Gaussian model for clustered binary data in which members of clusters in one latent class have a zero response with probability one, and members of clusters in a second latent class yield correlated outcomes. Response probabilities in terms of random-effects models are formulated, and maximum marginal likelihood estimation procedures based on Gaussian quadrature are developed. Application to esophageal cancer data in Chinese families is presented.
BonneyGE. Regressive logistic models for familiar disease and other binary traits. Biometrics1986; 42: 611–625.
2.
BonneyGE. Logistic regression for dependent binary observations. Biometrics1987; 43: 951–973.
3.
PrenticeRL. Binary regression using an extended beta-binomial distribution, with discussion of correlation induced by covariate measurement errors. J Am Stat Assoc1986; 81: 321–327.
4.
PrenticeRL. Correlated binary regression with covariates specific to each binary observation. Biometrics1988; 44: 1033–1048.
BowmanDGeorgeEO. A saturated model for analyzing exchangeable binary data: applications to clinical and developmental toxicity studies. J Am Stat Assoc1995; 90: 871–879.
7.
GeorgeEOBowmanD. A full likelihood procedure for analysing exchangeable binary data. Biometrics1995; 51: 512–523.
8.
FitzmauriceGMLairdNM. A likelihood based method for analysing longitudinal binary responses. Biometrika1993; 80: 141–151.
9.
LiangKYZegerSL. Longitudinal data analysis using generalized linear models. Biometrika1986; 73: 13–22.
10.
ZegerSLLiangKY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics1986; 42: 121–130.
11.
PrenticeRLZhaoLP. Estimating equations for parameters in means and covariances of multivariate discrete and continuous responses. Biometrics1991; 47: 825–839.
12.
LiangKLZegerSLQaqishBF. Multivariate regression models for categorical data. J R Stat Soc Ser B1992; 54: 3–40.
13.
AndersonDAAitkinM. Variance component models with binary response: interviewer variability. J R Stat Soc Ser B1985; 47: 203–210.
14.
RosnerB. Multivariate methods for clustered binary data with more than one level of nesting. J Am Stat Assoc1989; 84: 373–380.
15.
GibbonsRHedekerD. Application of random effects probit regression models. J Consult Clin Psychol1994; 62: 285–296.
16.
GibbonsRHedeklerD. Random effects probit and logistic regression models for three-level data. Biometrics1997; 53: 1527–1537.
17.
CnaanALairdNMSlasorP. Using the general linear mixed model to analyse unbalanced repeated measures and longitudinal data. Stat Med1997; 16: 2349–2380.
18.
PinheiroJCBatesDM. Approximations to the log-likelihood function in the nonlinear mixed-effects model. J Comput Gr Stat1995; 4: 12–35.
19.
PinheiroJCBatesDM. Mixed-effects models in S and S-PLUS. Statistics and Computing Series, New York, NY: Springer, 2000. .
20.
Ridout ML, Demetrio CGB and Hinde J. Models for count data with many zeros. In: Proceedings of XlXth International Biometric Society Conference, Cape Town, South Africa, pp.179–192. Washington DC, USA: International Biometric Society, 1998.
21.
BrillingerDRPreislerMK. Maximum likelihood estimation in a latent variable problem. Studies in econometrics, time series and multivariate statistics, New York: Academic Press, 1983, pp. 31–65.
22.
RoseCEMartinSWWannemuehlerKAet al.On the use of zero-inflated and hurdle models for modeling vaccine adverse event count data. J Biopharm Stat2006; 16: 463–481.
23.
HuMCPavlicovaMNunesEV. Zero-inflated and hurdle models of count data with extra zeros: examples from an HIV-risk reduction intervention trial. Am J Drug Alcohol Abuse2011; 37: 367–375.
24.
HallDB. Zero-inflated Poisson and binomial regression with random effects: a case study. Biometrics2000; 56: 1030–1039.
25.
LewseyJDThomsonWM. The utility of the zero-inflated Poisson and zero-inflated negative binomial models: a case study of cross-sectional and longitudinal DMF data examining the effect of socio-economic status. Commun Dentistry Oral Epidemiol2004; 32: 183–189.
26.
AlbertJMWangWNelsonS. Estimating overall exposure effects for zero-inflated regression models with application to dental caries. Stat Methods in Medical Research2011; 23: 257–278.
27.
HurKHedekerDHendersonWet al.Modeling clustered count data with excess zeros in health care outcomes research. Health Serv Outcomes Res Methodol2002; 3: 5–20.
28.
LeeAHWangKScottJAet al.Multi-level zero-inflated Poisson regression modelling of correlated count data with excess zeros. Stat Meth Med Res2006; 15: 47–61.
29.
LoeysTMoekerkeBDe SmetOet al.The analysis of zero-inflated count data: beyond zero-inflated Poisson regression. Br J Math Stat Psy2012; 65: 163–180.
30.
FultonKALiuDHaynieDLet al.Mixed model and estimating equation approaches for zero inflation in clustered binary response data with application to a dating violence study. Ann Appl Stat2015; 9: 275–299.
31.
Kwagyan J. Further investigation of the disposition model for correlated binary outcomes. PhD Thesis, Department of Statistics, Temple University, Philadelphia, PA.
32.
RubinDB. Inference and missing data. Biometrika1976; 63: 581–592.
33.
WilliamsDA. Extra-binomial variation in logistic linear models. Appl Stat1982; 31: 144–148.
34.
MooreDF. Asymptotic properties of moment estimators for overdispersed counts and proportions. Biometrika1986; 73: 583–588.
35.
MooreDF. Modelling the extraneous variance in the presence of extra-binomial variation. Appl Stat1987; 36: 8–14.
36.
CoxDR. The regression of binary sequence. J R Stat Soc B20: 215–242.
37.
CoxDR. The analysis of binary data, London: Chapman & Hall, 1970.
38.
AndersonDAAitkinM. Variance component models with binary response: interviewer variability. J R Stat Soc Ser B (Methodological)1985; 47: 203–210.
39.
GilmourARAndersonRDRaeAL. The analysis of binomial data by a generalized linear mixed model. Biometrika1985; 72: 593–599.
40.
ZegerSLKarimR. Generalized linear models with random effects; a Gibbs sampling approach. J Am Stat Assoc1991; 86: 79–86.
McCullochCE. Maximum likelihood algorithms for generalized linear mixed models. J Am Stat Assoc1997; 92: 162–170.
43.
KuhnELavielleM. Maximum likelihood estimation in nonlinear mixed effects models. Comput Stat Data Anal2005; 49: 1020–1038.
44.
ImSGianolaD. Mixed models for binomial data with an application to lamb mortality. Appl Stat1988; 37: 196–204.
45.
CrouchACSpielgelmanE. The evaluation of integrals of the form application to logistic-normal models. J Am Stat Assoc1990; 85: 464–469.
46.
AbramowitzMStegunIA. Handbook of Mathematical Functions-with Formulas, Graphs, and Mathematical Tables, New York: Dover publications, 1970.
47.
BockRDAitkinM. Marginal maximum likelihood estimation of item parameters: application of an EM algorithm. Psychometrika1981; 46: 443–459.
48.
AkaikeH. Maximum likelihood identification of Gaussian autoregressive moving average models. Biometrika1973; 60: 255–265.
49.
VuongQH. Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica1989; 57: 307–333.
50.
LiJY. Epidemiology of esophageal cancer in China. Natl Cancer Inst Monogr1982; 62: 113–120.
51.
KhurooMSZargarSAMahajanRet al.High incidence of oesophageal and gastric cancer in Kashmir in a population with special personal and dietary habits. Gut1992; 33: 11–15.
52.
RasoolSAGanaiBSameerASet al.Esophageal cancer: associated factors with special reference to the Kashmir Valley. Tumori2012; 98: 191–203.
53.
SelfSGLiangKY. Asymptotic properties of maximum likelihood estimators and likelihood ratio tests under nonstandard conditions. J Am Stat Assoc1987; 82: 605–610.
54.
SnijdersTBoskerR. Multilevel analysis: an introduction to basic and advanced multilevel modeling, Thousand Oaks, CA: Sage, 1999, pp. 90–91.
55.
HedekerDSiddiquiOHUFB. Random-effects regression analysis of correlated grouped-time survival data. Stat Meth Med Res2000; 9: 161–179.
56.
GuptaPGuptaRTripathiR. Analysis of zero-adjusted count data. Comput Stat Data Anal1996; 23: 207–218.
