Modelling fertility:an application of count regression models

    在线阅读 下载PDF 导出详情
    摘要 Oftenthelifecycledataoccurascountofthevitaleventsandarerecordedasintegers.Thepurposeofthisarticleistomodelthefertilitybehaviorbasedonreligious,educational,economic,andoccupationalcharacteristics.TheresponsesofclassifiedgroupsaccordingtothesedeterminantsareexaminedforsignificantinfluenceonfertilityusingPoissonregressionmodel(PRM)basedontheNationalFamilyHealthSurvey-3dataset.TheobservedandpredictedprobabilitiesunderPRMindicatemodalvalueoftwochildrenforthePoissondistributionmodeleddata.Presenceofdominanceoftwochildinthedatamotivatestheauthorstoadoptmultinomialregressionmodel(MRM)inordertolinkfertilitywithvarioussocioeconomicindicatorsresponsibleforfertilityvariation.Choiceoftheexplanatoryfactorsislimitedtotheavailabilityofdata.Trendsandpatternsofpreferenceforbirthcountssuggestthatreligion,caste,wealth,femaleeducation,andoccupationarethedominantfactorsshapingtheobservedbirthprocess.Empiricalanalysissuggeststhatboththemodelsusedinthestudyperformsimilarlyonthesampledata.However,fittingofMRMbytakingbirthcountoftwoascomparisoncategoryshowsimprovedAkaikeinformationcriterionandconsistentAkaikeinformationcriterionvalues.CurrentworkcontributestotheexistingliteratureasitattemptstoprovidemoreinsightintothedeterminantsofIndianfertilityusingPoissonandMRM.
    机构地区 不详
    出版日期 2015年04月14日(中国Betway体育网页登陆平台首次上网日期,不代表论文的发表时间)
    • 相关文献
    Baidu
    map