PropertyValue
?:abstract
  • Count data, as the name suggests, is data resulting from counting. It takes non-negative integer values and its distributions are generally right-skewed. The assumption of normality is not showed. Therefore, classical regression methods cannot be used. It is often encountered that there are more zeros than expected in these data. This greater than expected of zero values is referred to as zero inflation. It is recommended to apply zero inflated regression models in the analysis of zero inflated data. If appropriate methods are not used, results such as obtaining biased parameter estimates and inconsistent results may occur. The most commonly used models are zero inflated Poisson (ZIP) and zero inflated negative binomial (ZINB) regression models. Another method used in the analysis of count data is hurdle models. These models are two-stage models. The first stage is considered the transition stage, and there is a system that shows 0 values versus other positive values. There is a binary representation where zeros are 0 and other positive values are 1. Another stage is the event stage, that is, the stage where there are only positive values. Similar to zero inflated models, if the values in the event phase fit the Poisson distribution, it is called Poisson Hurdle (PH), if it fits the negative binomial distribution, it is called Negative Binomial Hurdle (NBH) regression model. In this study, 2018 data of the Income and Living Conditions Survey (SILC) was used. While the number of residences was taken as the dependent variable, age, gender, education level and income variables were considered as independent variables. In the study, the data was analyzed using ZIP, ZINB, PH and NBH regression models. It was decided which model among the obtained models better represented the data set by using at the Akaike Information Criterion and Log Likelihood values. (xsd:string)
?:author
?:comment
  • (SILC) (xsd:string)
?:dataSource
  • EU-SILC-Bibliography (xsd:string)
?:dateModified
  • 2024 (xsd:gyear)
?:datePublished
  • 2024 (xsd:gyear)
?:editor
?:fromPage
  • 11 (xsd:string)
is ?:hasPart of
?:inLanguage
  • english (xsd:string)
?:isbn
  • 978-625-8368-61-1 ()
?:name
  • A Comparison of Regression Models For Count Data in EU-Silc Survey (xsd:string)
?:publicationType
  • inproceedings (xsd:string)
?:publisher
?:reference
?:sourceCollection
  • Prooceedings Book of IRSYSC2023 (xsd:string)
?:sourceInfo
  • Bibsonomy (xsd:string)
  • In Prooceedings Book of IRSYSC2023, edited by Tuna, Elif and Akoğul, Serkan and Atakul, Serhat, 11-11, Turkish Statistical Instutute, 2024 (xsd:string)
?:startDate
  • 02.11.-05.11.2023 (xsd:gyear)
?:studyGroup
  • European Union Statistics on Income and Living Conditions (EU-SILC) (xsd:string)
?:tags
  • 2024 (xsd:string)
  • FDZ_GML (xsd:string)
  • SILC (xsd:string)
  • SILC_input2024 (xsd:string)
  • SILC_pro (xsd:string)
  • english (xsd:string)
  • inproceedings (xsd:string)
  • transfer24 (xsd:string)
?:toPage
  • 11 (xsd:string)
rdf:type
?:url