<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE ArticleSet PUBLIC "-//NLM//DTD PubMed 2.7//EN" "https://dtd.nlm.nih.gov/ncbi/pubmed/in/PubMed.dtd">
<ArticleSet>
<Article>
<Journal>
				<PublisherName>West Asia Organization for Cancer Prevention (WAOCP), APOCP's West Asia Chapter.</PublisherName>
				<JournalTitle>Asian Pacific Journal of Cancer Prevention</JournalTitle>
				<Issn>1513-7368</Issn>
				<Volume>16</Volume>
				<Issue>11</Issue>
				<PubDate PubStatus="epublish">
					<Year>2015</Year>
					<Month>11</Month>
					<Day>01</Day>
				</PubDate>
			</Journal>
<ArticleTitle>Probability Sampling Method for a Hidden Population Using Respondent-Driven Sampling: Simulation for Cancer Survivors</ArticleTitle>
<VernacularTitle></VernacularTitle>
			<FirstPage>4677</FirstPage>
			<LastPage>4683</LastPage>
			<ELocationID EIdType="pii">31144</ELocationID>
			
			
			<Language>EN</Language>
<AuthorList>
</AuthorList>
				<PublicationType>Journal Article</PublicationType>
			<History>
				<PubDate PubStatus="received">
					<Year>1970</Year>
					<Month>01</Month>
					<Day>01</Day>
				</PubDate>
			</History>
		<Abstract>When there is no sampling frame within a certain group or the group is concerned that making its populationpublic would bring social stigma, we say the population is hidden. It is difficult to approach this kind ofpopulation survey-methodologically because the response rate is low and its members are not quite honest withtheir responses when probability sampling is used. The only alternative known to address the problems causedby previous methods such as snowball sampling is respondent-driven sampling (RDS), which was developed byHeckathorn and his colleagues. RDS is based on a Markov chain, and uses the social network information ofthe respondent. This characteristic allows for probability sampling when we survey a hidden population. Weverified through computer simulation whether RDS can be used on a hidden population of cancer survivors.According to the simulation results of this thesis, the chain-referral sampling of RDS tends to minimize as thesample gets bigger, and it becomes stabilized as the wave progresses. Therefore, it shows that the final sampleinformation can be completely independent from the initial seeds if a certain level of sample size is secured evenif the initial seeds were selected through convenient sampling. Thus, RDS can be considered as an alternativewhich can improve upon both key informant sampling and ethnographic surveys, and it needs to be utilized forvarious cases domestically as well.</Abstract>
		<ObjectList>
			<Object Type="keyword">
			<Param Name="value">Respondent-driven sampling</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">cancer survivors</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">social network</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">hidden population</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">Simulation</Param>
			</Object>
		</ObjectList>
<ArchiveCopySource DocType="pdf">https://journal.waocp.org/article_31144_78c84df7a661b5bdf8095c10a8511fa2.pdf</ArchiveCopySource>
</Article>
</ArticleSet>
