• The Korean Journal of Applied Statistics
• /
• v.23 no.3
• /
• pp.533-543
• /
• 2010

Asking sensitive questions by a direct survey method causes non-response bias and response bias. Non-response bias arises from interviewees refusal to respond and response bias arises from giving incorrect responses. To rectify these biases, Warner (1965) introduced a randomized response model which is an alternative survey method for socially undesirable or incriminating behavior questions. The randomized response model is a procedure for collecting the information on sensitive characteristics without exposing the identity of the respondent. Many survey researchers have proposed diverse variants of the Warner randomized response model and applied their model to collect the information of sensitive questions. Using an optimal allocation, we proposed three-stage stratified randomized response technique which is an extension of the Kim and Elam (2005) two-stage stratified randomized response technique. In this study, we showed that the estimator based on the proposed response model is more efficient than Kim and Elam (2005). But by adding one more survey step to the Kim and Elam (2005), our proposed model may have relatively less privacy protection compared to the Kim and Elam (2005) model.

• Communications for Statistical Applications and Methods
• /
• v.4 no.2
• /
• pp.411-419
• /
• 1997

본 연구에서는 확률화응답모형이 가지는 두가지 한계에 대하여 고찰하였다. 첫째 민감한 속성을 갖는 모비율의 추정시 모비율 $\pi_A$가 매우 작은 값일 경우, 즉 희귀속성일 경우 확률화응답모형을 적용하게 되면 비밀보장의 효과를 감안한다고 해도 직접질문법에 비해 비효율적일 수 있음을 지적하였다. 둘째로 비밀보장에서 오는 이점과 그로 인한 효율의 손실이라는 서로 상충되는 면을 객관적으로 고려하는데 있어서 한계가 있음을 지적하였다.

• The Korean Journal of Applied Statistics
• /
• v.29 no.5
• /
• pp.897-906
• /
• 2016

We augment an unrelated quantitative attribute to Bar-Lev et al.'s model (2004) which is composed of sensitive quantitative variable and scrambled one to present a multiplicative unrelated quantitative randomized response model(MUQ RRM). We also establish theoretical grounds to estimate the sensitive quantitative attribute according to circumstances irrespective of known or unknown unrelated quantitative attribute. Finally, we explore the relationship among the suggested model, Eichhorn-Hayre model, Bar-Lev et al.'s model and Gjestvang-Singh's model, and compare the efficiency of our model with Bar-Lev et al.'s model.

• Journal of the Korean Data Analysis Society
• /
• v.20 no.6
• /
• pp.2895-2905
• /
• 2018

We present a mixed multiplicative quantitative randomized response model which added a unrelated quantitative attribute and forced answer to the multiplicative model suggested by Bar-Lev et al. (2004). We also try to set up theoretical grounds for estimating sensitive quantitative attribute according to circumstances whether or not the information for unrelated quantitative attribute is known. We also extend it into the stratified mixed multiplicative quantitative randomized response model for stratified population along with two allocation methods, proportional and optimum allocation. We can see that the various quantitative randomized response models such as Eichhorn-Hayre's model (1983), Bar-Lev et al.'s model (2004), Gjestvang-Singh's model (2007) and Lee's model (2016a), are one of the special occasions of the suggested model. Finally, We compare the efficiency of our suggested model with Bar-Lev et al.'s (2004) and see that the bigger the value of $C_z$, the more the efficiency of the suggested model is obtained.

• The Korean Journal of Applied Statistics
• /
• v.12 no.1
• /
• pp.143-152
• /
• 1999

본 논문에서는 응답자가 민감한 속성을 가지고 있지 않으면 직접 "예"라고 응답하고, 민감한 속성을 가지고 있으면 Greenberg et al.(1969)의 무관질문모형의 확률장치를 이용하여 선택된 질문에 응답을 하는 새로운 무관확률화응답모형을 제안하였다. 그리고, 제안한 모형이 Mangat(1994)의 관련질문모형보다 효율적인 되는 조건을 제시하였고, 수치적으로 효율성을 비교하였다. 또한, Leysieffer와 Warner(1976)의 위험함수와 Flinger et al.(1977)의 사생활 보호 측도를 이용하여 제안한 모형이 Mangat의 관련질문모형에 비하여 개인의 사생활을 보호해 주는 측면에서 더 효율적임을 보였다.효율적임을 보였다.

• The Korean Journal of Applied Statistics
• /
• v.28 no.6
• /
• pp.1113-1120
• /
• 2015

We propose a multi-proportions randomized response model by stratified simple random sampling for surveys of sensitive issues of a polychotomous population composed of several stratum. We also systemize a theoretical validity to apply multi-proportions randomized response model (Abul-Ela et al.' model, Eriksson's model) to stratified simple random sampling and derive the estimate and its dispersion matrix of the proportion of sensitive characteristic of population using the suggested model. Two types of sample allocations (proportional allocation and optimum allocation) are considered under the fixed cost. In efficiency, the Eriksson's model by stratified sampling are compared to the Abul-Ela et al.' model.

• Communications for Statistical Applications and Methods
• /
• v.5 no.1
• /
• pp.99-105
• /
• 1998

본 논문에서는 매우 민감한 조사에서 모집단이 여러 개의 집락으로 구성되어 있을 때, 모집단으로부터 집락을 단순임의추출한 후 추출된 각 집락에서 다시 조사단위의 표본을 추출하는 2단계 집락추출법에 확률화응답모형을 적용하였다. 그리고, 일정한 비용 하에서 분산을 최소로 하는 1단계 집락의 수와 2단계 집락에서 추출된 조사단위의 수의 최적값을 구하여 최소분산의 형태를 도출하였다.

• The Korean Journal of Applied Statistics
• /
• v.12 no.2
• /
• pp.585-591
• /
• 1999

Warner(1965)의 확률화 응답 모형을 두 번 연속사용하여 응답자들이 일관성 있는 응답을 했다는 가설을 검정하는 검정통계량을 제안했다. 이것은 양측과 단측 대립가설 모두 검정하는데 이용할 수 있으며, 제안된 검정통계량의 조건분포는 정규분포에 근사한다. 이 검정통계량의 조건부 검정력 함수와 비조건부 검정력 함수를 구하였다.

• The Korean Journal of Applied Statistics
• /
• v.27 no.2
• /
• pp.239-247
• /
• 2014

For a sensitive survey in which the population is composed by several strata with quantitative attributes, we present an additive stratified quantitative attribute randomized response model which applied stratified random sampling instead of simple random sampling to the models of Himmelfarb-Edgell's additive quantitative attribute model and Gjestvang-Singh's. We also establish theoretical grounds to estimate the stratum mean of sensitive quantitative attributes as well as the over all mean. We deal with the proportional and optimal allocation problems in each suggested model and compare the relative efficiency of the suggested two models; subsequently, Himmelfarb-Edgell's model is more efficient than Gjestvang-Singh's model under the condition of stratified random sampling.

• The Journal of the Korea Contents Association
• /
• v.17 no.10
• /
• pp.65-71
• /
• 2017

It is true that there is a possibility of distortion in the statistical surveys or actual surveys depending on which investigator, what purpose, and how research method. Even statistical results are more likely to be 'lying', and statistics on crime or delinquent are sometimes referred to as 'whopper'. There are many reasons for not trusting statistics on crime or delinquent, but one of the main causes is the existence of a hidden crime or an unreported crime. In order to overcome these hidden crime problems, victim surveys or self-report surveys are being used. However, this method also has the problem of underreporting or overreporting depending on the type of crime. Because investigations into crime, delinquency, and deviant behavior are very sensitive, the subjects have a psychological burden. A randomized response model has been developed and used in the field of statistics as a way to induce a true answer to the sensitive content which is burdensome to reveal the experiences of the survey subjects. This technique is a very useful way to solve the problems of victim surveys or self-report surveys. Nevertheless, there are very few cases in the field of criminology in Korea. Therefore, in order to examine the applicability of the randomized response model in the field of criminology, this study used the randomized response model to actually measure the content of prostitution for college students and the effectiveness of the randomized response model was confirmed.