Gumbel Mixed Distribution and Mixed Logit Model in Dichotomous Choice Contingent Valuation Study

SOO-IL KIM(김수일) 에너지경제연구원 2005 에너지경제연구 Vol.4 No.1

이분선택형 조건부 가치측정법에서 지불의사액을 추정하기 위해 주로 사용되는 로짓 모델은 두 선택조건하에서 설문 응답자의 확률효용(혹은 확률지출)이 독립적이고 공분산이라는 i.i.d. 가정을 전제로 모형을 추정한다. 지극히 제한적인 모델의 가정을 완화하고 가정의 적합성을 테스트하기 위해 본 논문에서는 Gumbel Mixed Model이라는 다중변수분포를 사용하여 일반화된 추정모형을 제안하였다. 일반화된 추정모형은 기존의 로짓모델이나 이분산 로짓모델을 포함하고 있으며 독립공분산가정을 간단히 검정할 수 있는 장점이 있다. 기존 연구에 사용된 두개의 자료를 이용하여 추정한 결과, 설문에 포함된 두 선택조건 하에서 응답자의 확률지출이 독립적이기는 하나 분산이 같지 않은 경우가 발견되었다. 본 논문의 결과는 기존 가정을 만족하지 않는 설문자료의 경우 로짓함수를 추정에 사용할 수 없음은 물론, 로짓모델을 통하여 구한 지불의사액을 비용편인 분석에 사용할 경우 전혀 다른 정책결정을 할 수도 있다는 것을 시사한다. The simple logit model often used for estimating willingness to pay from dichotomous choice contingent valuation (CV), assumes that the respondent's evaluations of the two states are stochastically independent and homoskedastic. Relaxing restrictive assumptions suggests a generalized estimation technique that utilizes a Gumbel mixed model. Nested within this generalized model are the heteroskedastic logit model and the simple logit. The nesting structure allows for straightforward tests of the homoskedastic-independent error assumptions. Estimation results from sample data show that the logistic distribution, at least theoretically, may not be a suitable distribution for the dichotomous choice CV study and that heteroskedasticity or correlation may provide policy implication in benefit-cost analysis different from the simple logit model.

A generalized mixed-effects model for vaccination data

최재성 한국데이터정보과학회 2004 한국데이터정보과학회지 Vol.15 No.2

This paper deals with a mixed logit model for vaccination data. The effect of a newly developed vaccine for a certain chicken disease can be evaluated by a noninfection rate after injecting chicken with the disease vaccine. But there are a lot of factors that might affect the noninfecton rate. Some of these are fixed and others are random. Random factors are sometimes coming from the sampling scheme for choosing experimental units. This paper suggests a mixed model when some fixed factors need to have different experimental sizes by an experimental design and illustrates how to estimate parameters in a suggested model.

선택실험법 자료에서의 선호이질성 분석을 위한 혼합로짓모형 및 잠재계층모형의 활용

유병국 ( Byong Kook Yoo ) 한국환경경제학회·한국자원경제학회(구 한국환경경제학회) 2012 자원·환경경제연구 Vol.21 No.4

조건부 로짓(Conditional Logit: CL)모형은 모형추정 및 결과해석이 비교적 용이하다는 장점으로 널리 사용되는 반면에 응답자의 선호이질성(preference heterogeneity)을 충분히 반영하지 못한다는 한계를 가지고 있다. 본 연구에서는 최근 지배적인 방식으로 제시되고 있는 두개의 모형인 혼합로짓(Mixed Logit; ML)모형과 잠재계층모형(Latent Class Model; LCM)을 사용하여 우리나라 보령댐 주변 습지보호에 있어서 응답자간 선호이질성을 설명하고자 하였다. 6대광역시와 보령시 표본에 대하여 응답자별 이질성의 존재여부를 검토한 결과 두 지역간 뚜렷한 차이가 있음을 알 수 있었다. 즉 보령시의 경우에는 응답자간 선호이질성이 뚜렷하게 나타난데 반하여 6대광역시의 경우 응답자간 선호이질성이 거의 나타나지 않았다. 따라서 6대광역시의 경우에는 모수추정을 위해 CL 모형의 사용이 가능하나 보령시의 경우에는 선호이질성을 반영하기 위해 ML모형이나 LCM에 근거한 모수추정이 요구된다. 선호이질성의 원인을 규명하기 위하여 교차항이 있는 혼합로짓모형과 잠재계층모형을 고려할 수 있다. 교차항이 있는 혼합로짓모형의 경우 관찰되지 않은 개인단위의 이질성을 설명할 수 있는 장점이 있다. 그러나 두모형을 비교한 결과 LCM이 교차항이 있는 ML모형이 제공하지 않는 추가적인 정보를 보여주는 것으로 나타나고 있다. 따라서 본 연구에서의 응답자간 선호이질성은 혼합로짓모형에 의한 개인적인 수준보다 잠재계층모형에 의한 계층단위에서 더 잘 설명될 수 있다고 할 수 있다. Conditional Logit (CL) model is widely used since its model estimation and interpretation of results of the model is relatively easy, on the other hand, it has the limit of preference heterogeneity of respondents being not fully considered. In this study we used the two models, Mixed Logit (ML) Model and Latent Class Model (LCM) to explain preference heterogeneity of respondents for protection for Boryeong Dam wetland. As a result of the examination for heterogeneity in Boryeong city and six metropolitan areas, we found there was significant difference between two regions. While there was explicit preference heterogeneity within respondents in Boryeong city, we found little heterogeneity within respondents in six metropolitan areas. Thus in the case of six metropolitan areas, CL model can be used for parameter estimation while in the case of Boryeong city, WTP estimates are based on parameter estimates from ML model to reflect the heterogeneity within respondents. Additionally, ML model with interaction and 2-class LCM for respondents in Boryeong city were used to explain the sources of the heterogeneity. The ML model with interaction has advantage of explaining individual unobserved heterogeneity. However The comarison between these two models reflects the fact that LCM provided added information that was not conveyed in the ML model with interaction. Thus, Preference heterogeneity within respondents in this study may be better explained by class level through LCM rather than indiviual level through ML model.

A Cumulative Logit Mixed Model for Ordered Response Data

최재성 한국데이터정보과학회 2006 한국데이터정보과학회지 Vol.17 No.1

This paper discusses about how to build up a mixed-effects model using cumulative logits when some factors are fixed and others are random. Location effects are considered as random effects by choosing them randomly from a population of locations. Estimation procedure for the unknown parameters in a suggested model is also discussed by an illustrated example.

A marginal logit mixed-effects model for repeated binary response data

최재성 한국데이터정보과학회 2008 한국데이터정보과학회지 Vol.19 No.2

This paper suggests a marginal logit mixed-effects for analyzing repeated binary response data. Since binary repeated measures are obtained over time from each subject, observations will have a certain covariance structure among them. As a plausible covariance structure, 1st order auto-regressive correlation structure is assumed for analyzing data. Generalized estimating equations(GEE) method is used for estimating fixed effects in the model.

Effect of electricity generation mix on battery electric vehicle adoption and its environmental impact

Choi, Hyunhong,Shin, Jungwoo,Woo, JongRoul Elsevier 2018 Energy Policy Vol.121 No.-

<P><B>Abstract</B></P> <P>Battery electric vehicles (BEVs) are gaining much attention as the next technology paradigm in the transport sector because it can mitigate the environmental problems, such as greenhouse gas (GHG) emissions, of conventional internal combustion engine vehicles. Many countries are attempting to promote the consumer adoption of BEVs in this sense by providing subsidies and expanding their related infrastructure. The expected environmental effect of BEVs is an important factor in increasing the consumer adoption of BEVs, and the environmental impact of BEVs is directly related to the electricity generation mix. In this study, we analyze how the consumer adoption behavior of BEVs and their environmental impact can be changed by improving the environmental performance of BEVs based on the electricity generation mix. To this end, we estimate consumer preferences in vehicles by using a discrete choice experiment survey and a mixed logit model. Then, on the basis of the estimation results, we simulate changes in consumers’ vehicle adoption behavior according to various electricity generation mixes and measure the environmental impact of these changes. Analysis results show changing the electricity generation mix to renewable-oriented mix can promote BEV's market share up to 10% and reduce GHG emissions up to 5% by 2026.</P> <P><B>Highlights</B></P> <P> <UL> <LI> Effect of power mix on electric vehicle (EV) adoption and its environmental impact. </LI> <LI> Power mix significantly affects EV adoption and GHG emissions in transport sector. </LI> <LI> Renewable power mix promotes EV adoption and reduces GHG emissions most effectively. </LI> <LI> Fossil-fuel power mix promotes EV more than nuclear power mix, but emits more GHG. </LI> </UL> </P>

An analysis of different types of advertising media and their influence on consumer preferences and purchase intentions for digital cameras

안중하,이창원,이종수,Michael Paik 한국마케팅과학회 2012 마케팅과학연구 Vol.22 No.4

This study analyzes consumer preferences for different types of advertising media: print media, the Internet, salespersons, and a new type of media1– an innovative information delivery device based on Microsoft’s multi-touch screen and a single-touch sensor bar. We chose three kinds of digital cameras for our study and analyzed how different types of advertising media influence purchase intention. We used mixed logit model and multinomial logit–OLS two-stage estimation method for this analysis. Our results show that the Internet had the highest influence on consumers’ purchase intentions. The new type of media had the second highest influence. This study reveals that advertisements which use new types of media can positively influence consumers. These findings can inspire companies to use new types of advertising media to develop their advertising strategy portfolios. This proposed model is applicable to other products and types of advertising media in order to analyze consumer preferences and purchase intentions.

사회서비스 지속성이 기혼여성의 경제활동참여 및 참여형태에 미치는 영향에 관한 종단연구: 혼합다항로지스틱 회귀모형(mixed-effects multinominal logistic regression model)의 활용

진선미 ( Jin Sun-mi ),강은나 ( Kang Eun Na ),장용석 ( Jang Yong-seok ) 한국보건사회연구원 2011 保健社會硏究 Vol.31 No.3

본 연구의 목적은 만 7세 미만 미취학아동을 양육하는 여성 1,020명을 대상으로 사회서비스 지속성이 기혼여성의 경제활동 참여 여부와 형태에 미치는 영향을 분석하는 것이다. 분석자료는 한국복지패널조사(Korea Welfare Panel Study) 제1차년도(2006 년)부터 제4차년도(2009년)까지의 자료를 활용하였다. 사회서비스 지속성이 미취학자녀를 둔 기혼여성의 경제활동 참여여부에 미치는 영향을 파악하기 위해 패널로짓분석 (Panel logit analysis)을 실시하였으며, 경제활동 참여 형태에 대한 영향력 분석은 혼 합 다항로지스틱 회귀모형(Mixed-effects multinominal logistic regression model)을 이용하였다. 분석결과는 다음과 같다. 첫째, 본 연구의 독립변수인 사회서비스 지속성은 기혼여성의 경제활동 참여여부 및 형태에 유의미한 요인으로 나타났다. 경제활동 참여 형태와 관련하여 사회서비스 지속성은 비임금근로 및 임시직 ? 비정규직 취업에 긍정적 영향을 미쳤으나, 상용직 취업에는 부정적인 영향을 미치는 것으로 나타났다. 둘째, 미 취학아동을 양육하는 여성의 경제활동참여 여부에 영향을 미치는 요인으로는 연령, 교육수준, 배우자유무, 거주지역, 과거고용경험, 배우자의 월평균 소득이 유의미하게 나타났다. 셋째, 미취학아동을 양육하는 여성의 경제활동참여 형태는 연령, 교육수준, 배우자 유무, 거주지역, 과거고용경험이 비임금근로 및 임시직 · 비정규직에 유의미한 영향을 미 치는 것으로 나타났다. 이러한 결과를 통하여 본 연구에서는 현재 제공되고 있는 사회 서비스의 성격과 효과성을 규명하고, 기혼 여성의 고용가능성 및 노동시장 지위향상을 위한 정책적 그리고 실천적 함의를 제시하고자 하였다. This study is to examine the effects of social service continuity on married women`s labor participation and employment status. The samples of this study were 1,020 married women with children under 7 years of age, and this study used data from the first wave of Korea Welfare Panel Survey(KWPS, 2006) to the fourth wave of KWPS(2009). As statistical methods, Panel Logit Analysis and Mixed-effects multinominal logistic regression model were used to test the factors influencing married women`s labor participation and their employment patterns. The results were as follows. First, social service continuity has a statistically significant positive effect on married women`s labor participation. With relation to employment status, social service continuity has a significant positive effect on self-employed and temporary/irregular job status and has a significant negative effect on regular job status. Second, the economic participation of married women is determined by age, education, spouse`s monthly income, living location, and previous occupations. Third, the employment status of married women is significantly affected by age, education level, spouse`s monthly income, and previous occupations. Based on these empirical findings, the study has discussed some policy and practical implications to improve married women`s labor force participation.

A Generalized Mixed-Effects Model for Vaccination Data

Choi, Jae-Sung 한국데이터정보과학회 2004 한국데이터정보과학회지 Vol.15 No.2

This paper deals with a mixed logit model for vaccination data. The effect of a newly developed vaccine for a certain chicken disease can be evaluated by a noninfection rate after injecting chicken with the disease vaccine. But there are a lot of factors that might affect the noninfecton rate. Some of these are fixed and others are random. Random factors are sometimes coming from the sampling scheme for choosing experimental units. This paper suggests a mixed model when some fixed factors need to have different experimental sizes by an experimental design and illustrates how to estimate parameters in a suggested model.

A Generalized Mixed-Effects Model for Vaccination Data

Jae Sung Choi 한국데이터정보과학회 2004 한국데이터정보과학회지 Vol.15 No.2

This paper deals with a mixed logit model for vaccination data. The effect of a newly developed vaccine for a certain chicken disease can be evaluated by a noninfection rate after injecting chicken with the disease vaccine. But there are a lot of factors that might affect the noninfecton rate. Some of these are fixed and others are random. Random factors are sometimes coming from the sampling scheme for choosing experimental units. This paper suggests a mixed model when some fixed factors need to have different experimental sizes by an experimental design and illustrates how to estimate parameters in a suggested model.