• Journal of Distribution Research
• /
• v.4 no.3
• /
• pp.47-64
• /
• 2000

This study introduces the mother logit model to predict consumer's store choices. The model is not based on the IIA assumptions and thus accounts for substitution among similar alternatives. The choice data as an input to the model is obtained through the conjoint-type choice experiment. The model is applied to consumer's choice of fastfood stores in the context where new store enters the market. The analysis shows that the substitution effects are significant and therefore the mother logit model predicts better than the IIA model. The mother logit model will be useful as well for the market structure analysis in capturing cannibalization among several brands.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.37 no.2
• /
• pp.437-444
• /
• 2017

The purpose of this study is to analyze the factors affecting the reverse mortgage utilization and satisfaction of the elderly. Based on the survey data of the reverse mortgage demand in 2016, we carried out empirical analysis using the binary logit model and the ordered logit model. First of all, as a result of the empirical analysis using the binary logit model, the determinants of using the reverse mortgage were age, region, assets, household member, children with financial help, and education level. As a result of the empirical analysis using the ordered logit model, the determinants of the satisfaction level of the reverse mortgage were estimated to be age, gender, and region. Based on the results of the empirical analysis, it is necessary to find a way to increase the participation rate of the reverse mortgage and to improve the satisfaction of the user.

• Journal of Korean Society of Transportation
• /
• v.28 no.6
• /
• pp.43-54
• /
• 2010

In recent years there have been studies on tour based approaches for freight travel demand modelling. The purpose of this paper is to analyze tour type choice behavior of commercial vehicles which are divided into round trips and chained tours. The methods of the study are based on the decision tree and the logit model. The results indicates that the explanation variables for classifying tour types of commercial vehicles are loading factor, average goods quantity, and total goods quantity. The results of the decision tree method are similar to those of logit model. In addition, the explanation variables for tour type classification of small trucks are not different from those for medium trucks', implying that the most important factor on the vehicle tour planning is how to load goods such as shipment size and total quantity.

• Journal of Korean Society of Transportation
• /
• v.17 no.5
• /
• pp.57-65
• /
• 1999

본 연구에서는 지금가지 통근통행자의 활동참여 분석(activity Participation analysis)이 주로 일과후 활동참여(postwork activity participation)에 초점이 주어졌다는 점을 감안하여 일과전 활동참여(preworkactivity Participation)를 분석하고자 하였다. 이러한 연구배경하에서 본 연구는 통근통행자의 통근통행 이전의 비 통근통행 발생여부와 교통수단 선택의 행태를 네스티드 로짓모형을 이용하여 분석하였다. 본 연구는 모형의 경험적 추정을 위해 1995년 미국 Nationwide Personal Transportation Survey (NPTS)의 일부분으로 수집된 뉴욕 대도시지역의 설문조사자료를 이용하였으며, 추정된 경험적 모형을 바탕으로 의미 있는 결과를 논의하였다. 본 연구에서 추정된 통근통행 이전의 비 통근통행 발생여부와 교통수단 선택의 네스티드 로짓모형은 통계적 측면이나 행태적 측면에서 볼 때 대체로 타당한 분석결과를 보여준다. 교통수단 선택모형(조건부 모형)의 추정결과를 보면 나이, 운전면허 소지여부, 가구내 차량보유대수, 가구내 직장인수, 연간 가구소득, 교통수단의 통행시간이 통근통행자의 교통수단 선택에 의미 있는 영향을 미치는 것으로 확인되었다. 한편 통근통행 이전의 비통근통행 발생여부 선택모형(한계모형)의 추정결과를 보면 나이 성별, 가구내 차량보유대수, 가구의 생애주기가 통근통행자의 비 통근통행 발생여부에 의미 있는 영향을 미치는 것으로 확인되었다. 본 연구로부터의 중요한 발견은 추정된 네스티드 로짓모형의 inclusive value의 추정계수값을 살펴봄으로써 가능한데, 모형의 경험적 추정결과는 본 연구에서 가설화된 네스티드 로짓모형구조의 타당성을 입증시켜 준다. 아울러 직장인의 교통수단 선택은 통근통행 이전의 비 통근통행 발생여부의 선택과 밀접한 관련을 가지면서 이루어진다는 사실을 확인할 수 있었다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.11
• /
• pp.151-156
• /
• 2019

This study aims to develop a distress prediction model and to evaluate distress prediction power for the food services industry by using 2017 food service industry financial ratios. Samples were collected from 46 food service industries, and we extracted 14 financial ratios from them. The results show that, first, there are eight ratios (financial ratio, current ratio, operating income to sales, net income to assets, ratio of cash flows, income to stockholders' equity, rate of operating income, and total asset turnover) that can discriminate failures in food service industries and the top-level food service industries. Second, by using these eight financial ratios, the logit function classifies the top-level food service industries, and failures in the food service industry can be estimated by using logit analysis. The verification results as to accuracy in the estimated logit analysis indicate that the model's distress-prediction power is 89.1%.

• Environmental and Resource Economics Review
• /
• v.14 no.1
• /
• pp.51-73
• /
• 2005

The purpose of this study is estimating the value of recreation of the eighteen national parks in Korea. A conditional logit model and a nested logit model have been estimated for the purpose. The data used for the study have been collected via a national level off-site survey. In addition, the annual aggregate data on the number of visitors to each park have been combined with the survey data to derive more reliable estimates. The paper finds that there are substantial differences in preferences for mountain and marine national parks. Not only the value of each park but also the values of the main characteristics of the parks are estimated.

• Journal of Korean Society of Transportation
• /
• v.28 no.1
• /
• pp.51-61
• /
• 2010

The main objective of this paper is to analyze change of travel patterns according to public transportation reform. The paper uses data collected from Gyeongsan City. So far few researches, especially using before-and-after individual travel survey, have been conducted on analyzing effects of public transportation reform. For this research, some descriptive statistical analyses and statistical hypotheses tests were conducted. Furthermore, some empirical logit models were estimated for analyzing the individual effects of the public transportation reform. Finally, some important foundings, policy implications, and limitations of this research are discussed.

• Journal of Korean Society of Transportation
• /
• v.15 no.4
• /
• pp.35-51
• /
• 1997

The main objective of this research is to develop urban workers' daily travel pattern choice model. For this research, a hovel pattern choice model was empirically estimated by using a survey data collected from Kyongsan and Yeungchun City. For this research, a nested logit model structure was employed. For the model specification, it is hypothesized that urban workers' daily travel pattern choice behavior is represented by two stages of choices with single-destination or multi destination travel pattern choice as the higher stage, and the number of tours as the lower stage. The urban workers' daily travel pattern choice model developed in this research yields intuitively reasonable results. From the empirical results, it is found to be sensible to represent urban workers' daily travel patterns as the nested logit model structure Hypothesized in this research. furthermore, future directions of model development are suggested.

• Environmental and Resource Economics Review
• /
• v.19 no.3
• /
• pp.547-569
• /
• 2010

The purpose of this study is to estimate the benefits associated with the opportunities of multiple destination trips using a mixed logit model. The opportunities that affect the destination choice can be one of the external environments of recreational sites. The data for this study were taken from a survey conducted to the multiple destination trip participants at Seorak National Park in Korea. The results of the research show the specific values of complementary opportunities in the study area.

• The Korean Journal of Applied Statistics
• /
• v.15 no.2
• /
• pp.355-366
• /
• 2002

Nonparametric statistical inference for the parameter of logit model were examined. Usually nonparametric approach is milder than parametric approach based on normal theory assumption. We compared the two nonparametric methods for legit model, the bootstrap and random permutation in the sense of coverage probability. Monte Carlo simulation is conducted for small sample cases. Empirical power of hypothesis test and coverage probability for confidence interval estimation were presented for simple and multiple legit model respectively. An example were also introduced.