• 무역학회지
• /
• 제47권2호
• /
• pp.69-88
• /
• 2022

This paper aims to predict Busan's regional product and employment using the logistic regression models and machine learning models. The following are the main findings of the empirical analysis. First, the OLS regression model shows that the main industries such as electricity and electronics, machine and transport, and finance and insurance affect the Busan's income positively. Second, the binomial logistic regression models show that the Busan's strategic industries such as the future transport machinery, life-care, and smart marine industries contribute on the Busan's income in large order. Third, the multinomial logistic regression models show that the Korea's main industries such as the precise machinery, transport equipment, and machinery influence the Busan's economy positively. And Korea's exports and the depreciation can affect Busan's economy more positively at the higher employment level. Fourth, the voting ensemble model show the higher predictive power than artificial neural network model and support vector machine models. Furthermore, the gradient boosting model and the random forest show the higher predictive power than the voting model in large order.

• 국제지역연구
• /
• 제12권3호
• /
• pp.321-339
• /
• 2008

본 연구는 소비자 선택행위에 대한 설명과 예측에 유용한 이산적 선택모델들 중 널리 알려진 다항 로짓모형(Multinomial Logit Model)을 통해 소비자의 특성과 구매 행동을 분석하는데 그 목적이 있다. 본 연구에서는 미용화장지 시장을 중심으로 제조업자 브랜와 유통업자 브랜드간 경쟁상황을 파악하고 어떤 인구 통계적 특성과 마케팅 변수들이 소비자들의 브랜드 선택에 영향을 미치는가를 파악하고자 한다. 이를 위하여 개개인의 시점별 구매기록이 담겨진 패널(panel) 자료를 연구목적에 맞도록 재구성하고 소비자의 구매시점별 브랜드 선택 확률 및 이에 영향을 미치는 요인들에 대한 분석을 진행하였다. 실증분석 결과 구매자의 연령과 소득이 높을수록 제조업자 브랜드 구매확률이 높아지는 것으로 나타났다. 또한 구매 장소가 달라짐에 따라 브랜드 선택에 중요한 영향을 미치는데, 백화점이나 대형할인점에서는 유통업자 브랜드를 선택하는 것으로 확인되었다. 판매촉진의 경우에는 고소득자는 오히려 구매확률이 낮아지는 것으로 나타났으며, 판매촉진의 형태가 사은품일 경우보다 가격할인이 일어날 경우에 구매확률이 더 높아지는 것으로 나타났다. 한편, 구매단계별 분석결과를 살펴보면, 유명제조업자 브랜드 충성고객은 가격이 높을수록 구매확률이 증가하고, 유통업자 브랜드를 지속적으로 구매하는 고객은 가격할인을 통한 판매촉진에 영향을 많이 받는 것으로 나타났다.

• 한국유통학회지:유통연구
• /
• 제16권4호
• /
• pp.29-64
• /
• 2011

상품의 이용가능성은 소비자 욕구를 충족시킬 수 있는 중요한 점포 능력 중의 하나이다. 그러나 구매하고자 하는 상품이 없는 품절(stockouts)상황이 발생했을때 소비자는 구매의사 결정의 어려움에 직면하고 이러한 어려움은 소비자의 부정적인 반응, 예컨대 점포에 대한 불만족을 야기한다. 본 연구에서는 상품의 이용가능성 상설, 즉 품절상황에서의 소비자 대체선택행동을 이끌어 내는 결정요인들을 파악하여, 이러한 변수들이 품절에 대한 소비자반응과 행동에 어떠한 영향을 미치는가를 규명하고자 한다. 세부적으로 본 연구에서는 기존 연구들과 달리 구매환경의 상황적인 특성(예: 구매 긴급성, 당황정도), 점포 특성(예: 상품구색, 매장편의성), 소비자 특성(예: 브랜드애호도, 점포애호도) 등에 초점을 맞추어 이러한 요인들이 품절시 소비자반응(상품대체, 구매연기, 점포전환 등의 대체선택행동)에 미치는 영향을 실증적으로 밝히는 것을 목적으로 하였다. 특히 본 연구에서는 선행연구에 비해 상품구색, 전반적인 점포 가격 이미지 등의 점포특성 변수들을 추가적으로 고려하여 분석하였다. 이러한 연구목적을 달성하기 위하여 대형할인마트, 슈퍼마켓 등의 다양한 소매업태에서 품절상황을 경험하고 나오는 353명의 소비자들로부터 자료를 수집하여 다항로짓분석(multinomial logit analysis)을 이용하여 연구모형과 연구가설을 검증하였다. 소비자 선택행동에 대한 MNL 모델 추정결과 품절상황에서 상품대체, 구매연기, 점포전환 행동에 대해 구매의 긴급성, 당황 정도, 구매량, 사전 계획구매, 상품구색, 전반적 점포가격 이미지, 브랜드애호도, 점포애호도 요인들이 소비자 선택행동에 유의한 영향을 미치는 것으로 나타났다. 구체적으로 살펴보면, 품절상황시 소비자의 점포 내 상품대체행동을 이끄는 결정요인으로 구매의 긴급성, 당황정도, 구매량, 사전 계획구매, 상품구색의 다양성, 저렴한 점포가격이미지, 브랜드애호도, 점포애호도 등이 유의하였으며, 구매연기를 유도하는 결정요인으로 구매의 긴급성, 구매량, 브랜드애호도 등으로 판명되었다. 품절시 소비자가 다른 점포로 전환하게 하는 요인은 구매의 긴급성, 구매량, 사전 계획구매, 상품구색의 다양성, 저렴한 점포가격이미지, 브랜드애호도, 점포애호도 등으로 파악되었다. 마지막으로 품절시 점포 편의성과 판매원 서비스 요인들은 소비자 대체 행동에 유의한 영향을 주지 못하였다.

• 유통과학연구
• /
• 제13권3호
• /
• pp.79-91
• /
• 2015

Purpose - First, the study identifies and analyzes consumer preferences with regard to health foods and supplements. Second, it identifies and analyzes consumer preferences with regard to the properties of Vitamin C supplements. Third, in order to provide a basic data for the development of Vitamin C supplements and to measure how consumers value the properties of different Vitamin C products, a consumer survey was conducted through the choice-based conjoint model. Based on the results, the research estimates consumers' relative product-related priorities as well as price levels and willingness to pay (WTP) for different product types, and makes suggestions regarding consumer-oriented new product development and progressive directions for the successful launch of health foods and supplements. Research design, data, and methodology - This study aims to define the attributes of health foods and supplements based on several characteristics including their natural ingredients, product price, product originality, natural ingredient content, and additional functional ingredients, and makes suggestions regarding strategic market pricing and product development for health foods and supplements according to customer attitudes and characteristics. The research used choice-based conjoint analysis methodology based on the Multinomial Logic Model and collected 94 questionnaires filled out by users of Korean Vitamin C supplements. Results - Product price is the most influential factor among the five analyzed properties. When consumers buy Vitamin C products, the relative significance level of four of the examined properties is as follows: 40.9% for product price, 23.3% for product originality, 21.9% for natural ingredient content, and 13.9% for additional functional ingredients. Vitamin C content is excluded as it is not a statistically significant factor. It is interesting that supplement manufacturers and retailers consider Vitamin C content to be very important whereas consumers do not regard it as an important factor at the time of purchase. The results for the marginal willingness to pay (MWTP) for each property of Vitamin C supplements show that consumers are willing to pay an additional 11,146 Korean won for a 50% increase in the natural ingredient content. With regard to product originality, consumers are willing to pay an additional 11,301 Korean won for products manufactured in Europe than for products manufactured in China. Moreover, consumers show a greater preference for products manufactured in Korea than in Europe. However, consumers are not willing to pay more for additional Vitamin C or additional functional ingredients added to Vitamin C products. Conclusions - According to the results of consumer research on Vitamin C supplements, which represent a popular health food supplement in Korea, most Korean health food and supplement companies are not consumer- or market-oriented when developing new products. Companies gather information from either R&D specialists or sales managers and their opinions are highly reflected in new product development. The study's results will help companies recognize the importance of understanding consumers' unmet needs in advance to develop new products in the future.

• 한국환경과학회지
• /
• 제23권8호
• /
• pp.1455-1462
• /
• 2014

The purpose of this study is to examine green consumer behavior (green product purchasing behavior and green consumption life) affected by demographical characteristics, and cognition degree of green life (cognition of a green indicator, a green life catalyst system, and environmental problems). It's also to promote strategy and suggest effective activation plans for the vitalization of green consumer behavior. To carry out the task, verification of credibility, multiple regression analysis, two-step cluster analysis, and multinomial logistic analysis were used. The results are as follows: First, the factors that effect green product purchasing behavior were gender, age, cognitive of a green indicator, carbon points system, electricity peak hour system, and seriousness of environmental damage due to lifestyle. Second, the factors that effect green lifestyle were gender, age, carbon grade indicator system, cognition of a green system, and the seriousness of environmental damage due to lifestyle. Third, the comparative group characteristic analysis showed low rates for careless green consumer behavior groups compared to the passive green consumer behavior groups in cognition of a green indicator, green system, and environmental problems. For active green consumer behavior groups, the analysis showed high rates in cognition of carbon grades, eco-labeling, electricity peak hour system, and environmental damage due to lifestyle. In order to encourage green consumer behavior, it's evident that cognition of a green indicator, a green life catalyst system, and environmental problems need to be improved through strategic education and continuous encouragement.

• 한국의류학회지
• /
• 제35권1호
• /
• pp.13-25
• /
• 2011

This study investigates the factors influencing consumer's channel selection for luxury fashion items, specifically the effect of (i) perceived luxury benefits, (ii) perceived online shopping risks, (iii) demographics, and (iv) purchasing behavior. A survey questionnaire was developed and implemented to measure the perceived luxury benefits, perceived online shopping risks, purchasing behaviors, and consumer demographics. A total of 396 responses were analyzed by one-way ANOVA, cross-tab, and multinomial logit analysis with SPSS18.0. The results are as follows. First, those who shop in offline shopping channels tend to be heavy buyers that look for product quality and conspicuousness. They perceive low risks from online shopping and purchase few bag items offline. Second, those who shop online tend to be men and perceive the high benefits of economic value. Third, those who shop in multi channels tend to be men, search for information via the Internet, and purchase few accessory items. Implications for multichannel strategies are suggested.

• Child Health Nursing Research
• /
• 제25권4호
• /
• pp.487-495
• /
• 2019

Purpose: We aimed to identify associations between the frequency of electronic cigarette (e-cigarette) use and smoking-related characteristics among adolescent smokers. Methods: We used data from the 14th Korea Youth Risk Behavior Survey. The participants were 3,722 smokers who had smoked cigarettes on at least one day during the past 30 days. The frequency of e-cigarette use was defined as the number of days that participants used e-cigarettes during the past 30 days, and was divided into four categories: 0, 1~2, 3~9, and ${\geq}10$ days. We used the $x^2$ test and multinomial logistic regression to examine the associations between the frequency of e-cigarette use and smoking-related characteristics. Results: Those who attempted to quit smoking, had used a heated tobacco product (HTP) at least once, or initiated e-cigarette use before middle school were more likely to use e-cigarettes for 3~9 days in the past 30 days. More frequent e-cigarette use (${\geq}10$ days) was associated with a higher frequency and greater amount of cigarette use, lifetime HTP use, and earlier initiation of e-cigarette use. Conclusion: Smoking-related characteristics differed by the frequency of e-cigarette use. Therefore, intervention programs targeting adolescent smokers should consider differences in smoking-related characteristics based on the frequency of e-cigarette use.

• The Journal of Asian Finance, Economics and Business
• /
• 제8권3호
• /
• pp.917-929
• /
• 2021

This study highlights the importance of remittance, the factors which affect the percentage of income remitted, and the investment options available to reduce remittance. For the year 2019, the remittances from Oman totaled $9.1 billion. Oman was among the top remitters with remittance to gross domestic product (GDP) ratio of 11.9%. A survey was conducted on 300 resident expats in Oman. The descriptive analysis shows that the maximum remittance percentage is between 21%-40% of income. The multinomial logistic regression results show that outward remittance depends on gender, age, occupation, number of dependent in Oman, and the number of dependent in the home country. Regarding investment, the most preferred investment option is business, followed by real estate and financial services. Age and education are found to affect investment options. This paper tries to fill the literature gap, especially for the case of Oman, by exploring what determines the level of remittance and the preferred choices for retaining funds. This study adds to the existing literature, as no previous study is available regarding how personal factors can influence the level of remittance and investment in a small oil-exporting developing country like Oman. The study will be helpful to policymakers and academicians in devising policies to retain and invest the outwards remittances in Oman.

• The Journal of Asian Finance, Economics and Business
• /
• 제8권3호
• /
• pp.1217-1230
• /
• 2021

Using a multinomial logit model with the panel-data set of Vietnam manufacturing firms, this paper investigates the impacts of foreign direct investment (FDI) - small and medium enterprises (SMEs) linkages and other factors on SMEs' participation in the global value chain (GVC). We consider GVC firms are those engaging in any of the three modes including (i) using domestic inputs to export (D2E), (ii) using imported inputs to produce for the domestic market (I2P), (iii) using imported inputs to export (I2E). We discover that FDI-SME linkages statistically encourage Vietnamese SMEs to integrate into the GVC via I2P and I2E, while no statistical association between FDI-SME linkage and D2E participation is found. GVCs participation likelihood is also positively correlated with the introduction of new product introduction. The establishment of firms' production facilities in industrial zones and foreign ownership are both reported to be significantly decisive factors to SMEs' decisions on GVC participation. Besides, there is a strong association between firms' attributes, i.e. employment, capital intensity as well as financial access, and their participation in the GVC. Local governance quality (proxied by the Provincial Competitiveness Index) and the share of skilled labor at the province-level can facilitate firms' integration into GVCs, while greater market concentration may be a hurdle to such potential.

• Journal of the Korean Data and Information Science Society
• /
• 제11권1호
• /
• pp.83-90
• /
• 2000

이원분할표의 두 범주형 변수에 대한 독립성을 검정할 때 흔히 카이제곱 검정통계량이 사용된다. 표본추출 모형이 다항이나 곱다항인 경우 이 검정통계량이 독립성 가정하에서 근사적으로 카이제곱 분포를 따르게 되는 것은 잘 알려진 사실이다. 두 주변값이 모두 주어진 경우 독립성 가정하에서 표본추출 모형은 다중 초기하분포가 되며 앞의 모형과 마찬가지로 카이제곱 통계량에 근거한 검정을 사용할 수 있다. 이 연구에서는 주변값이 주어진 경우에 카이제곱 통계량의 소표본 분포를 대표본 분포인 카이제곱 분포와 비교하고자 한다. 표본크기가 작은 몇 개의 경우에 대해 카이제곱 통계량의 소표본 분포를 직접 계산해보았다. 표본크기가 큰 몇 개의 경우는 간단한 몬테칼로 알고리듬을 통해 소표본 분포를 생성하고 카이제곱 확률도와 콜모고로브-스미노브 단일표본 검정을 이용하여 대표본 분포와의 일치성을 알아보았다.