Multinomial Logit Research Articles

Exploring the impact of time pressure on motorized two-wheeler riders’ over-speeding behavior

ABSTRACT Despite being inexpensive and the most accessible travel mode, motorized two-wheelers (MTW) are more prone to crashes than other travel modes. Over-speeding is one of the principal causes of such MTW-related crashes. Past studies are nonexistent in exploring the impact of time pressure and psychological constructs on MTW riders’ over-speeding behavior. Therefore, this paper captures the impact of the time pressure situations and identifies psychological segments (based on travel time-related anxiety) on the over-speeding behavior among MTW riders. A two-step modeling approach, including latent class analysis and multinomial logit model (MNL), was utilized for a stated preference survey of 513 Indian MTW riders. The latent class analysis identified three different psychological segments, i.e. minimally anxious, moderately anxious, and highly anxious. Results showed that over-speeding behavior is more likely to increase with travel time-related anxiety levels and hurriedness. The findings are expected to offer potential policy implications for mitigating speeding-related violations.

Contraceptive uptake in postpartum people with and without opioid use disorder and opioid use with co-occurring substance use

BackgroundUsing contraception to delay pregnancy allows people with opioid use disorder (OUD) to choose when they are ready to continue their families. Yet, postpartum contraceptive uptake among people with OUD has not been well characterized. MethodsAnalyses used 73,811 pregnancy episodes among 61,221 people (2016–2021) from the St. Louis University-SSM Virtual Data Warehouse. OUD was defined from the year prior and through pregnancy. Contraceptive uptake was defined within 90-days after delivery. We used Generalized Estimating Equations-type multinomial logit models to assess association of OUD +/- co-occurring substance use disorders (SUDs) with any contraception (yes/no) and type of contraception (effective – pills, patch, ring, injection; or highly effective – long-acting reversible, LARC methods [intrauterine device, implant] and sterilization). ResultsThe sample was 66.0 % white and average age was 27.7 years (±5.6). 32.5 % of pregnancies were followed by contraception initiation, 2.3 % had an OUD diagnosis, and 1.3 % OUD with co-occurring SUD. There was no association between OUD and postpartum contraception receipt, but OUD was associated with decreased highly effective compared to effective method initiation (aOR=0.76; 95 % CI: [0.64–0.91]). OUD plus co-occurring SUD was associated with decreased uptake across all contraception types (aOR=0.81[0.70–0.93]), specifically, highly-effective methods (aOR=0.48[0.38–0.61]). ConclusionsOverall postpartum contraception uptake among people with OUD is comparable to uptake in the non-OUD population. People with OUD plus co-occurring SUDs are particularly unlikely to receive contraception. The reasons people choose contraceptive methods are complex and may differ by SUD severity. More information is needed to understand factors that impact postpartum contraception initiation.

Digital Divide Among Medicare Beneficiaries and the Diabetes Prevention Program

IntroductionSuccessful delivery of the virtual Medicare Diabetes Prevention Program (MDPP) is influenced by a beneficiary's access to a computer and use of the Internet. MethodsUsing the 2020 nationally representative Medicare Current Beneficiary Survey, a three-level categorical dependent variable was created: (1) has a computer AND uses Internet, (2) has a computer OR uses Internet, and (3) has no access to either (reference group). A survey-weighted multinomial logit model was performed to examine associations between socio-demographics, comorbidities, and computer access and Internet use in 2023. ResultsOf study beneficiaries aged ≥65 years with BMI ≥25 kg/m2 and no history of diabetes (n=3,875), 70.8% had a computer AND used Internet; 14.3% had a computer OR used Internet; and 14.9% had no computer AND did not use Internet. Hispanics and non-Hispanic Blacks were less likely (OR=0.28, 95%CI [0.17-0.43]) than non-Hispanic Whites to have a computer AND use Internet. Beneficiaries with less education (<high school) were less likely to have a computer AND use Internet (OR=0.04, [0.03-0.06]) or have a computer OR use Internet (OR=0.21, [0.14-0.33]) than those with more than a high school education. Beneficiaries with lower incomes (<$25,000) were less likely to have a computer AND use Internet (OR=0.27, [0.20-0.35]) or have a computer OR use Internet (OR=0.58, [0.41-0.80]) than those with higher incomes (≥$25,000). ConclusionsApproximately 30% of potentially MDPP-eligible beneficiaries (representing ∼6 million beneficiaries) lacked full digital access. Efforts to encourage enrollment in the virtual MDPP must address digital disparities for beneficiaries at-risk for type 2 diabetes.

Reinforcement learning for Multi-Flight Dynamic Pricing

Dynamic pricing is essential for airline revenue management, requiring quick adaptation to fluctuating market environments and complex customer behaviors. This study addresses the Multi-Flight Dynamic Pricing (MFDP) problem, which presents unique challenges due to interdependent demand between multiple flights and high dimensionality. Traditional studies often assume that the demand function modeling customer behavior is either known in advance or follows a predefined structure, failing to capture the dynamic nature of pricing decisions. To fill this gap, we develop deep reinforcement learning (DRL) algorithms—Deep Q-Network (DQN), Advantage Actor-Critic (A2C), Proximal Policy Optimization (PPO), and Trust Region Policy Optimization (TRPO). By formulating the MFDP problem as a Markov Decision Process (MDP), we design an innovative utility function for the Multinomial Logit (MNL) model that captures realistic features of the airline market, such as competition from high-speed rail, the effect of reference fares, and travel time. We compare the performance of our DRL algorithms with traditional algorithms, including Dynamic Programming (DP), Price Pooling (PP), Inventory Pooling (IP), and Inventory and Price Pooling (IPP). Our experiments demonstrate that DRL algorithms alleviate the curse of dimensionality faced by traditional algorithms, expedite the learning process, and deliver satisfactory performance without relying on predefined demand functions. Among these algorithms, TRPO shows superior performance, achieving 99% of the theoretical optimal revenue, proving its adaptability and stability in dynamic pricing applications. We also highlight the importance of considering the null price in the action space of MFDP problems. The larger the market scale, the more pronounced the effect of the null price in accelerating RL algorithm convergence, leading to more efficient computational resource utilization.

Determinants of organic food purchase behaviour in the European Union: a cross-country analysis guided by the theory of planned behavior

PurposeThe purpose of this study is to analyse the impact of behavioural determinants, such as consumer attitude, subjective norm, behavioural control and sustainable buying behaviour, on organic food purchase behaviour in different EU countries.Design/methodology/approachA structured survey of more than 5000 individuals from Germany, Netherlands, Spain, Italy and Czech Republic was conducted to elicit consumer responses regarding behavioural determinants of organic food selection. Both a structural equation modelling approach and a multinomial logit model were employed to analyse the data.FindingsConsumer attitudes, subjective norms and behavioural control were shown to significantly influence the frequency of organic food purchasing behaviour for the whole sample and to varying degrees across each of the countries. The effect of a sustainable buying behaviour was found to be more pronounced for Germany, Netherlands and Italy than for Czech Republic and Spain.Research limitations/implicationsAnalysis is based on self-reported data on frequency of organic food purchases, which might be subject to recall bias. Overestimation may arise as consumers tend to report increased purchasing due to the “desirable” qualities of organic food. We tried to ensure that the translations bore similar meanings cross country although some miscomprehension by survey respondents was a possibility.Practical implicationsTo stimulate consumer demand for organic products, it is important to design policies and interventions that take into account the impact of both behavioural factors and demographic attributes on consumer decision-making across various locations. For example, market analysts and policymakers may find it useful to exploit the impact of consumer attitude towards the qualities of organics to induce sales of such food in their locations.Originality/valueThis paper combines unique survey data from five EU countries using a common model to explore and compare consumer behavioural preferences for multiple organic food products. In addition to a classic triad of behavioural factors believed to influence organic food purchases, sustainable buying behaviour as a determinant of organic purchasing behaviour was included in the analysis.

Developing Problematic Performance Value Scores: Binding Routine Activity Performance, Environmental Barriers, and Health Conditions.

Community design features, such as sidewalks and street crossings, present significant challenges for individuals with disabilities, hindering their physical performance and social integration. However, limited research has been conducted on the application of Universal Design (UD) to address these challenges, particularly concerning specific demographic groups and population cohorts. Understanding the influence of environmental features on physical performance is crucial for developing inclusive solutions like UD, which can enhance usability and social integration across diverse populations. This study aims to bridge this gap by investigating the complex relationships between environmental barriers, health conditions, and routine activity performance. An index was developed to evaluate users' UD performance based on functional capacity, providing scientifically rigorous and objectively measured evidence of UD effectiveness in creating inclusive built environments. Using data from the Problematic Activities Survey (PAS) conducted in the U.S., Canada, and Australia and targeting individuals with and without functional limitations, multinomial logit models were employed to estimate the probabilities of encountering performance problems. This analysis led to the development of the Problematic Performance Value (PPV) score. The results demonstrated significant disparities in PPVs across various health conditions, particularly concerning curb ramps. Individuals facing mobility issues in their legs/feet, arms/hands, or back/neck encounter more pronounced challenges, especially when curb ramps lack proper design elements. Similarly, individuals with vision impairments face heightened difficulties with traffic signals, particularly due to issues with audible signal systems. These findings underscore the importance of addressing micro-level environmental challenges to accommodate individuals with varying functional capacities effectively. By providing insights into the most problematic daily activities encountered by diverse populations, the PPV score serves as a valuable indicator for guiding environmental design improvements and promoting equitable space usage. This can be used to guide improved UD solutions and decide areas of concentration by providing generalized information on specific environmental features that contribute to user performance.

Young novice drivers’ road crash injuries and contributing factors: A crash data investigation

Objective Collisions are a significant cause of injury and fatality among young novice drivers. Using real crash data, this study further explores the multifaceted and complex nature of young novice drivers’ crash injury risk by synthesizing different driver attributes and crash scenarios in order to update and validate previous research findings and provide more feasible recommendations for preventive measures. Methods Detailed data on traffic crash of young novice drivers were extracted from the National Automobile Accident In-Depth Investigation System (NAIS) in China, and a mixed research methodology using a Random Forest and multinomial logit modeling framework was used in order to explore and study the important influences on traffic crash injuries of young novice drivers in Songjiang District, Shanghai, during the period from 2018 to 2022. Results The results of the study showed that human, vehicle, road and environmental characteristics contributed 36.83%, 22.65%, 17.07% and 23.45% respectively to the prediction of crash injury level of novice drivers. Among the various single factors, driver negligence was the most important factor affecting the crash injury level of novice drivers. Age of the vehicle, crash location, road signal condition and time of crash all had a significant effect on the crash injury level of young novice drivers (95% of the confidence level). Conclusions The study comprehensively analyzed young novice driver crash data to reveal the crash injury risk and its severity faced by young novice drivers in different contexts, and suggested targeted safety improvements. There are similarities and differences with the results of previous studies, in which there are new contributions to understanding the driving risks of young novice drivers in daytime and nighttime.

Comparison of injury severity level involving drivers with and without commercial driver’s licence involved in small vehicle crashes

As vehicle operators, some drivers may have significant impacts on crash-injury severity, and it may be of interest to understand the underlying mechanism of how driver characteristics affect the crash-injury severity. Therefore, this study explored the effect of driver characteristics on crash-injury severity, specifically focusing on drivers with and without commercial driver’s licences (CDL) when driving small vehicles (passenger car, SUV, and pickup truck) with consideration of drivers’ heterogeneity. Crash data from Maryland between January 2015 and December 2018 were analysed, considering factors such as roadway, driver, vehicle, environmental, crash, and temporal characteristics. Furthermore, the random parameters multinomial logit model was adopted to analyse how significant factors affect the crash-injury severity of drivers with and without CDLs. The study identified common features in injury severity among these two types of drivers, as well as notable differences based on their licence types. These findings may provide valuable insights for developing effective long-term strategies to reduce crash.

Determining factors and barriers to the uptake of climate change adaptation strategies of agriculture and aquaculture farm households in Myanmar

PurposeClimate change is an undeniable reality that threatens people’s livelihoods. Flooding and saltwater intrusion, along with the rising sea levels, are affecting agricultural and aquaculture livelihoods in Myanmar’s coastal areas. Although climate change adaptation is gaining popularity as a resilience strategy to cope with the negative effects of climate change, both agriculture- and aquaculture-farmers are more often deterred from implementing climate change adaptation strategies due to practical availability and socioeconomic barriers to adaptation. This study aims to evaluate the barriers and factors that influence farm household’ choice of climate change adaptation measures.Design/methodology/approachThis study was conducted with 599 farm households (484 rice-farmers and 115 fish farmers) based in the coastal areas of Myanmar during 2021–2022 to explore the farmer’s choice of climate change adaptation measures and the determining factors. The multinomial logit regression (MLR) model was used to examine the factors influencing the farmers’ choice of climate change adaptation strategies.FindingsThe study found out that farm households use a variety of adaptation methods at the farm level, with building embankment strategy (23.4%) in agriculture and net-fencing measure (33.9%) in fish farming being the most popular adaptation strategies. Farmers’ decisions to adopt climate change adaptation strategies are influenced by factors such as distance to market, education level of the household head, remittance income and the availability of early warning information, among others. The study also discovered that COVID-19 has had an impact on the employment opportunities of household members and the income from farming as well had a consequential effect on the adoption of climate change adaptation measures. Furthermore, lack of credit (42.4%), labor shortage (52.8%), pest and disease infestation (58.9%), high input costs (81%) and lower agricultural product prices (73%) were identified as major barriers to the adoption of climate change adaptation measures by both agriculture and aquaculture farm households.Originality/valueThis study demonstrates that the COVID-19 pandemic and farm-level barriers are the major factors influencing farm households’ choice of climate change adaptation measures, and that removing practical farm-level barriers and encouraging the adoption of adaptation techniques as potential COVID-19 recovery actions are required. This study also highlighted that the adaptive capacity of agriculture and aquaculture farm households should be strengthened through formal and informal training programs, awareness raising, the exchange of early warning information and the development of proper credit scheme programs.

Examining the Retail Delivery Choice Behavior in a Technology-Aware Market

This study aims to provide valuable insights into consumer preferences for delivery services in online shopping in Chile. The COVID-19 pandemic has accelerated the evolution of delivery and logistics services, leading to increased competition among online stores. Chile, with its highly digitally enabled population and a competitive landscape of online retailers, serves as an ideal reference case for Latin America. By analyzing key delivery attributes such as delivery time, order arrival time range, compensation policies for delivery delays, and delivery prices, we offer valuable insights into consumer behavior. These insights will, in turn, inform the formulation of effective strategies within the online shopping industry. We examine the following aspects: (a) The willingness of consumers to pay for the service attributes; (b) The relative importance assigned to these attributes by consumers; and (c) The relationship between consumer preferences and socioeconomic characteristics. Using Multinomial Logit Models and a database from a Discrete Choice Experiment, we have discovered that the most significant attributes of delivery service are the time until product arrival and the existence of compensation in case of delivery delays. Additionally, we found that consumers are willing to pay more for the same delivery service if the product is large, as large products generally have higher prices. Furthermore, we observed that delivery time preferences vary by gender and for small products, and price sensitivity varies according to educational level, household size, and socioeconomic status. To the best of our knowledge, no previous research of this kind has been conducted for Chile.

Unveiling determinants of household lighting preferences in rural Tanzania: insights for sustainable energy access

This paper investigates the determinants and prospects of household lighting choices in rural Tanzania using a Multinomial Logit Regression Model. The analysis is based on data from 4671 households, focusing on three lighting options: electricity, solar energy, and candle lighting. The results reveal significant factors influencing these choices, including household head characteristics, household size, marital status, education, employment status, number of rooms, and income. Key findings indicate that the age of the household head negatively influences the likelihood of choosing grid-electricity, while having a male head of household significantly reduces the probability of opting for any lighting option. Larger household size is negatively associated with choosing electricity and candle lighting. Marital status shows that married households are more likely to use candle lighting. Employment status positively impacts the likelihood of adopting all three lighting options, with employed household heads being more likely to choose modern lighting solutions. Income levels are crucial, as higher income significantly increases the probability of selecting electricity and candle lighting, but not solar energy. These findings provide valuable insights for policymakers and stakeholders aiming to enhance sustainable energy access in rural Tanzania. It highlights the importance of addressing socio-economic factors to promote the adoption of modern and sustainable lighting technologies.

Optimizing Inventory and Pricing for Substitute Products with Soft Supply Constraints

This paper presents a profit optimization model for substitute products in a competitive, time-sensitive market with scarcity and shifting user preferences. The model maximizes profit, considering production costs and inventory maintenance. It uses a discrete choice model to represent demand, sensitivity to price, availability, and changing preferences. A two-phase PSO-type metaheuristic solution tackles the nonlinear, recursive model, efficiently managing inventories and evolving consumer preferences. The model integrates production decisions, inventories, and sales prices, considering scarcity conditions and user preferences. It uses a multinomial logit for the consumers’ demand function with soft exogenous constraints, which influence utility and change consumption preferences and choices. This research offers a tool for companies to manage stock, production, and pricing in a context where goods are substitutes, providing a new perspective on business strategy.

Patient Preferences for Operative Versus Nonoperative Treatment of LC1 Pelvis Fracture: A Discrete Choice Experiment.

To quantify how patients with lateral compression type 1 (LC1) pelvis fracture value attributes of operative versus nonoperative treatment. Discrete choice experiment. Three US Level 1 trauma centers. Adult survivors of an LC1 pelvis treated between June 2016 and March 2023 were identified from institutional registries. The choice experiment was administered as a survey from March through August 2023. Participants chose between 12 hypothetical comparisons of treatment attributes including operative or nonoperative care, risk of death, severity of pain, risk of secondary surgery, shorter hospital stay, discharge destination, and independence in ambulation within 1 month of injury. The marginal utility of each treatment attribute, for example, the strength of participants' aggregate preference for an attribute as indicated by their survey choices, was estimated by multinomial logit modeling with and without stratification by treatment received. Four hundred forty-nine eligible patients were identified. The survey was distributed to 182 patients and collected from 72 patients (39%) at a median 2.3 years after injury. Respondents were 66% female with a median age of 59 years (IQR, 34-69 years). Before injury, 94% ambulated independently and 75% were working; 41% received operative treatment. Independence with ambulation provided the highest relative marginal utility (21%, P < 0.001), followed by discharge to home versus skilled nursing (20%, P < 0.001), moderate versus severe postdischarge pain (17%, P < 0.001), shorter hospital stay (16%, P < 0.001), secondary surgery (15%, P < 0.001), and mortality (10%, P = 0.02). Overall, no relative utility for operative versus nonoperative treatment was observed (2%, P = 0.54). However, respondents strongly preferred the treatment they received: operative patients valued operative treatment (utility, 0.37 vs. -0.37, P < 0.001); nonoperative patients valued nonoperative treatment (utility, 0.19 vs. -0.19, P < 0.001). LC1 pelvis fracture patients valued independence with ambulation, shorter hospital stay, and avoiding secondary surgery and mortality in the month after their injury. Patients preferred the treatment they received rather than operative versus nonoperative care.

Parking preferences of tourists in Sun Moon Lake scenic area

Sun Moon Lake is a famous tourist attraction in Taiwan and around the world. However, as Sun Moon Lake is surrounded by mountains and has limited land to develop, traffic congestion around the lake area is commonplace during peak holiday hours. This study focuses on the parking choices of visitors to Sun Moon Lake and develops a stated preference (SP) instrument with multiple scenarios to evaluate parking preferences under various financial and service level scenarios. We estimate Multinomial Logit (MNL) and Mixed Logit (ML) models, accounting for the panel nature of the data (PDML)to identify preferences for parking choices of visitors to Sun Moon Lake. The focus is on understanding how parking price, travel time, walking time, scenery, and transfers between public transport affect visitors’ parking choices. Unlike the findings of studies in metropolitan areas, which often find that parking price was the deciding factor, visitors’ parking decisions in the tourist area were more concerned with time factors, such as the time to search for places to park and traffic congestion, possibly due to the less frequent use of tourist venues. Although raising parking price can suppress parking demand in the scenic area, other parking management mechanisms work better, such as the construction of new and suitable outer parking lots with transfer buses to relieve the heavily congested traffic in the scenic area. In addition, we find that using the parking space in the area can be improved by beautifying the landscaping between the parking lots and the tourist spots, enhancing the pleasure of traveling along the routes, introducing multiple transfer modes, and providing real-time traffic information to tourists.

Group Penalized Multinomial Logit Models and Stock Return Direction Prediction

Group Penalized Multinomial Logit Models and Stock Return Direction Prediction

ARTIFICIAL INTELLIGENCE ROLE IN OPTIMIZING ELECTRIC VEHICLE CHARGING PATTERNS, REDUCE COSTS, AND IMPROVE OVERALL EFFICIENCY: A REVIEW

The global popularity of electric cars (EVs) as a sustainable means of transportation, reliable and efficient charging infrastructure is essential. Traditional electric vehicle charging involves connecting the car to a power source and waiting for the battery to charge. However, AI has allowed us to improve charging patterns, reduce costs, and boost efficiency. This article examines how AI algorithms are changing electric car charging. If electric vehicle (EV) charging and discharging are not coordinated, the power supply infrastructure will be overrun. Demand response like dynamic pricing might encourage electric vehicle owners to participate in scheduling initiatives. Thus, EV charging and discharging scheduling and dynamic pricing model research are crucial. Artificial intelligence-based models for EV charging predictions and scheduling have been the focus of researchers. These models outperform linear, exponential, and multinomial logit optimization approaches. Due to the novelty and ongoing development of EVs returning electricity to the power grid, vehicle-to-grid (V2G) systems have received little attention. Thus, a complete analysis of EV charging and discharging research is needed to identify gaps and improve future studies. This study categorizes EV charging and discharging studies into forecasting, scheduling, and pricing techniques. The work links forecasting, scheduling, and pricing processes and identifies research gaps in EV discharge scheduling and dynamic pricing models.

Coping With Intimate Partner Violence: The Effect Of Women’s Resources On Their Response To Violence In Turkey

Abstract: The aim of this study is to investigate the interaction between the personal and social resources of women who have experienced intimate partner violence (IPV) and their responses to two specific types of violence, namely; physical and sexual violence. The study utilizes survey data from the "2014 Domestic Violence against Women in Turkey Survey", a follow-up to the 2008 nationwide survey on violence against women, which includes a wide range of socio-demographic background questions as well as questions on IPV and various types of violence against women and girls. This quantitative research represents one of the rare studies focusing on women's individual and institutional responses to IPV and it is also based on findings from the most recent field study on Domestic Violence against Women in Turkey. The sample design of the research is a multi-staged stratified cluster sampling and 11,247 households completed the survey from 12 statistical regions, urban and rural areas of Turkey. During the field visits, 13,310 women were identified as eligible for interviews, and 8,960 women were approached for face-to-face interviews. The response rate for the women interviews was recorded at 83.3 percent. Multinomial logit models were employed to elucidate the relationship between the responses of women who had been subjected to violence and their individual resources. The findings of the study reveal that individual resources such as age, higher education, urban residence, and being in regular employment with social security significantly increase the likelihood that women will engage in at least one institutional or non-institutional response to IPV.

An Approach to Assessing Farm-Scale Adaptation to Climate Change: The Case Study of Prespa Park

Developing appropriate adaptation practices and coping mechanisms for climate change and evaluating the variables affecting households’ choices are critical for ensuring sustainable agricultural production. In addressing the above issues, this paper presents the results of a multi-method approach at the farm level conducted in the case study of Prespa Park. The data collected by a participatory process were analyzed using descriptive statistics, the 5-point Likert scale, the Delphi method, and a multinomial logit model. The typology of coping and adaptation practices to climate change preferred by households and the reasons for those who failed to adapt are presented. Results revealed that improving technologies for increasing soil health was the most preferred adaptation practice, followed by planting early maturing and drought-tolerant food crop varieties, practicing water-saving irrigation methods or technologies, planting agroforestry systems, and finally perennial agriculture. In terms of coping mechanisms, engaging in off-farm activities was the most used, followed by collecting fuel wood for sale, selling assets like livestock, increasing water storage capacity, and changing farming structure. Small farm holdings, financial constraints, limited off-farm employment opportunities, inadequate infrastructure and technology, and a lack of information about adaptation practices were identified as the main barriers to undertaking adaptation. Performing the multinomial logit analysis, the variables that positively and significantly improve households' ability to adapt to climate change were identified and evaluated. The results of this study should help policymakers and climate change planners come up with better practices for the agricultural sector to adapt to the effects of climate change.

Perceptions and knowledge of ecosystem services in urban river systems, Eastern Cape, South Africa

The ecosystem services concept has been studied in rural contexts but not as much in urban landscapes, particularly for aquatic ecosystems in Africa. There is little knowledge of how people perceive ecosystem services (ES) supplied by rivers in urban centres and the factors influencing their perceptions. To fill this gap, the research presented here addresses two objectives: 1) to assess local communities’ perception and knowledge of different types of ecosystem services of an urban river in South Africa, 2) to assess the influence of particular demographic characteristics (age and gender), socio-economic characteristics (education, housing type, income, and years of residence), and prior knowledge of the ecosystem services concept on people’s perception of the ecosystem services provided by the Swartkops River in the Eastern Cape of South Africa. To address these objectives, a team of ten trained field staff collected 181 surveys between May–June 2021 in communities within the river catchment. We used multinomial logit models to analyse the relationships between demographic and socio-economic variables, and people’s perception of ecosystem services. For the majority of the ecosystem services, there was almost an equal split between the perception “sometimes” and “never” that the river provides a particular ecosystem service, and people were less likely to say that the river always provide a particular service. Prior knowledge of the ecosystem services concept, years of residence, age, income and housing type significantly influence people’s perception of ecosystem services. Prior knowledge, was important for the intangible services, whereas years of residence was more important for tangible services with direct use value. Level of formal education was not a significant variable. An important implication of this study is that ecological interventions can be designed through an understanding of factors influencing people’s perceptions of ecosystem services.

Group-constrained assortment optimization under the multinomial logit model

AbstractWe study an assortment problem under the multinomial logit model with two new types of group constraints that are motivated by a joint project with the German car manufacturer BMW. Under group constraints, products are either attributed to exactly one group or to several groups at once and there is either a bound on the number of products offered per group or on the number of groups from which products are offered. We formulate both optimization problems as binary fractional linear program and provide reformulations that can be solved using state-of-the-art solvers. Finally, we conduct a numerical study and find that all instances of the products-per-group constrained problem as well as small to medium size instances of the number-of-offered-groups constrained problem can be solved within fractions of a second, whereas large instances of the latter problem might take some seconds to be solved.