Search

When a firm selects an assortment of products to offer to customers, it uses a choice model to anticipate their probability of purchasing each product. In practice, the estimation of these models is subject to statistical errors, which may lead to significantly suboptimal assortment decisions. In "Randomized Assortment Optimization," Wang, Peura, and Wiesemann show that the standard approach of deterministically selecting a single assortment to offer is not always optimal in this setting: Instead, the firm can do better by selecting an assortment randomly according to a prudently designed probability distribution. The authors show how an optimal randomization strategy can be determined for common choice models, improving performance in realistic data-driven settings. The results suggest that more general versions of the assortment optimization problem—incorporating business constraints, more flexible choice models and/or more general forms of uncertainty—tend to be more receptive to the benefits of randomization. When a firm selects an assortment of products to offer to customers, it uses a choice model to anticipate their probability of purchasing each product. In practice, the estimation of these models is subject to statistical errors, which may lead to significantly suboptimal assortment decisions. Recent work has addressed this issue using robust optimization, where the true parameter values are assumed unknown and the firm chooses an assortment that maximizes its worst-case expected revenues over an uncertainty set of likely parameter values, thus mitigating estimation errors. In this paper, we introduce the concept of randomization into the robust assortment optimization literature. We show that the standard approach of deterministically selecting a single assortment to offer is not always optimal in the robust assortment optimization problem. Instead, the firm can improve its worst-case expected revenues by selecting an assortment randomly according to a prudently designed probability distribution. We demonstrate this potential benefit of randomization both theoretically in an abstract problem formulation as well as empirically across three popular choice models: the multinomial logit model, the Markov chain model, and the preference ranking model. We show how an optimal randomization strategy can be determined exactly and heuristically. Besides the superior in-sample performance of randomized assortments, we demonstrate improved out-of-sample performance in a data-driven setting that combines estimation with optimization. Our results suggest that more general versions of the assortment optimization problem—incorporating business constraints, more flexible choice models and/or more general uncertainty sets—tend to be more receptive to the benefits of randomization. Funding: Z. Wang acknowledges funding from the Imperial College President's PhD Scholarship programme. W. Wiesemann acknowledges funding from the Engineering and Physical Sciences Research Council [Grants EP/R045518/1, EP/T024712/1, and EP/W003317/1]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/opre.2022.0129. [ABSTRACT FROM AUTHOR]

Due to the sheer number of available choices, online retailers frequently use tiered assortment to present products. In this case, groups of products are arranged across multiple pages or stages, and a customer clicks on "next" or "load more" to access them sequentially. Despite the prevalence of such assortments in practice, this topic has not received much attention in the existing literature. In this work, we focus on a sequential choice model that captures customers' behavior when product recommendations are presented in tiers. We analyze multiple variants of tiered assortment optimization (TAO) problems by imposing "no-duplication" and capacity constraints. For the offline setting involving known customers' preferences, we characterize the properties of the optimal tiered assortment and propose algorithms that improve the computational efficiency over the existing benchmarks. To the best of our knowledge, we are the first to study the online setting of the TAO problem. A unique characteristic of our online setting, absent from the one-tier MNL bandit, is partial feedback. Products in the lower priority tiers are not shown when a customer has purchased a product or has chosen to exit at an earlier tier. Such partial feedback, along with product interdependence across tiers, increases the learning complexity. For both the noncontextual and contextual problems, we propose online algorithms and quantify their respective regrets. Moreover, we are able to construct tighter uncertainty sets for model parameters in the contextual case and thus improve the performance. We demonstrate the efficacy of our proposed algorithms through numerical experiments. This paper was accepted by J. George Shanthikumar, data science. Supplemental Material: The data files and e-companion are available at https://doi.org/10.1287/mnsc.2023.4940. [ABSTRACT FROM AUTHOR]

Valence is a crucial concept in studying spatial voting and party competition. The widely adopted approach is to rely on intercepts of vote choice models and to infer, based on their size and direction, how valence affects party strategies in empirical settings. The approach suffers from fundamental statistical flaws. This contribution provides the statistical fundamentals to advance the empirical modeling of valence. It proposes an appropriate modeling approach to interpret intercepts as valences and alternate specifications to parameterize the effects of valence. [ABSTRACT FROM AUTHOR]

Incidental and discrete emotions can influence tourist decision‐making and their preferences for destinations with different types of attractions and activities. In this explorative study, we apply a choice model and the cognitive appraisal theory of emotion to examine the influence of those negative emotions that emerged during Australia's lockdown on tourists' anticipated destination preferences. We find that negative emotions were influenced by both pre‐lockdown emotions and lockdown conditions. Negative emotions exert a significant impact on tourists' selection of destinations and influence their attitudes toward social distancing measures. Travel motivations and gender also affect tourist preferences. Finally, we address the theoretical implications of this study regarding the impact of incidental emotions on travel preferences and make practical suggestions to practitioners on providing targeted products and adjusting management and marketing strategies. [ABSTRACT FROM AUTHOR]

One of the main factors affecting the safety of signalised intersections is the stop/go behaviour during the yellow interval. Although previous research has exhaustively examined drivers' stop/go decision-making, the expected autonomous vehicles' (AVs') stop/go behaviour has not yet been thoroughly investigated. Through a series of simulation experiments developed for conventional and autonomous vehicles using different car-following, lane-changing, lateral placement and stop/go model parameter values, we examine here whether the default VISSIM stop/go parameter values can adequately replicate the observed drivers' behaviour at the considered intersection and assess the suitability of using the currently available options, albeit referring to human drivers, to simulate the expected stop/go behaviour of AVs. We also propose a policy framework for determining the desired behaviour of AVs in yellow interval, which is integrated into an AVs logic and achieved in the last simulation to explore the effect of automation on the stop/go outcome and, hence, on the safety level of signalised intersections. Several data analysis and modeling techniques were used for the formulation of certain scenarios, including binary choice models. The default stop/go parameter values were found unfit to replicate the observed stop/go behaviour and subjected to calibration. Compared to the currently available options, the proposed AVs logic proved to produce the most accurate results, in terms of the stop/go simulation outcome. Regarding the impact of automation on the stop/go outcome, the simulation experiments showed that AVs preferred a more conservative behaviour in favor of road safety, as indicated by the significant reduction (≈15%) in the number of vehicles crossing the stop line during the yellow light and zero instances of red light violation. However, compared to the conservative drivers represented by the default stop/go parameter values, AVs preferred a more rational behaviour in favor of intersection capacity without compromising road safety.

Farmers should increasingly adopt more water‐efficient irrigation technologies—such as drip irrigation—as a result of climate warming and aggravating water scarcity. We analyze how Italian farmers adapt to climate change by changing their irrigation technology mix. We apply a two‐stage econometric model to data from 5876 Italian farms. We find that farmers' initial reaction to increasing temperatures is reducing their surface‐irrigated fractions. When temperatures increase further, farmers switch toward more sprinkler irrigation. Our results show that farmers are not autonomously moving to drip irrigation in response to climate change, suggesting that government incentives are needed to encourage this transition. [ABSTRACT FROM AUTHOR]

This research presents an analytical analysis of the factors related to the preference for using electric vehicles, aimed at estimating their potential adoption in urban freight transport. We employed choice modeling to assess the trade-offs among various attributes associated with Electric Freight Vehicles (EFVs) and predicted the probabilities of EFV adoption within urban freight fleets. This estimation is based on an industry survey and real delivery data from a retail firm. Our findings indicate that electric vehicles could be utilized in approximately 32% of the deliveries for the case company. This, in turn, corresponds to a notable 26% reduction in CO2 emissions resulting from delivery operations. Furthermore, our model enabled us to expand the scope of our analysis to the city level, using Istanbul as a specific example. Our illustration demonstrates that, under the current circumstances, electric vehicles have the potential to account for roughly 25% of all deliveries in Istanbul. [ABSTRACT FROM AUTHOR]

One of the main factors affecting the safety of signalised intersections is the stop/go behaviour during the yellow interval. Although previous research has exhaustively examined drivers' stop/go decision-making, the expected autonomous vehicles' (AVs') stop/go behaviour has not yet been thoroughly investigated. Through a series of simulation experiments developed for conventional and autonomous vehicles using different carfollowing, lane-changing, lateral placement and stop/go model parameter values, we examine here whether the default VISSIM stop/go parameter values can adequately replicate the observed drivers' behaviour at the considered intersection and assess the suitability of using the currently available options, albeit referring to human drivers, to simulate the expected stop/go behaviour of AVs. We also propose a policy framework for determining the desired behaviour of AVs in yellow interval, which is integrated into an AVs logic and achieved in the last simulation to explore the effect of automation on the stop/go outcome and, hence, on the safety level of signalised intersections. Several data analysis and modeling techniques were used for the formulation of certain scenarios, including binary choice models. The default stop/go parameter values were found unfit to replicate the observed stop/go behaviour and subjected to calibration. Compared to the currently available options, the proposed AVs logic proved to produce the most accurate results, in terms of the stop/go simulation outcome. Regarding the impact of automation on the stop/go outcome, the simulation experiments showed that AVs preferred a more conservative behaviour in favor of road safety, as indicated by the significant reduction (z15%) in the number of vehicles crossing the stop line during the yellow light and zero instances of red light violation. However, compared to the conservative drivers represented by the default stop/go parameter values, AVs preferred a more rational behaviour in favor of intersection capacity without compromising road safety. [ABSTRACT FROM AUTHOR]

Online educational platforms increasingly allow learners to consume content at their own pace with on-demand formats, in contrast to the synchronous content of traditional education. Thus, it is important to understand and model learner engagement within these environments. Using data from four business courses hosted on Coursera, the authors model learner behavior as a two-stage decision process, with the first stage determining across-day continuation (vs. quitting) and the second stage determining within-day choices among lectures, quizzes, and breaks. By modeling the heterogeneity across learners pursuing lecture and quiz completion goals, the authors capture different patterns of consumption that correspond to extant theories of goal progress within an empirical field setting. They find that most individuals exhibit a learning style whereby lecture utility changes as an inverted U-shaped function of current progress. This model may also be used as an early detection system to anticipate changes in engagement, and it enables the authors to relate learning styles to final performance outcomes and enrollment in additional courses. Finally, the authors examine how quiz-taking varies across learners in different courses and between those who have paid versus not paid for the option to earn a course certificate. [ABSTRACT FROM AUTHOR]

The objective of this manuscript is to investigate the determinants influencing the selection of over-the-top (OTT) platforms as opposed to traditional television mediums—cable, Internet protocol television (IPTV), and satellite broadcasting—for the consumption of content such as television shows and films. Employing data extracted from the 2020 Media Panel comprising 423,851 observations garnered from personal media diaries, this study scrutinizes the impacts of individual attributes, environmental conditions, and temporal factors on platform choice. The findings reveal a temporal influence characterized by a "Friday effect" and a heightened preference for OTT platforms during early afternoon (12:00–16:00) and late-night hours (00:00–04:00). Notably, the likelihood of selecting OTT platforms is significantly augmented during the late-night period in comparison to other time frames. In relation to individual characteristics, variables such as male gender, younger age, higher educational attainment, and elevated income levels were positively correlated with a predilection for OTT platforms. Additionally, environmental variables such as possession of an unlimited data plan and ownership of a tablet personal computer also emerged as significant predictors for OTT preference. Furthermore, the presence of a beam projector during late-night hours and residing in a household with multiple occupants during afternoon hours also served as contributing factors for OTT utilization. In conclusion, the study offers critical insights for stakeholders in both traditional television and burgeoning OTT markets, providing data-driven recommendations for the strategic allocation of resources in consideration of day-of-week and time-of-day variables. [ABSTRACT FROM AUTHOR]

Limited information, time, or capacity may prevent customers from acting as utility maximizers when making purchase decisions. Rather, they would settle for a good enough option; that is, they stop searching and make a purchase as soon as they find an acceptable alternative. We incorporate this behavior in an assortment-optimization problem. Whereas different approaches to modeling customer choice are adopted in assortment planning, all assume customers are utility maximizers. Our work bridges the research streams of assortment planning and bounded rationality, particularly satisficing behavior. In addition, we define a limit for the search budget of customers, in which customers leave without purchase after examining a certain number of items. This assumption brings a new perspective to the assortment-planning literature, enabling us to capture the choice-overload effect. We prove that the firm's problem of finding the optimal assortment is NP-hard. We further establish certain structural properties of the optimal decision, which allows us to reformulate the model as a mixed-integer program. We analytically derive a tight upper bound on the percentage loss in the firm's expected profit for small instances when it assumes incorrectly that customers are utility maximizers. For larger instances, we take a numerical approach to determine the loss. Our results indicate that firms offering low-involvement products, among those dealing with satisficing customers, are more likely to face substantial profit loss if they ignore this behavior. Supplemental Material: The e-companion is available at https://doi.org/10.1287/deca.2022.0063. [ABSTRACT FROM AUTHOR]

The objective of this manuscript is to investigate the determinants influencing the selection of over-the-top (OTT) platforms as opposed to traditional television mediums—cable, Internet protocol television (IPTV), and satellite broadcasting—for the consumption of content such as television shows and films. Employing data extracted from the 2020 Media Panel comprising 423,851 observations garnered from personal media diaries, this study scrutinizes the impacts of individual attributes, environmental conditions, and temporal factors on platform choice. The findings reveal a temporal influence characterized by a “Friday effect” and a heightened preference for OTT platforms during early afternoon (12:00–16:00) and late-night hours (00:00–04:00). Notably, the likelihood of selecting OTT platforms is significantly augmented during the late-night period in comparison to other time frames. In relation to individual characteristics, variables such as male gender, younger age, higher educational attainment, and elevated income levels were positively correlated with a predilection for OTT platforms. Additionally, environmental variables such as possession of an unlimited data plan and ownership of a tablet personal computer also emerged as significant predictors for OTT preference. Furthermore, the presence of a beam projector during late-night hours and residing in a household with multiple occupants during afternoon hours also served as contributing factors for OTT utilization. In conclusion, the study offers critical insights for stakeholders in both traditional television and burgeoning OTT markets, providing data-driven recommendations for the strategic allocation of resources in consideration of day-of-week and time-of-day variables.

Plant quarantine pests worldwide cause considerable economic damage due to direct plant losses, eradication costs, and contamination measures. Although these losses can threaten the survival of a farm, no country to date has a universal compensation solution that encompasses all agricultural sectors. This paper aims to assess various compensation options by, firstly, calculating potential monetary losses caused by selected quarantine pests, which have not been studied in this context before. Secondly, we calculate farmers' willingness to pay (WTP) for different components of compensation options in plant production using a discrete choice experiment (DCE). We find that monetary losses caused by quarantine pests vary significantly across products and sectors and are highest for intensive vegetable production. The DCE reveals a strong preference for the state compensation and heterogeneity of preferences among farmers depending on their experience with risk management options and perception of pest occurrence. The WTP analysis indicates that requiring preventive plant protection measures as a prerequisite for receiving compensation significantly deters farmers from participating in such a system, as its WTP exhibits the most negative impact compared to all other factors. Due to sectoral differences and path-dependent nature of compensation programs, a universal solution is rather not feasible. Sector-specific solutions must be considered, which does not rule out similar compensation approaches among sectors. Our findings are useful for designing compensation programs for quarantine plant pest cases. [ABSTRACT FROM AUTHOR]

Choice models are used by applied economists for many purposes, such as non‐market valuation or estimating willingness to pay for novel food and product attributes. Mixed‐logit models allow researchers to account for preference heterogeneity and complex decision‐making processes when modelling choices. In mixed‐logit models, parameters of monetary attributes such as prices typically are assumed to follow a negative lognormal random distribution to ensure that the marginal utility of a price increase is strictly negative. However, this practice can cause means and standard deviations of welfare estimates to 'explode' to unfeasibly large levels, as the model assumes there are some marginal utilities of cost approaching zero. This paper examines whether cost non‐attendance, which occurs when respondents ignore costs in stated‐preference studies, could be a cause of inflated welfare estimates when a lognormal cost parameter is used. A two‐class equality‐constrained latent‐class model is proposed, in which the cost parameter is fixed at zero for a cost non‐attender class and is specified as a random lognormal parameter for cost attenders. This proposed model produces mean welfare estimates that are 17 times lower than a mixed‐logit model with a lognormal cost parameter, and 10% lower than a model with a non‐random cost parameter. These results suggest that cost non‐attendance can result in inflated welfare estimates when employing a lognormal cost parameter, and that accounting for cost non‐attendance could be a simple, parsimonious solution to this problem. [ABSTRACT FROM AUTHOR]

Host-generated property images as a visual channel reveal substantial information about properties. Selecting proper images to display can lead to higher demand and increased rental revenue. In this paper, we define, estimate, and optimize the impacts of Airbnb photos on customers' renting decisions. We apply ResNet-50, a convolutional neural network model, to build two separate, supervised learning models to evaluate the image quality and room types posted by Airbnb hosts. Then, we characterize the overall impacts of photo layout by the room type featured in the photo, photo quality, and order of display on the listings' web pages. To address two estimation challenges in the Airbnb setting, namely, censored demand and changing consideration sets, we propose a novel pairwise comparison model that utilizes customers' booking sequence data to consistently estimate the impact of photo layout on customers' renting decisions. Our estimation results suggest that the cover image has a significantly larger impact than noncover photos and a high-quality bedroom cover image leads to the largest increase in demand. Furthermore, we build a nonlinear integer programming optimization problem and develop an algorithm to determine the optimal photo layout. Our counterfactual analysis suggests that a listing's unilateral adoption of optimal photo layout leads to 11.0% more bookings on average. Moreover, depending on the neighborhood and market size, when listings simultaneously switch to the optimal photo layout, they get booked for two to five additional days in a year on average, which boosts revenue by $500 to $1,100. This paper was accepted by Swaminathan, Jayashankar, operations management. Funding: This research was partially sponsored by the MIT Data Science Lab and also benefited from generous support provided by Zalando. Supplemental Material: The online companion and data are available at https://doi.org/10.1287/mnsc.2022.4616. [ABSTRACT FROM AUTHOR]

The purpose of this study was to empirically investigate the effectiveness of the Blended Learning Choice Model on student academic performance. In response to the requirements necessitated by the COVID-19 pandemic, an adaptive blended learning method was developed. This model combined pedagogical needs with digital technology using online learning activities to implement student learning in a digital marketing course at a private undergraduate liberal arts college located in the northeastern United States. The Choice Model of instruction was used for two sections of the digital marketing course during the fall 2020 and spring 2021 semesters. During the fall 2021 semester, the two sections returned to 100% in-person instruction. This research compared the academic performance of students by their choice of delivery instruction method, synchronous remote or classroom, in two sections during the fall 2020 and spring 2021 semesters. For each semester and section, the overall grade point average (GPA) of each student was compared to the GPA earned in the digital marketing course. Students were assigned to one of two groups depending on their selected choice of a remote synchronous delivery method or in-person classroom instruction. A two-sample t-test was performed to compare the grades of students in the remote section to the in-person section. The combined results of the fall 2020 and spring 2021 semesters indicated that students enrolled in the course via remote delivery demonstrated significantly lower GPA compared to students with in-person instruction. [ABSTRACT FROM AUTHOR]

This paper proposed and tested an interpolation method that took advantage of low-fidelity passive positioning data, such as mobile phone (MP) data, to estimate urban tourist visits to tourism sites, which utilized the domain knowledge of regularities in tourist behavior and preferences. The central component of the approach was a model of tourists choosing itineraries. A dataset of 80 Shanghai (China) tourist trajectories was collected through simultaneous MP and global positioning system (GPS) tracking. The results showed that the itinerary choice models that used both types of tracking data were quite similar and behaviorally reasonable. The most important finding was that visits that were estimated with the MP data and model were as accurate as those that were estimated via the GPS data and model. Tourists generate itineraries during their trips. Therefore, monitoring and understanding the itineraries is important for tourism planning and management. The universal mobile phone (MP) positioning data are valuable resources when collecting tourist itineraries. However, a major weakness of the data is the low positioning fidelity. Its position might deviate from an individual's true location by hundreds of meters or even kilometers, and the frequency of the data logging could be hours apart, which significantly limits the quality of the collected itineraries. An approach was developed to more reliably estimate urban tourist itineraries that was based on MP data, which considered the tourists' behavioral characteristics and tour preferences. The approaches could generate itineraries as accurately as when the high-fidelity GPS data were used, and the estimated visits to the tourism sites agreed moderately with the actual visits. The approaches could be further developed to continuously monitor urban tourist activities and provide a relatively low-cost and high-efficiency tool for tourism practices which utilizes MP data better. [ABSTRACT FROM AUTHOR]

After people migrate to a different religious landscape, they will have multiple candidate religions to choose from, and some people may convert to a new religion. This study argues that both individual-level characteristics and a religion's organizational traits are involved in religious conversion. Using cross-sectional samples of Chinese overseas students and scholars in the US in 2018 (n = 1911), we deployed mixed multinomial models to demonstrate how conversion, measured as intergenerational and personal religious changes, is associated with individual-level factors and organizational traits of religions. We found that the choice between Christianity, Buddhism, and Chinese folk religion, compared with no religion, is associated with unique individual-level characteristics. For religious organizational traits, missionary intensity and organized activity intensity are generally associated with a higher likelihood of conversion. This study distinguishes the different levels of the operating mechanism in conversion and points out an interactive and heterogenous model for individuals' choice of various competing religions. [ABSTRACT FROM AUTHOR]

Discrete choice models have recently attracted significant attention to model demand in revenue-management applications, as they can capture the fact that if a product is unavailable, then some customers substitute for this product, whereas others leave the system without a purchase. Although a more sophisticated choice model may capture the choice process of the customers more faithfully, a simpler choice model may result in tractable optimization problems when finding the optimal assortment of products to offer or prices to charge. One approach for coming up with sophisticated choice models is to mix existing ones, where the different segments of customers choose under the different choice models in the mixture. In "Revenue Management Under a Mixture of Independent Demand and Multinomial Logit Models," Cao, Rusmevichientong, and Topaloglu demonstrate that mixing the independent demand and multinomial logit models can significantly increase the modeling flexibility of each of these choice models, while keeping the corresponding operational assortment optimization problems tractable. We consider assortment optimization problems when customers choose under a mixture of independent demand and multinomial logit models. In the assortment optimization setting, each product has a fixed revenue associated with it. The customers choose among the products according to our mixture choice model. The goal is to find an assortment that maximizes the expected revenue from a customer. We show that we can find the optimal assortment by solving a linear program. We establish that the optimal assortment becomes larger as the relative size of the customer segment with the independent demand model increases. Moreover, we show that the Pareto-efficient assortments that maximize a weighted average of the expected revenue and the total purchase probability are nested, in the sense that the Pareto-efficient assortments become larger as the weight on the total purchase probability increases. Considering the assortment optimization problem with a capacity constraint on the offered assortment, we show that the problem is NP-hard, even when each product consumes unit capacity, so that we have a constraint on the number of offered products. We give a fully polynomial-time approximation scheme. In the assortment-based network revenue-management problem, we have resources with limited capacities, and each product consumes a combination of resources. The goal is to find a policy for deciding which assortment of products to offer to each arriving customer to maximize the total expected revenue over a finite selling horizon. A standard linear-programming approximation for this problem includes one decision variable for each subset of products. We show that this linear program can be reduced to an equivalent one of substantially smaller size. We give an expectation-maximization algorithm to estimate the parameters of our mixture model. Our computational experiments indicate that our mixture model can provide improvements in predicting customer purchases and identifying profitable assortments. Funding: The work of Y. Cao was partly supported by the National Natural Science Foundation of China [Grant 72131010]. The work of H. Topaloglu was partly supported by the National Science Foundation [Grant CMMI-1825406]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/opre.2022.2333. [ABSTRACT FROM AUTHOR]

Goals are constructs that direct choice behavior by guiding a decision maker toward desirable (or away from undesirable) end states. Often, consumers are motivated to satisfy multiple goals within a single choice. Although previous research has recognized this possibility, it has not directly formulated models of choice as a multigoal problem. The authors develop such a model, referred to as the multiple-goal-based choice model, which incorporates (1) simultaneous multiple goal pursuit and (2) context-driven goal adaptation but (3) does not require apriori identification of the number or nature of the goals. Goal adaptation within a single choice instance, allied to repeated choices, is the key to empirical identification of multiple latent goals. The proposed model is tested and supported using discrete choice experimental data on digital cameras through multiple validation exercises. The model can lead to significantly different policy implications with regard to consumers’ valuation for new product designs, compared with extant utility-based choice models. [ABSTRACT FROM AUTHOR]

This paper (a) analyses the audience of sports broadcasting from a programme repertoire perspective and (b) clarifies the market structure of sports broadcasting in the television industry by estimating a demand model. We first use item- and user-based modelling to investigate the TV programme repertoires of audiences and conducts cluster analysis to identify sports viewers. Then, we apply econometric choice models to clarify the competition structure of sports broadcasting content on the television market. Based on repertoire clustering, the first part of this study identifies different types of sports fans: fanatic viewers, common viewers, entrainment seekers, and information seekers. The second part shows that sports content is mainly selected by fanatic fans, but has a limited influence on other audiences compared to comedy or drama programmes. The results can help advertisers or sports marketers gain a better understanding of the sports content market and audiences to develop more appropriate marketing strategies. [ABSTRACT FROM AUTHOR]

A key feature of online gaming, which serves as an important measure of consumer engagement with a game, is level progression, wherein players make play-or-quit decisions at each level of the game. Understanding users' level-progression behavior is, therefore, fundamental to game designers. In this paper, we propose a dynamic model of consumer level-progression decisions to shed light on the underlying motivational drivers. We cast the individual play-or-quit decisions in a dynamic framework with forward-looking players and consumer learning about the evolution patterns of their operation efficiencies (defined as the average score earned per operation for passing a level). We develop a boundedly rational approach to model how individuals form predictions of their own operation efficiency and playing utility. This new approach allows researchers to flexibly capture players' over/unbiased/underestimation tendencies and risk-averse/-neutral/-seeking preferences—two features that are particularly relevant when modeling game-playing behavior. We develop an algorithm for estimating such a dynamic model and apply our model to level-progression data from individual players with one online game. We find that players in the sample tend to overestimate their operation efficiency as their predicted values are significantly higher than the mean estimates inferred from their playing history with their completed levels. Furthermore, players are found to be risk-seeking with a moderate amount of uncertainty. We uncover two segments of players labeled as "experiencers" versus "achievers"—the former tend to derive a higher utility from the playing process, and the latter are more goal-oriented and derive a higher benefit from completing the entire game. Two counterfactual simulations demonstrate that the proposed model can help adjust the uncertainty level and configure a more effective level-progression point schedule to better engage players and improve the game developer's revenue. This paper was accepted by David Simchi-Levi, marketing. [ABSTRACT FROM AUTHOR]

In this paper, we study queueing systems with delayed information that use a generalization of the multinomial logit choice model as its arrival process. Previous literature assumes that the functional form of the multinomial logit model is exponential. However, in this work we generalize this to different functional forms. In particular, we compute the critical delay and analyze how it depends on the choice of the functional form. We highlight how the functional form of the model can be interpreted as an exponential model where the exponential rate parameter is uncertain. Furthermore, the rate parameter distribution is given by the inverse Laplace–Stieltjes transform of the functional form when it exists. We perform numerous numerical experiments to confirm our theoretical insights. [ABSTRACT FROM AUTHOR]

The Aggregate Impact of Consumer Reviews on Market Outcome in Differentiated Products Market Jun B. Kim, Seoul National University, Republic of KoreaFollow Abstract We investigate the aggregate impact of consumer reviews on market outcome in a differentiated product category. We model consumers as Bayesian learners who use online consumer reviews to learn and update their beliefs on product quality before their choice. For our empirical analysis, we use aggregate-level, longitudinal data from Amazon.com in the digital camcorder category and estimate the demand parameters. Using model estimates, we conduct two simulation studies and quantify the impact of consumer reviews on the market outcome. We report that the products experience heterogenous market share changes: the standard deviation of market share changes across products and time is 16.7%, ranging from -40% to 20%. In addition, consistent with the previous findings in experience goods, the marginal effect of low consumer ratings is greater than that of high consumer ratings. We discuss model limitations and offer directions for further research.

Studies on the relationship between brand equity and long-term promotional effectiveness have many issues. Brand equity measures can be classified as either consumer-based (CB) or firm-based (FB) measures. Previous literature has mostly focused on one or the other type of measure, mostly due to computational difficulties or data limitations, making it difficult to determine whether the results can be generalized to all types of measure. In addition, long-term promotional effectiveness remains a debatable issue due to complicated long-term effects. Our study is the first to include both CB and FB brand equity measures in a long-term promotional effectiveness study. Using weekly scanner data in the breakfast category covering a three-year period, we are able to compute a variety of popular operationalizations of brand equity measures by employing the aggregate demand model and capture total long-term effects via the persistence model. Our major finding is that the general consensus that higher brand equity generates higher promotional effectiveness applies only to typical FB measures—not CB measures. Moreover, we show a positive accumulated long-term effectiveness and an absence of permanent effects for all our brands. These results offer new insights into the differentiation of major brand equity measures and long-term promotional effectiveness. [ABSTRACT FROM AUTHOR]

The car-oriented approach of cities in Iran has caused to decrease in the share of active travel modes in recent years. In contrast, cycling as an active travel mode can play a crucial role in intra-city movements and reduce motorized travel modes' disadvantages. On the other hand, active modes are more practical for short distances and are less desirable in long distances. In this regard, this study intends to investigate the factors affecting the choice of cycling in home-based trips shorter than three kilometers in Qom. For this study, a sample of 887 residents of Qom city was used. Results showed that four categories of variables, including socio-economic characteristics, travel characteristics, environmental and theoretical factors at the significance level of 10%, affect the choice of cycling on trips shorter than three kilometers. The marginal effect values indicated that having access to a suitable bicycle could increase the probability of choosing cycling by 23.4%. Also, trips for leisure, sports, and meeting acquaintances have tremendous potential for cycling. Results also showed that people living outside the city center were more likely to ride bicycles in trips with a distance of less than three kilometers. Also, higher densities of residents and parks within a radius of 800 meters around the residence and having a bicycle accident experience with positive and negative marks, respectively, affect the likelihood of choosing a bicycle.

Mobility-as-a-service (MaaS) is an innovative framework aiming to promote sustainable mobility via altering users travel behaviour. Despite its potential, empirical evidence of MaaS effectiveness in achieving sustainable goals is scarce due to limited real-world trials and commercial offers of MaaS. This study leverages high-quality data from the Sydney MaaS trial and the Application of Programming Interface (API) to develop an original Error Components-Random Parameter Logit Model that describe the choices of transport modes under the "MaaS era". Analysis results offer the first insight into how multimodal bundles reshape mode choices and the extent to which subscription bundles could be used as a powerful tool to nudge users towards more sustainable choices. The results show that multimodal bundles present an appealing alternative for the users and help them reduce their private car use. Simulations were conducted to provide guidance for designing multimodal bundles that are both attractive to users and friendly to the environment. [ABSTRACT FROM AUTHOR]

After people migrate to a different religious landscape, they will have multiple candidate religions to choose from, and some people may convert to a new religion. This study argues that both individual-level characteristics and a religion’s organizational traits are involved in religious conversion. Using cross-sectional samples of Chinese overseas students and scholars in the US in 2018 (n = 1911), we deployed mixed multinomial models to demonstrate how conversion, measured as intergenerational and personal religious changes, is associated with individual-level factors and organizational traits of religions. We found that the choice between Christianity, Buddhism, and Chinese folk religion, compared with no religion, is associated with unique individual-level characteristics. For religious organizational traits, missionary intensity and organized activity intensity are generally associated with a higher likelihood of conversion. This study distinguishes the different levels of the operating mechanism in conversion and points out an interactive and heterogenous model for individuals’ choice of various competing religions.

Thailand has aimed at becoming a "Seed Hub" in the region given that it is the second largest exporter of seed of field crops in Asia. Maize contributes the largest share of seed exports and is the prototype crop for seed industry promotion. One of the goals of the Seed Hub policy is to develop and export high quality Thai brand-name seeds. The structure of the maize seed industry, however, is such that it is concentrated in a few Multinational Companies (MNCs), and the question remains as to whether national and local companies could give farmers the same benefits as the multinational companies. This study aimed at comparing costs and returns across groups of maize seed companies and determining factors associated with farmers' participation in contracts. The results show that despite a higher cost of production, overall, MNCs give farmers the highest net income because of a higher productivity and higher price they give to the farmers from complying with the strict requirements of the contract specifications, followed by national companies, while local companies give the least. Requirements in household labor, investment in irrigation, size of farm, land rent, and age are important factors in farmers' participation in different groups of seed companies. [ABSTRACT FROM AUTHOR]

We estimate a choice-based aggregate demand model accounting for consumers" consideration sets, and study its implications on market structure. In contrast to past research, we model and estimate consumer demand using aggregate-level consumer browsing data in addition to aggregate level choice data. The use of consumer browsing data allows us to study consumer demand in a realistic setting in which consumers choose from a subset of products. We calibrate the proposed model on both data sets, avoid biases in parameter estimates, and compute the price elasticity measures. As an empirical application, we estimate consumer demand in the camcorder category and study its implications on market structure. The proposed model predicts a limited consumer price response and offers a more discriminating competitive landscape from the one assuming universal consideration set.

The goal of the study is to understand the role of social norm in purchase decisions where demand is revealed in the form of multiple-discreteness. Consumers are socially engaged in various activities through the expectation from others in their community. Actions or decisions are likely to reflect this influence. This implicit or explicit social norm is revealed as the rules, regulations, and standards that are understood, shared, endorsed, and expected by group members. When consumers’ decisions are in distance from the norm, they come to face discomfort such as shame, guilt, embarrassment, and anxiety. These pressure act as a constraint as opposed to utility in their decision making. In this study, the effect of social norms on consumer demand is captured via multiple constraint model where constraints are not only from budget equation but also from psychological burden induced by the deviation from the norm. The posterior distributions of model parameters were estimated via conjoint study allowing for heterogeneity via hierarchical Bayesian framework. Individual characteristics such as age, gender and work experience are also used as covariates for capturing the observed heterogeneity. The empirical results show the role of social norm as constraint in consumers’ utility maximization. The proposed model accounting for social constraint outperforms the standard budget constraint-only model in terms of model fit. It is found that people with longer job experience tend to be more robust and resistant to the deviation from the norm. Incorporating social norm into the utility model allows for another means to disentangle the reason for no-purchase as ‘not preferred’ and ‘not able to buy’.

The authors offer a new conceptualization and operational model of consumer choice that allows context-sensitive information usage and preference heterogeneity to be separately and simultaneously captured, thus transforming the axiom of full information use into a testable hypothesis. A key contribution of the proposed framework is the integration of two previously disjointed and often antagonistic research paradigms: (1) the economic rationality perspective, which assumes stable preferences and full information usage, and (2) the psychological bounded-rationality perspective, which allows context-sensitive preferences and information selectivity. The authors demonstrate that the two paradigms can and do coexist in the same decision-making space, even at the level of individual consumer choices. The proposed information archetype mixture model is tested in four studies that span different product categories and levels of task complexity. The findings have ramifications for choice modeling theory and implementation, beyond the disciplinary boundaries of marketing to applied economics and choice-focused social sciences. [ABSTRACT FROM AUTHOR]

Research on swearwords (used here as a hyperonym to include other phenomena and/or synonyms, including cursing, cussing, profanity and foul language) has been done internationally for many years in a variety of scientific disciplines. In contrast, very little to no research has been done on swearing in the South African context. In this article, two research questions on two specific Afrikaans swearwords, feeks and helleveeg, are answered, viz.: 1. What does the constructionalisation network of the words feeks (shrew) and helleveeg (harridan) look like?; and 2. What are users' opinions about feeks and helleveeg, with specific reference to selfreported frequencies, taboo values, prominence, and gender connotation of the referents? Since the Etimologiewoordeboek van Afrikaans (EWA) ("Etymology Dictionary of Afrikaans") was published in 2008, a number of newer etymological sources (such as the Etymologiewoordenboek van het Nederlands) have been added to Afrikaans linguists' arsenal of sources. With the help of this new information, as well as information from other sources, we are able to lay out clear constructionalisation paths for feeks and helleveeg. One of the interesting new insights is that these words have the same Proto-Indo-European stem, but that they came via different routes firstly to Dutch, and then to Afrikaans. We also make a number of suggestions for visual representations of constructionalisation networks. We argue that end-users still need to be consulted about the intelligibility of our proposed visual representations, for example whether information should be presented from present to past, or vice versa. We also argue that solutions (such as those from EtymologyExplorer) should be found that would allow such visual representations to be automatically scaled to larger data sets. Regarding the second question, we use data collected via short, online polls, as well as related sample sign tests, to compare the similarities and differences of people's perceptions about feeks and helleveeg. We show that these two words differ statistically significantly with respect to the following aspects: * Self-reported frequency (production and perception): Feeks is said/written/heard/ read more often than helleveeg; * Emotional valency: Feeks is more negatively charged than helleveeg; * Salience: Feeks is less salient than helleveeg; * Familiarity: Feeks is better known than helleveeg; and * Gender connotation: Feeks can only be used to refer to women, while helleveeg might also be used to refer to men. If we compare each of the two words in terms of self-reported production and perception on the one hand, and their taboo values for the self and for others on the other hand, then we conclude that respondents think that: * they themselves rarely use feeks, but they still encounter it more frequently (statistically significantly); * they use helleveeg very seldom, and also encounter it very seldom; and * both feeks and helleveeg are more offensive to other people than to themselves. The article concludes with a number of pertinent questions for ongoing and future research. [ABSTRACT FROM AUTHOR]

Word of Mouth (WOM) is powerful, and online reviews are often the most accessible WOM information source in electronic commerce. Maintaining favorable online reputation has been the top priority for businesses, and investments in improving online review valence have been increasing. Extensive studies explored how online reviews might influence sales, however, the results have been inconsistent. This study explores whether and how consumers might incorporate online reviews into decision making based on signaling theory and examines when online review valence influences sales and when it might not. In a signaling perspective, online reviews might serve as a product quality signal, and subsequently, consumers might incorporate less the online review information into decision making if other product information cues such as expert ratings or brands help to verify the product quality. The findings from 633,029 consumer decisions on a hotel-booking website indicate that product quality information cues moderate the effect of online reviews on purchase likelihood. Also, product quality information cues were highly endogenous in estimating the effect of online reviews on sales. Online reviews are not likely to be a significant influencer on sales if the seller signal product quality with convincing information cues. [ABSTRACT FROM AUTHOR]

Offering the ideal product that individuals will systematically repurchase (intensification) and satisfying their need for variety in order to make them cross-purchase within a brand-line's products (diversification) instead of switching toward competitors are two fundamental functions of brand-lines. Both rely on product variety, organized along breadth (complementary products) and depth (substitutes). We analyze the differing impacts that breadth and depth have on brand-line repurchasing behavior. Repurchasing is broken down into four components based on its motivation, intensification (repurchasing the same product) vs. diversification (cross-purchasing), and its temporal perspective, inter-trip (over shopping trips) vs. intra-trip (during purchasing trips). Whereas breadth demonstrates only positive effects on the repurchasing components, it has no protective effect against brand-switching. Conversely, depth has a negative impact on inter-trip product repurchasing, the most frequent repurchasing component, but offers a protective negative impact against brand-switching. We examine the moderating impacts of brand-line quality, brand-line alignability, and household size. [ABSTRACT FROM AUTHOR]

Sequential Recommendation Under the Multinomial Logit Model with Impatient Customers In many applications, customers incrementally view a subset of offered products and make purchasing decisions before observing all the offered products. In this case, the decision faced by a firm is not only what assortment of products to offer, but also in what sequence to offer the products. In "Assortment Optimization and Pricing Under the Multinomial Logit Model with Impatient Customers: Sequential Recommendation and Selection", Gao, Ma, Chen, Gallego, Li, Rusmevichientong, and Topaloglu propose a choice model where each customer incrementally view the assortment of products in multiple stages, and their patience level determines the maximum number of stages. Under this choice model, the authors develop a polynomial-time algorithm that finds a revenue-maximizing sequence of assortments. If the sequence of assortments is fixed, the problem of finding revenue-maximizing prices can be transformed to a convex program. They combine these results to develop an effective approximation algorithm when both the sequence of assortments and prices are decision variables. We develop a variant of the multinomial logit model with impatient customers and study assortment optimization and pricing problems under this choice model. In our choice model, a customer incrementally views the assortment of available products in multiple stages. The patience level of a customer determines the maximum number of stages in which the customer is willing to view the assortments of products. In each stage, if the product with the largest utility provides larger utility than a minimum acceptable utility, which we refer to as the utility of the outside option, then the customer purchases that product right away. Otherwise, the customer views the assortment of products in the next stage as long as the customer's patience level allows the customer to do so. Under the assumption that the utilities have the Gumbel distribution and are independent, we give a closed-form expression for the choice probabilities. For the assortment-optimization problem, we develop a polynomial-time algorithm to find the revenue-maximizing sequence of assortments to offer. For the pricing problem, we show that, if the sequence of offered assortments is fixed, then we can solve a convex program to find the revenue-maximizing prices, with which the decision variables are the probabilities that a customer reaches different stages. We build on this result to give a 0.878-approximation algorithm when both the sequence of assortments and the prices are decision variables. We consider the assortment-optimization problem when each product occupies some space and there is a constraint on the total space consumption of the offered products. We give a fully polynomial-time approximation scheme for this constrained problem. We use a data set from Expedia to demonstrate that incorporating patience levels, as in our model, can improve purchase predictions. We also check the practical performance of our approximation schemes in terms of both the quality of solutions and the computation times. [ABSTRACT FROM AUTHOR]

The brain disease model of addiction is widely endorsed by agencies concerned with treating behavioral disorders and combatting the stigma often associated with addiction. However, both its accuracy and its effectiveness in reducing stigma have been challenged. A proposed alternative, the "choice" model, recognizes the residual rational behavior control capacities of addicted individuals and their ability to make choices, some of which may cause harm. Since harmful choices are ordinarily perceived as blameworthy, the choice model may inadvertently help justify stigma. This paper seeks to fully naturalize the choice model by highlighting the determinants of voluntary action and thus increase its potential for destigmatizing addiction. In light of a deterministic understanding of behavior, it is unreasonable to suppose that addicted individuals could have made different choices in becoming addicted and in subsequent situations. To the extent that stigma is motivated by the supposition that addicted individuals could have chosen otherwise in actual situations, a deterministic understanding of addictive behavior promises to mitigate blame and stigma. [ABSTRACT FROM AUTHOR]

In this study, we consider a new customer choice model which we call the single transition choice model. In this model, there is a universe of products and customers arrive at each product with a certain probability. If the arrived product is unavailable, then the seller can recommend a subset of available products and the customer will purchase one of the recommended products or choose not to purchase with certain transition probabilities. The distinguishing features of the model are that the seller can control which products to recommend depending on the arrived product, and each customer either purchases a product or leaves the market after one transition. We study the assortment optimization problem under this model. Particularly, we show that it is NP‐Hard even if the customer can transition from each product to at most two products. Despite the computational complexity, we provide polynomial time algorithms or approximation algorithms for several special cases, such as when the customer can only transition from each product to at most a given number of products and the size of each recommended set is bounded. Our approximation algorithms are developed by invoking the submodularity arguments, or connecting the problem with maximum constraint satisfaction problem and applying randomized rounding techniques to its semidefinite programming relaxation. We also provide a tight worst‐case performance bound for revenue‐ordered assortments. In addition, we propose a compact mixed‐integer program formulation, which is efficient for moderate size problems. Finally, we conduct numerical experiments to demonstrate the effectiveness of the proposed algorithms. [ABSTRACT FROM AUTHOR]

Mallows-Smoothed Distribution over Rankings Approach for Modeling Choice Assortment optimization is an important problem arising in many applications. Key to this problem is a demand model able to specify the expected demand in response to a given offer set. In "Mallows-Smoothed Distribution over Rankings Approach for Modeling Choice," Désir, Goyal, Jagabathula, and Segev study the smoothing of the existing sparse rank-based choice model using Mallows kernels. Through a detailed case study, the authors show that smoothing can significantly improve the parsimony of sparse rank-based models. Furthermore, despite the exponential support size of the Mallows distribution, an efficient procedure to compute any choice probabilities is developed. This procedure is leveraged to formulate the assortment-optimization problem as a compact mixed integer program, leading to an efficient approach for solving problem instances of practical nature and scale. Finally, to complement this MIP formulation, several near-optimal algorithms are designed by unraveling various structural properties of the Mallows distribution. Assortment optimization is an important problem arising in many applications, including retailing and online advertising. The goal in such problems is to determine a revenue-/profit-maximizing subset of products to offer from a large universe of products when customers exhibit stochastic substitution behavior. We consider a mixture of Mallows model for demand, which can be viewed as a "smoothed" generalization of sparse, rank-based choice models, designed to overcome some of their key limitations. In spite of these advantages, the Mallows distribution has an exponential support size and does not admit a closed-form expression for choice probabilities. We first conduct a case study using a publicly available data set involving real-world preferences on sushi types to show that Mallows-based smoothing significantly improves both the prediction accuracy and the decision quality on this data set. We then present an efficient procedure to compute the choice probabilities for any assortment under the mixture of Mallows model. Surprisingly, this finding allows us to formulate a compact mixed integer program (MIP) that leads to a practical approach for solving the assortment-optimization problem under a mixture of Mallows model. To complement this MIP formulation, we exploit additional structural properties of the underlying distribution to propose several polynomial-time approximation schemes (PTAS), taking the form of a quasi-PTAS in the most general setting, which can be strengthened to a PTAS or a fully PTAS under stronger assumptions. These are the first algorithmic approaches with provably near-optimal performance guarantees for the assortment-optimization problem under the Mallows or the mixture of Mallows model in such generality. [ABSTRACT FROM AUTHOR]

We analyze and study competition in differentiated product market using public data source. Understanding competitive market structure is critical for firms to assess how their products compete against other firms in a given market. In this paper, we estimate consumer demand, extend clout and vulnerability framework, and study competition among multi-product manufacturers in differentiated product market. For our empirical analysis, we adopt choice-based aggregate demand model and estimate consumer demand while accounting for unobserved product characteristics. Once we estimate consumer demand, we compute full price elasticity matrix and investigate intra- and inter- manufacturer substitutions among consumers. This research offers a framework for marketers to analyze and understand market structures, leading them to informed decisions.

This paper explores the impact of COVID-19 on shopping behavior in two neighboring developing economies: Bangladesh and India. While the previous studies investigating the impact of COVID-19 on shopping behavior have relied on Revealed Preference (RP) data, this paper combines RP and Stated Preference (SP) data to develop joint RP-SP discrete choice models. This makes it possible to quantify the relative impact of the situational contexts on the choice of shopping modes of households and to capture the associated heterogeneity arising from the characteristics of the households. Further, comparison of the data and the estimated model parameters of the two countries with substantial socio-cultural similarities provide insights about how differences in the state of e-commerce can lead to different levels of inertia in continuing the pre-COVID behavior. The results will be useful to planners and policymakers for predicting the shopping modes in different future scenarios and formulating effective restriction measures. [ABSTRACT FROM AUTHOR]