key: cord-0744906-57dpzjbz authors: Zhang, Fan; Ji, Yanjie; Lv, Huitao; Ma, Xinwei; Kuai, Chenchen; Li, Wenhao title: Investigating factors influencing takeout shopping demand under COVID-19: Generalized additive mixed models date: 2022-04-22 journal: Transp Res D Transp Environ DOI: 10.1016/j.trd.2022.103285 sha: 4cf314d55c6ccea296f48844546b7dd3272f53ba doc_id: 744906 cord_uid: 57dpzjbz The COVID-19 pandemic severely hampered the freedom of shopping travel while increasing individuals’ interest in takeout. Although many studies have examined takeout shopping, the available literature provides insufficient evidence on the factors influencing takeout shopping demand under the COVID-19. In this study, generalized additive mixed models were developed based on sampling data of takeout orders in Nanjing before, during, and post the pandemic to measure the associations between takeout shopping demand and neighborhood characteristics at the business circle scale. The results show that population density, house prices, road density, and catering all have a significant impact on takeout shopping demand, while the roles of land use (residential and company indexes) before and post the pandemic are opposite. Besides, the factors influencing the recovery of the demand before and after the pandemic were analyzed. These findings provide important insights into the development of the takeout industry in the post-pandemic era. The COVID-19 pandemic is currently one of the greatest challenges to human life 28 and health, economic development, and social stability (Kummitha, 2020) . Beginning To address these questions, the takeout shopping demand and relative change 93 related to the pandemic in Nanjing were analyzed. The detailed descriptive statistics 94 were performed to examine the spatiotemporal characteristics of takeout shopping 95 demand. Then, the cross-section analysis of takeout shopping demand before, during, 96 and after the COVID-19 was carried out using three generalized additive mixed models 97 (GAMMs), which have proved to be suitable for complex nonlinear situations and 98 provide good prediction accuracy (Hastie, 2015) . Finally, a GAMM was fitted to 99 investigate the factors influencing the recovery of takeout shopping demand, with the 100 relative change in takeout shopping demand before and after the pandemic as the 101 dependent variable, weather and seasonality being the controlling factors. The fixed 102 effects are composed of socio-demographics, the characteristics of takeout business 103 circles, the built environment of the business circle radiation areas, and time-varying 104 variables. Notably, the analysis of takeout shopping demand characteristics was 105 conducted at the business district scale to examine the association between takeout 106 shopping demand and the built environment. 107 The following section presents a review of the existing literature on takeout 108 shopping. Section 3 provides the research design, including the study area, data sources, 109 and modeling approach. Section 4 discusses the model results. Finally students were unable to return to colleges, while more people who were forced to work 178 from home increased their takeout shopping frequency to devote more time to jobs (Li 179 et al., 2020). Furthermore, the catering and retail industries have been hit hard, and 180 shifting from offline to online operations became an important way to stay afloat 181 (iiMedia, 2020). As a result, takeout delivery not only met consumers' needs but also 182 employed those who deliver orders (Noor and Renwick, 2020) . District has a high demand for ordering takeout food online throughout the city, 228 followed by Gulou and Qinhuai Districts in the urban area (Zhang, 2020) . Delivery of 229 goods has also expanded beyond traditional foods to include desserts, beverages, 230 flowers, fruits, vegetables, and other daily necessities. It is worth noting that the 243 2020, a total of 93 local cases were confirmed in Nanjing, and all were cured by the end 244 of March 2020. In May 2020, Nanjing resumed normal production and life order. Until 245 May 2021 (as of this writing), no confirmed cases have been found in Nanjing, which 246 is also known as the post-pandemic era. Therefore, to better explain the impact of the 247 pandemic on takeout shopping demand, we chose takeout order data for three 248 representative periods, namely November 1 to 7, 2019 (before the pandemic), March 1 249 to 7, 2020 (during the pandemic) and November 1 to 7, 2020 (the post-pandemic period). 250 It should be noted that the three chosen periods do not include holidays, major events, 251 severe weather, and other interfering factors. in this study is takeout shopping demand, and the variance of the demand is larger than 354 the mean, so the GAMM assumption follows the negative binomial (NB) distribution. 355 The demand relative change, which is assumed to follow the Gaussian distribution, is (2) is replaced by non-parametric function terms of unknown form, and the basic 369 framework of GLM model is still retained. The average delivery distance (Euclidean distance (Liu et al., 2020)) is 1,749 395 meters with 38.8%, 82.2%, and 98.6% within 1km, 2km, and 3km, respectively (Fig.3) . 396 Orders with a delivery distance of 1,400 to 1,600 meters account for the highest 397 proportion. The average delivery time of orders is 28min, among which, those took 24-398 27min account for the highest proportion (Fig.4) . In addition, the average distance and 399 duration of delivery increased during the pandemic. We believe that such changes are 400 due to fewer nearby takeout supplies being available. (Table 3) . A summary of the findings from this study and other 425 studies is presented in Table 4 . 426 In terms of socio-demographics, the population density in the three models all has The authors declare that they have no known competing financial interests or personal 660 relationships that could have appeared to influence the work reported in this paper. 661 Measuring consumers' level of satisfaction 667 for online food shopping during COVID-19 in Italy using Consumer adoption of online 670 food delivery ordering (Ofdo) services in pakistan: The impact of the covid-19 671 pandemic situation How to do the preliminary research of 674 commercial real estate? E-commerce, 36 678 transportation, and economic geography Geographic distribution of e-shopping: application 682 of structural equation models in the Twin Cities of Minnesota Nonlinear effects of built environment on intermodal 685 transit trips considering spatial heterogeneity Does covid-19 affect the behavior of 688 buying fresh food? Evidence from Wuhan, China Comprehensive evaluation and spatial pattern of 691 regional urbanization level in Henan What is the busiest time for takeout orders? Decision-making Factors of Middle-class Outshopping Behavior in Guangzhou 696 City The effect of COVID-19 and subsequent social distancing on travel 698 behavior Impact of the COVID-19 crisis on short-and consumer behavior Influencing Factors of Online Shopping Frequency 702 of Residents and Spatial Differences of These Factors in Guangzhou City A comparison of online and in-706 person activity engagement: The case of shopping and eating meals Non-linear relationships 709 between built environment characteristics and electric Influential factors and spatial effects of individuals' 713 shopping behavior in Nanjing: A comparative study on four types of goods E-716 shopping in the Netherlands: Does geography matter? Environ Exploring the impact of socio-demographic 719 characteristics, health concerns, and product type on home delivery rates and 720 expenditures during a strict COVID-19 lockdown period: A case study from 721 Home-deliveries before-during COVID-19 724 lockdown: Accessibility, environmental justice, equity, and policy implications To buy or not buy food online: The impact 728 of the COVID-19 epidemic on the adoption of e-commerce in China Can delivery companies keep up with the post-covid-19 e-731 commerce boom? gam: Generalized Additive Models Prediction of Total Demand of O2O Food Delivery Customers Based 736 on SARIMA-SVR Model Examining spatiotemporal changing 738 patterns of bike-sharing usage during COVID-19 pandemic A big-data 741 driven approach to analyzing and modeling human mobility trend under non-742 pharmaceutical interventions during COVID-19 pandemic Research report on Merchants in China's food and beverage takeaway 745 market during the epidemic period China Food Delivery Industry Development Report 2019 and 747 the First Effect of Geographical Distribution of Online 749 Customers on Operational Decisions for Take-out Service Retailers Empirical Analysis Based on Order Data from Eleme Autonomous delivery vehicles to 753 fight the spread of Covid-19 -How do men and women differ in their 754 acceptance? To be online or in-store: Analysis of retail, grocery, 757 and food shopping in New York city Spatial Attributes and Patterns of Use in 760 Smart technologies for fighting Pandemics: The techno and 763 human driven approaches in controlling the virus transmission Relationships between the 38 766 online and in-store shopping frequency of Davis, California residents Review of online food delivery platforms and 769 their impacts on sustainability Assessing regional risk of COVID-19 infection 772 from Wuhan via high-speed rail Integrating Dijkstra's 774 algorithm into deep inverse reinforcement learning for food delivery route 775 planning Factors associated with home-based e-working and e-778 shopping in Nanjing Accessibility or Innovation? Store Shopping Trips 780 versus Online Shopping Transportation technologies, sharing 783 economy, and teleactivities: Implications for built environment and travel Is it safe and ethical to order meals online during the 787 coronavirus outbreak? Factors affecting customer satisfaction and 791 loyalty in online food delivery service during the COVID-19 pandemic: Its 792 relation with open innovation Spatial Pattern of Online and Offline 795 Catering Service in Nanjing：A Case Study of Dianping and Eleme Factors Influencing College Students' Choice of Takeout 798 Vendors-Study in Chengdu How is 801 COVID-19 reshaping activity-travel behavior? Evidence from a comprehensive 802 survey in Chicago How does 804 purchasing intangible services online influence the travel to consume these 805 services? A focus on a Chinese context The influence of the built 808 environment on online purchases of intangible services: Examining the 809 mediating role of online purchase attitudes Does e-shopping for 812 intangible services attenuate the effect of spatial attributes on travel distance and 813 duration? Does e-shopping replace shopping 816 trips? Empirical evidence from Chengdu Innovation diffusion 819 hypothesis or efficiency hypothesis: Spatial penetration of online-to-offline e-820 commerce in China based on Spatial diffusion and penetration about 822 the takeaway O2O e-commerce in China: Feasibility of innovation diffusion 823 hypothesis and efficiency hypothesis The association between spatial attributes and 826 neighborhood characteristics based on Meituan take-out data: Evidence from 827 shanghai business circles Moving towards "mobile warehouse": 830 Last-mile logistics during COVID-19 and beyond Commuting before 833 and after COVID-19 Exploratory Analysis of Factors Affecting 836 Levels of Home Deliveries Before , During , and Post COVID-19 Operation research and optimization of Meituan intelligent Adoption of delivery 841 services in light of the COVID pandemic: Who and how long? Transp. Res. Part 842 A Policy Pract An XGBoost-enhanced 844 fast constructive algorithm for food delivery route planning problem The Impacts of Information and Communication 847 Technologies(ICT) on the Spatial Distribution of Urban Customer Services: A Case Study of Online Takeaway Industry in Beijing Impacts of food accessibility and built environment on on-850 demand food delivery usage Generalized additive models: an introduction with R How does same-day-delivery online shopping reshape 40 854 social interactions among neighbors in Nanjing? The impacts of same day delivery online shopping on 857 local store shopping in Nanjing Study on the influencing factors and spatial 860 characteristics of residents' online consumption in Nanjing To survive or to thrive? China's 863 luxury hotel restaurants entering O2O food delivery platforms amid the COVID-864 19 crisis Analysis of factors influencing delivery e-867 bikes' red-light running behavior: A correlated mixed binary logit approach Nanjing's take-out orders ranked first in the province in 2019 The influence of built environment on the choice of 874 residents' shopping modes: A case study of Nanjing The association between spatial 876 attributes and e-shopping in the shopping process for search goods and 877 experience goods: Evidence from Nanjing The Role of High-speed Rail and Air Travel in the Spread of 880 COVID-19 in China Is compulsory home quarantine less effective than centralized 882 quarantine in controlling the COVID-19 outbreak? Evidence from Hong Kong Analysis on the Spatial Pattern of Catering Facilities in Nanjing Based 885 on POI data Ji 893 Yanjie: Supervision, Funding acquisition, Writing-Reviewing and Editing Writing-Reviewing and Editing. Kuai Chenchen: 895 Data curation, Validation. Li Wenhao: Writing-Reviewing and Editing