Search

The successful management of crowdsourcing risk has gained considerable attention because of the challenges involved in obtaining the satisfactory performance from crowdsourcing activities. However, the mechanism of how various critical risk factors affect crowdsourcing performance has been poorly investigated. Empirical evidence based on a dataset of 169 samples from China shows that three technical (i.e., complexity, requirement, and task) and three social (i.e., seeker, solver, and relationship) subsystem risks not only directly but also interactively influence the crowdsourcing task performance. Technical subsystem risk factors, while worsen performance, diminish the negative effects of social subsystem risk factors on the crowdsourcing task performance. The diminishing effects of three technical subsystem risk factors are also differentiated. High requirement and task risks lessen the negative effects of seeker, solver, and relationship risks on the task performance, whereas the high complexity risk only weakens the negative effect of the solver risk. Therefore, various crowdsourcing risk factors affect the performance in different ways and directions (either negatively or positively). Risk can perform dual roles in influencing the performance. Hybrid strategies that integrate social and technical subsystem elements may be effective in achieving a superior task performance for the crowdsourcing risk management. Managers of crowdsourcing platforms can also optimize social and technical environments to improve the fit between social and technical systems.

PurposeThis study aims to investigate the determinants of herd behavior in online hotel service evaluations, focusing on the cultural and geographic distance characteristics of customers.Design/methodology/approachOn the basis of 381,462 TripAdvisor reviews of hotels in the USA written by more than 100,000 customers from 92 countries, this study uses the empirical analysis to explore the collective roles of cultural distance, geographic distance and hospitality experience on herd behavior in online hotel ratings.FindingsCultural and geographic distances between customers and product and service locations positively affect herding and these two effects are substitutable. The hospitality experience of customers attenuates the impacts of distances on herding. These results are robust for multiple hotel service ratings.Practical implicationsFindings help hotels understand perceptual biases of customers on hotel services under the social influence and consequently develop effective marketing strategies to boost hotel revenues and increase profitability.Originality/valueThe research contributes to hospitality and online review literature by understanding how cultural and geographic distances shape online hotel service evaluations under the root of the uncertainty of decision-making and the observation of others’ behavior. The research also contributes to the distances in international business literature by deepening the understanding of the substitution and heterogeneity of distance effects. Methodologically, a time-varying and monotonously increasing variable is constructed to depict customers’ hospitality experience. The extensive data volume ensures the generalizability of our results.

This study investigates the heterogeneous effects of online reviews on hotel revenue when online reviews and American Customer Satisfaction Index (ACSI) coexist. Using a sample of reviews for 303 H...

Cost stickiness, which is also termed cost asymmetry, describes the asymmetric relationship between revenue and cost. In this study, we examine whether high-speed railway (HSR) connection affects the cost stickiness of tourism firms. Employing a sample of 324 Chinese tourism firms from 2003 to 2018 and applying a difference-in-difference (DID) method, we find that the cost stickiness of tourism firms increases after HSR connection. Our results also reveal that the relationship between HSR connection and cost stickiness is more pronounced in firms with higher free cash flow (FCF), higher labor costs, and in state-owned enterprises (SOEs). Our research advances an in-depth understanding of the cost behavior in tourism firms and sheds light on the policy effect of HSR connection.

The ability to understand the causes of customers' complaints is critical for hotels to improve their service quality, customer satisfaction, and revenue. This study adopts a novel structural topic model text analysis method to analyze 27,864 hotel reviews in New York City, and show that it leads to improved inference on consumer dissatisfaction. Our results reveal 10 topics, whose appearances in the negative reviews are substantially higher than those in the positive reviews. In addition, we investigate how customer complaints vary across different hotel grades. Results indicate that customer complaints for high-end hotels are mainly related to service issues, whereas customers of low-end hotels are frequently annoyed by facility-related problems. This research contributes to the hospitality literature by enhancing our understanding of the aspects of hotel customers’ dissatisfaction through rigorous statistical analysis, their correlations, and importance for different hotel grades.

In this study, we focus on the effects of doctors’ provision of online health consultation service (OHCSV) on the review valence provided by patients. We develop a unique panel dataset based on over 3,000,000 reviews for 493,548 doctors on a leading online healthcare community in China. Coarsened exact matching and difference-in-difference estimators are used to determine causality. Findings corroborate that doctors’ provision of OHCSV increases the review valence posted by patients and significantly changes the valence structure by increasing and decreasing the proportion of positive and negative valences, respectively. Effects of OHCSV also differ among various review dimensions and doctors’ hierarchical levels. Doctors at lower hierarchical levels experience stronger positive provision effect than their counterparts. Such findings contribute to literature on sustainable doctor–patient relationship and expert service, and similarly generate implications for the sustainability of health platforms.

Agglomeration benefits firms because it minimizes consumers' search costs and information asymmetry. However, the effects of agglomeration may change due to the considerable amount of information available in online platforms. Thus, drawing upon the theoretical lens of information transparency and agglomeration theory, we investigate how online reviews change the effect of agglomeration on firm performance. A total of 11,528 observations of 467 hotels show that agglomeration density positively affects firm revenue in the hotel sector. However, such an effect decreases with the increase in online review quantity (including volume and length). This change is manifested in two forms, namely, minimizing information asymmetry and increasing competition cost due to information transparency. Moreover, the positive effect of agglomeration density decreases due to the quantity of high-quality (few spelling errors and easy to understand) online reviews, but not with low-quality. We further discover that the substitution effect of online review quantity on agglomeration density is stronger for independent hotels than for branded ones, whereas the effect is similar among hotels with various qualities. Our findings provide implications for the management of online channels in changing the local agglomeration effectiveness.

This study investigates the differential roles of negative and positive reviews on the growth of branded and independent hotels and their underlying interplays. We construct a dataset by combining the revenue data with TripAdvisor review data of hotels in four Texas cities. Based on the resource-based view of firm growth and signaling theory, we obtain several novel findings. Positive reviews exert greater positive effects on the growth of independent hotels than branded hotels, and high growth in turn brings them additional positive reviews. Such a positive feedback loop helps explain why independent hotels benefit more from positive reviews than branded hotels. Negative reviews negatively affect the growth of independent hotels. Although a high growth of branded hotels leads to more negative reviews, the latter does not jeopardize their growth significantly due to the protecting role of brands.

This study explores the antecedents and consequences of health information privacy concerns in online health communities by integrating the dual calculus and protection motivation theories. On the basis of survey data from 337 users, health information privacy concerns, together with informational and emotional support, significantly influence personal health information (PHI) disclosure intention. Privacy concerns are negatively influenced by two coping appraisals (i.e., response efficacy and self-efficacy) and positively affected by two threat appraisals (i.e., perceived vulnerability and perceived severity). The perceived health status differentially moderates the effects of privacy concerns and informational support on the PHI disclosure intention.

This study investigates the collective influences of cultural, hotel, and reviewer characteristics on online ratings in the hotel sector. Based on over 243,000 TripAdvisor reviews for hotels in 24 US cities, we empirically find a negative relationship between the reviewers' power distance and their online hotel ratings, thereby indicating that cultural factor plays a significant role in the customers’ online rating behavior. The negative effect of power distance on online hotel ratings is weaker for chained hotels than for independent hotels. This negative effect is also weaker for reviewers with more travel experience than for those with less travel experience. The robustness check demonstrates that these findings are applicable for ratings on product features that involve staff interactions, such as service, value, rooms, and cleanliness.

Purpose Discovering the research topics and trends from a large quantity of library electronic references is essential for scientific research. Current research of this kind mainly depends on human justification. The purpose of this paper is to demonstrate how to identify research topics and evolution in trends from library electronic references efficiently and effectively by employing automatic text analysis algorithms. Design/methodology/approach The authors used the latent Dirichlet allocation (LDA), a probabilistic generative topic model to extract the latent topic from the large quantity of research abstracts. Then, the authors conducted a regression analysis on the document-topic distributions generated by LDA to identify hot and cold topics. Findings First, this paper discovers 32 significant research topics from the abstracts of 3,737 articles published in the six top accounting journals during the period of 1992-2014. Second, based on the document-topic distributions generated by LDA, the authors identified seven hot topics and six cold topics from the 32 topics. Originality/value The topics discovered by LDA are highly consistent with the topics identified by human experts, indicating the validity and effectiveness of the methodology. Therefore, this paper provides novel knowledge to the accounting literature and demonstrates a methodology and process for topic discovery with lower cost and higher efficiency than the current methods.

Hotel managers increasingly compose rote responses (e.g., similar responses repeated from prior responses) to reviews when dealing with customers' opinions while improving efficiency and saving cost. However, such responses may be ineffective because they are unspecific and add little information to reviewers' comments. To address this dilemma, we examine the degree of managers' rote responses to identify the similarity of managers' response texts. Basing on panel data of 334,671 reviews and 169,794 manager responses from 868 hotels in TripAdvisor, we find that customers leave fewer reviews and lower review valence during the subsequent month, when the level of roteness of managers’ responses is higher. However, rote responses only decrease the volume of subsequent positive reviews but have no influence on the volume of negative reviews. Moreover, such effects are weaker for chain hotels than for independent ones. Therefore, managers should differentiate their responses to online reviews, particularly for independent hotels.

Using data of online ticket sales for attractions in the seven provinces of South Central China, this study focuses on the impact of COVID-19 on tourists' destination preferences after the end of lockdown. Empirical results reveal that tourists' destination preferences have changed significantly, which holds under a number of robustness checks. Specifically, we find that tourists avoid traveling to destinations with more confirmed cases of COVID-19 relative to their places of origin, especially Hubei Province, and prefer destinations close to home, especially local attractions. The empirical findings have significant implications for managers and policymakers in tourism and we provide potential mechanisms for these findings based on signaling, risk perception, and prospect theory.

The hospital system is a complex service system in which patients and hospitals interact and make their decisions based on bounded rationality and information. In this paper, we develop a generative agent-based model to simulate the behavior of a hospital service system. Our model combines agent-based simulation and queueing models to mimic the hospital service processes. Our goal is to simulate and understand the growth and size distribution of hospitals. This simulation model includes agents for supply elements (i.e., hospitals with different resources and expansion strategies) and demand elements (i.e., patients with different preferences for their hospital selections) in a hospital service system. Three important questions are investigated: 1) what is the emergent size distribution of hospitals? 2) what key factors influence the size distribution? and 3) how sensitive is the size distribution to these key factors? Simulation results show that the size distribution is neither power law nor lognormal. Rather, the distribution is leptokurtic, and more skewed than normal but less skewed than lognormal. This result contradicts those in extant literature on the size distributions of human and natural systems, including cities, firms, power-grids, and citation network. We conduct a set of experiments to identify the generative mechanisms for the hospital size distribution and to test the robustness of the results. The model was validated empirically by using a U.S. hospital size dataset.

This study investigates if reviewers' pattern of rating is consistent over time and predictable. Two interesting results emerge from the econometric analyses using publicly available data from TripAdvisor.com . First, reviewers' rating behavior is consistent over time and across products. Furthermore, most of the variation in their future rating behavior can be explained by their rating behavior in the past rather than by the observed average rating. Second, reviews by reviewers with higher absolute bias in rating in the past receive more helpful votes in future. We further divide the bias in rating into intrinsic bias (driven by intrinsic reviewer characteristics) and extrinsic bias (driven by influences beyond intrinsic bias) and document that intrinsic bias plays a more significant role in influencing helpful votes for reviews than extrinsic bias. Our results are robust to different product categories and different definition of bias. Overall our results indicate that in the online review context, the observed average rating or an attention grabbing strategy may not be as important as believed in the past. This study provides insights into reviewers' rating behavior and prescribes actionable items for online vendors so that they can proactively influence online opinion instead of passively responding to them.

In this study, we utilize the employee online review data from Glassdoor to examine whether stock market capitalizes the intangible asset value of employee satisfaction for high-contact service firms. We hypothesize that in the hospitality and tourism industry, employee satisfaction can efficiently motivate employees to deliver high-quality service and improve the employee retention, thereby leading to greater shareholder value. Our finding is consistent with this prediction that stock market investors indeed incorporate the intangible value of employee satisfaction into the valuation models. In addition, we find that the value of such intangible asset manifests in firm’s subsequent profitability, resulting in improved return on asset (ROA). We further decompose ROA into profit margin and asset turnover to explore the pathways by which employee satisfaction affects shareholder value. We find that employee satisfaction can improve both profit margin and asset turnover for high-contact service firms. Overall, our study suggests that employee satisfaction can be an important intangible asset that contributes to the service firms’ long-term value.

The study uses a novel psychological text-mining approach, language style matching (LSM), to examine the effect of the social interaction of top management team (TMT) members on conservative accounting reporting practices. We posit that similar language styles help to form consensus and social integration, leading to better cooperation in group decision-making. Using 10,531 firm-year observations, we show that there is a positive association between TMT LSM and accounting conservatism, suggesting that TMT members who share similar linguistic styles respond quickly to corporate news on firm value. This implies that TMTs with better language style alignment facilitate social interactions among members to form a consensus view, enhancing decision making on how a firm should react to various corporate events that affect its value. Additional analyses indicate that the impact of LSM on conservative accounting reporting behavior is more pronounced when firms face a more competitive business environment, severe information asymmetry, and higher financial distress. By examining TMT characteristics from the dynamic psycholinguistic perspective, we shed light on the understanding of how executives' cognitive values collectively affect a firm's conservative reporting decisions.

We use the system GMM estimator to analyze the size-growth relationship of 21,948 stores from 14 industries on Taobao.com.Data analysis shows that Gibrat's law tends to be rejected and the smaller stores grow faster.The applicability of the law is contingent upon store age and size: it tends to hold for larger and older stores but fails for smaller and younger ones.We explain the size-growth relationship from the viewpoints of trust building and store survival. This study examines Gibrat's law regarding size-growth relationships in the consumer-to-consumer (C2C) online marketplace. Using dynamic panel data models, we analyze 21,948 e-merchants from 14 industries on Taobao.com. The data analysis shows that Gibrat's law holds for large and mature stores when their size and age exceed certain threshold, but it generally fails to apply to stores whose size and age are below certain threshold. For those small stores, they grow faster than large ones in the C2C e-commerce. Results of the study provide insights into the competitive dynamics and industry structure of the C2C online marketplace.

This paper studies the temporal evolution of the size distribution of China's listed companies. We first identify a Pareto distribution for the upper-tail distribution. Unexpectedly, we observe that overall the Pareto coefficients decreased over the years from 2001 to 2013, which has not been reported previously in the literature. In particular, the Pareto coefficients dropped significantly during 2001 to 2008, and then fluctuated at the lowest level after 2008. A decreasing Pareto coefficient implies that the firm size inequality of the China's listed companies continuously increases during these years. By analyzing the relationship between the growth and size of firms based on a panel data model, we find that one possible reason causing the Pareto coefficients to decrease is that large firms grow faster than small ones, which is in particularly true during the non-tradable shares reform period. Furthermore, estimation results of the panel data model show that after 2008 large firms grew not as fast as they would before 2008, indicating a possible negative outcome due to the global financial crisis, which affected the growth of large firms. In addition, we examine the newly listed companies and discover that the newly listed companies with size greater than the lower bound of Pareto distribution also contribute to the decrease of the Pareto coefficients.

International audience; This study investigates how free add-on services affect customers’ perceived value in horizontal and vertical competition. We collected 349,879 reviews about over 3000 hotels in 25 U.S. cities from TripAdvisor. Using three balanced data sets generated by coarsened exact matching, the ordered logistic regressions show that free hotel add-on services (including free breakfast, parking, and WiFi) positively affect consumers’ perceived value. However, increased horizontal and vertical competition differentially weakens the positive effects of free add-on services. We not only observe a negative moderating effect of horizontal competition, but also identify three patterns of the marginal effects of these three add-ons in horizontal competition. The moderating effect of vertical competition exists from the higher-grade hotel segment to a lower-grade hotel, but such an effect is insignificant from the lower-grade hotel segment to a higher-grade hotel. Therefore, hotel managers should consider diverse external competitive environments and design appropriate differentiated service strategies.

This study examines the collective effects of voice and physician characteristics on patient satisfaction based on 35,597 voice-based medical services provided by physicians in a mobile health community. Results show patient satisfaction is positively influenced by the physician’s speech rate but negatively affected by the average spectral centroid of consultation voice. A fast speaker and speech with neural emotion are more likely to be associated with higher patient satisfaction than a slow speaker and speech with low and high emotion states. However, these effects are weak for physicians with high professional capital, which suggests a substitute role for voice characteristics.

This paper considers the competition of service centers within a service system. Each service center is modeled as a multiple-server queue. Service centers compete with each other in a fair manner and adjust their service resources (number of servers) to accommodate the demand obtained from the competition. An agent-based approach is used to model the competition and adjustment processes. Our objective is to study the size distribution of service centers under a fair competition, i.e., to answer the question, will the size distribution be uniform, normal, or skewed? It turns out that a skewed distribution (e.g., power law distribution) is obtained. Two data sets of hospital sizes are analyzed and are found to have a skewed distribution. An analytical argument is provided to explain how a fair competition could give rise to a skewed (i.e., unfair) size distribution.

This paper first gives a brief interview of AVS, especially on prediction modes of motion estimation by sub-pixel and quarter-pixel interpolation. Furthermore a new sub-pixel interpolation algorithm--EISPI based on analyzing some traditional sub-pixel interpolation algorithms is discussed in details. EISPI mainly adopts the method of threshold decision to divide the pixels into two kinds and then applies different interpolation algorithms separately. The simulation results show that EISPI can not only increase the precision of the interpolation but also save the time of the interpolation obviously.

The traditional PageRank(PR) just takes into account the web link structure, when distributing rank scores it treats all links equally, which results in topic drift. In this paper, Latent Semantic Model(LSM) is used to calculate the similarity between web pages, and the LSMPageRank(LPR) algorithm is introduced. In this algorithm, the value of parent page is distributed to the child on the basis of page similarity between them. The experiment which combines with Nutch shows that the LSMPageRank algorithm performs better than the PageRank algorithm and retrieves better result set.

Due to the lack of individual-level data, CA models that are based on individual-level theory are difficult to be validated quantitatively. In this paper, we present an approach by which we can derive the parameters of CA model from aggregate-level data so that the presented model can be used for the prediction of Internet users' quantity in the future. The result of simulation suggests that the forecasting precision of the presented model is nearly equal to the aggregate-level model (Bass model). Moreover, this model can reproduce the pattern of Internet diffusion in the aggregate-level by the interacting of individual level units, which could not be educed by the Bass model.