Latent Class Cluster Analysis Research Articles

Is tourist accommodation homogeneous? An analysis of the adults-only category through a three-step approach

PurposeThis paper aims to classify tourist accommodation using data from Booking.com and TripAdvisor and analyse the extent to which the different segments identified differ in terms of being adults-only.Design/methodology/approachIn total, 1,535 properties located in nine Spanish sun and beach destinations were examined using a latent class cluster analysis (LCCA). The bias-adjusted three-step approach was used to investigate the differences between belonging to adults-only accommodation or not among the identified clusters.FindingsResults show that adults-only accommodation tends to belong to the cluster with higher online ratings. In small Spanish islands, adults-only hotels account for a large share (more than 25%) of hotels.Research limitations/implicationsIt was not possible to analyse whether the higher rating was due to the accommodation being better or due to the tourists being more satisfied with their stay.Practical implicationsIn urban destinations, the model is not widely used. However, in coastal destinations, it is becoming more than a novelty or a new trend.Social implicationsIn small Spanish islands, people traveling with children are becoming a minority. Families may feel discriminated against and express dissatisfaction with this situation in the future.Originality/valueThis study covers the gap in the academic literature on this growing hotel segment.

Moral foundations in gender violence cases decided in Portuguese courts

Gender violence encompasses a multitude of morally problematic psychological, physical, and sexual behaviors that, in most countries, constitute criminal offenses. In this study, we investigate the association between moral foundations ( Care, Fairness, Loyalty, Authority, and Sanctity) and punitive responses to gender violence offenses. Our case study focuses on gender violence in Portugal, a country in which these offenses are a prevalent social problem. We collected data on gender violence legal cases decided in Portuguese courts between 2002 and 2022, and we used a latent class cluster analysis model to identify the complex patterns in the data and reduce such patterns to a distinct number of clusters. Four main clusters unravel latent relations between the foundations mapped in the legal narratives and corresponding punitive responses: (i) Affirmative with suspended prison time (moral rhetoric rooted in Authority); (ii) Mixed outcomes but no prison time (moral rhetoric rooted in Sanctity); (iii) Affirmative with lengthy prison time large compensation (moral rhetoric rooted in Loyalty and Care); and (iv) Affirmative with court fines (moral rhetoric rooted in Fairness). The moral foundations provide a valuable lens to understand the problem of gender violence, but further research is needed to establish the causal mechanisms between morality and punitive responses to gender violence.

Pleasure, quality or status? an analysis of drivers of purchase of fresh pork in China

What are consumers aiming to get when they buy fresh meat? Is it the emotional pleasure, the nutritional quality and functionality, or the status that goes with it? We examine this question for Chinese consumers buying fresh pork. In order to understand the driving forces for pork purchases, we use the concept of perceived value of a product, and distinguish emotional value, quality/performance and social value, together with price/value for money. We look at how perceived value of pork products in China is related to consumers’ attitude to these products and to their repeated purchase of these products. In addition, we look at how value perception and its role in determining attitude and purchase behavior differ between different consumer segments, distinguished based on their shopping behavior. An online survey was carried out in 5 Chinese 1st and 2nd tier cities. Respondents were segmented based on their usage of different ways of shopping for pork using latent class cluster analysis. Relationships between constructs were estimated using PLS. Quality/functional value was the strongest determinant of attitude, but emotional value was both the strongest direct determinant of purchase behavioral and the strongest determinant overall when taking effects mediated by attitude into account. Customer journey segments differed in their pattern of determinants of attitudinal and behavioral loyalty. We conclude that buying fresh meat is, for Chinese consumers, mostly driven by the expected pleasure and to a lesser degree by quality and functional properties like safety and healthiness. The latter do have an impact on consumers’ attitude to the product, but less so on their buying behavior, suggesting that attitude and purchase are driven by different mental processes. We discuss implications for future demand for fresh pork in China.

Analysis of factors influencing the degree of accidental injury of bicycle riders considering data heterogeneity and imbalance.

Bicycle safety has emerged as a pressing concern within the vulnerable transportation community. Numerous studies have been conducted to identify the significant factors that contribute to the severity of cyclist injuries, yet the findings have been subject to uncertainty due to unobserved heterogeneity and class imbalance. This research aims to address these issues by developing a model to examine the impact of key factors on cyclist injury severity, accounting for data heterogeneity and imbalance. To incorporate unobserved heterogeneity, a total of 3,895 bicycle accidents were categorized into three homogeneous sub-accident clusters using Latent Class Cluster Analysis (LCA). Additionally, five over-sampling techniques were employed to mitigate the effects of data imbalance in each accident cluster category. Subsequently, Bayesian Network (BN) structure learning algorithms were utilized to construct 32 BN models after pairing the accident data from the four accident cluster types before and after sampling. The optimal BN models for each accident cluster type provided insights into the key factors associated with cyclist injury severity. The results indicate that the key factors influencing serious cyclist injuries vary heterogeneously across different accident clusters. Female cyclists, adverse weather conditions such as rain and snow, and off-peak periods were identified as key factors in several subclasses of accident clusters. Conversely, factors such as the week of the accident, characteristics of the trafficway, the season, drivers failing to yield to the right-of-way, distracted cyclists, and years of driving experience were found to be key factors in only one subcluster of accident clusters. Additionally, factors such as the time of the crash, gender of the cyclist, and weather conditions exhibit varying levels of heterogeneity across different accident clusters, and in some cases, exhibit opposing effects.

Public preferences for the allocation of societal resources over different healthcare purposes

ObjectiveIncreasing healthcare expenditures require governments to make difficult prioritization decisions. Considering public preferences can help raise citizens’ support. Previous research has predominantly elicited preferences for the allocation of public resources towards specific treatments or patient groups and principles for resource allocation. This study contributes by examining public preferences for budget allocation over various healthcare purposes in the Netherlands. MethodsWe conducted a Participatory Value Evaluation (PVE) choice experiment in which 1408 respondents were asked to allocate a hypothetical budget over eight healthcare purposes: general practice and other easily accessible healthcare, hospital care, elderly care, disability care, mental healthcare, preventive care by encouragement, preventive care by discouragement, and new and better medicines. A default expenditure was set for each healthcare purpose, based on current expenditures. Respondents could adjust these default expenditures using sliders and were presented with the implications of their adjustments on health and well-being outcomes, the economy, and the healthcare premium. As a constraint, the maximum increase in the mandatory healthcare premium for adult citizens was €600 per year. The data were analysed using descriptive statistics and a Latent Class Cluster Analysis (LCCA). ResultsOn average, respondents preferred to increase total expenditures on all healthcare purposes, but especially on elderly care, new and better medicines, and mental healthcare. Three preference clusters were identified. The largest cluster preferred modest increases in expenditures, the second a much higher increase of expenditures, and the smallest favouring a substantial reduction of the healthcare premium by decreasing the expenditure on all healthcare purposes. The analyses also demonstrated substantial preference heterogeneity between clusters for budget allocation over different healthcare purposes. ConclusionsThe results of this choice experiment show that most citizens in the Netherlands support increasing healthcare expenditures. However, substantial heterogeneity was identified in preferences for healthcare purposes to prioritize. Considering these preferences may increase public support for prioritization decisions.

Identifying heterogeneity in university students’ transport mode choice

College students are argued to have some interest in adopting alternative mobility solutions. Even though previous works have examined the transport mode choice of college students, the inclusion of shared mobility services is scarcely found. Thus, this study aims to examine the transport mode choice of university students by revealing the heterogeneity in transport mode usage frequency incorporating socio-demographic and travel characteristics as covariates in the latent class cluster analysis (LCCA) framework. The analysis results in four clusters. Clusters of public transport users and walkers are mostly characterized by lower-income students who can reach the university in less than 11 minutes. Private car users mostly fall into a cluster where students generally come from higher-income families. Policies supporting sustainable transportation among college students are discussed, including the mobility points around the university and public transport stops.

Assessing Climate Disaster Vulnerability in Peru and Colombia Using Street View Imagery: A Pilot Study

Community and household vulnerability to natural hazards, e.g., earthquakes, hurricanes, and floods, is a concern that transcends geographic and economic boundaries. Despite the abundance of research in this field, most existing methods remain inefficient and face the challenge of data scarcity. By formulating and investigating the correlation between the household vulnerability and street view images of buildings, this research seeks to bridge the knowledge gap to enable an efficient assessment. Especially in developing countries, the widespread prevalence of outdated or inadequately enforced building codes poses a significant challenge. Consequently, a considerable portion of the housing stock in these regions fails to meet acceptable standards, rendering it highly vulnerable to natural hazards and climate-related events. Evaluating housing quality is crucial for informing public policies and private investments. However, current assessment methods are often time-consuming and costly. To address this issue, we propose the development of a rapid and reliable evaluation framework that is also cost-efficient. The framework employs a low-cost street view imagery procedure combined with deep learning to automatically extract building information to assist in identifying housing characteristics. We then test its potential for scalability and higher-level reliability. More importantly, we aim to quantify household vulnerability based on street view imagery. Household vulnerability is typically assessed through traditional means like surveys or census data; however, these sources can be costly and may not reflect the most current information. We have developed an index that effectively captures the most detailed data available at both the housing unit and household level. This index serves as a comprehensive representation, enabling us to evaluate the feasibility of utilizing our model’s predictions to estimate vulnerability conditions in specific areas while optimizing costs. Through latent class clustering and ANOVA analysis, we have discovered a strong correlation between the predictions derived from the images and the household vulnerability index. This correlation will potentially enable large-scale, cost-effective evaluation of household vulnerability using only street view images.

Sleep patterns according to genetically determined ethnicity in the population of São Paulo, Brazil

Sleep is a behavior expressed differently for each individual. However, studies have shown that some ethnic groups express common sleep patterns, which can be observed in different ethnic groups. Previous studies have shown the existence of sleep disparities in populations of different ethnicities. Most of these studies have considered self-reported ethnicity and assessed sleep subjectively. Therefore, the aim of this study was to evaluate sleep disparities in different ethnic groups based on an analysis of genetic ancestry and the use of objective sleep evaluation. To do this, we used data from the São Paulo Epidemiologic Sleep Study (EPISONO), which was undertaken in Brazil, a country that is known for its ethnic/racial diversity. All individuals completed a series of questionnaires, underwent full polysomnography and had their blood collected for DNA extraction. After genotyping and identifying samples with high-quality DNA suitable for genetic analysis, 31 ancestry-informative markers (AIMs) were selected. These markers exhibited substantial allelic frequency differences, enabling the characterization of the three primary founding populations of modern Brazil - Europeans, West-Africans, and Native Americans. Through this analysis, the genetic contribution of each of these ancestral groups was identified in respect of each participant. Based on this, a latent class cluster analysis (LCCA) was performed to define the three clusters that best classified the sample according to ethnic group: African (n = 255), Caucasian (n = 668) and Native American (n = 83). Applying the adjusted model for the confounding variables (age, socio-economic class and sex), statistically significant differences in sleep variables between ethnicities were found. Africans had higher sleep latency compared to the other groups (β = 4.46, CI = 1.18 to 7.74 and β = 7.83, CI = 3.50 to 12.15), while Caucasians had longer total sleep time (β = −16.47, CI = −29.94 to −2.99) and better sleep efficiency (β = −2.19, CI = −4.35 to −0.02) compared to Africans. Regarding the respiratory arousals index (β = −1.11, IC = −2.07 to −0.16) and periodic leg movements index (β = −7.48, CI = −12.08 to −2.88), both were higher among Caucasians compared to Africans. We were able to conclude that genetic ancestry might modulate sleep structure and the occurrence of sleep disorders.

Examining TOD node typology using k-means, hierarchical, and latent class cluster analysis for a developing country

Examining TOD node typology using k-means, hierarchical, and latent class cluster analysis for a developing country

Job quality and precarious employment among lesbian, gay, and bisexual workers: A national study

BackgroundEmployment outcomes among sexual minority (i.e., lesbian, gay, bisexual) workers are poorly understood, and previous research on this topic has focused almost exclusively on inequities in earnings, neglecting other important dimensions of job quality. We address this gap by describing and comparing the job quality of straight and sexual minority workers in Canada. MethodsData are from the 2016 General Social Survey: Canadians at Work and Home, the only national survey providing both a measure of sexual orientation and a multidimensional view of job quality in Canada. We identified 25 unique job quality indicators (e.g., temporary employment; job insecurity; health benefits; low income; job satisfaction; job control; discrimination). Latent class cluster analysis was used to establish a typology of job quality describing standard, flexible, and precarious employment types. We used multivariable regression methods to examine the association between sexual orientation and job quality. ResultsSexual minorities reported lower job quality than their straight counterparts along many dimensions, with bisexual people reporting the lowest job quality. While inequities were generally observed among both sexual minority men and women, they sometimes differed in magnitude by gender. The prevalence of precarious employment was nearly three times higher among lesbian, gay, and bisexual workers (PR: 2.94, CI: 1.89–4.58 among all sexual minorities; PR: 3.04, CI: 1.71–5.43 among gay/lesbian workers; and PR: 2.81, CI: 1.45–5.47 among bisexual workers) compared to their straight counterparts. ConclusionInequities in job quality among sexual minorities persist despite comprehensive human rights protections in Canada. These inequities are pervasive, extending well beyond conventional indicators such as dollars earned and hours worked. Multi-pronged interventions are needed that move beyond simply ensuring that sexual minority workers are employed. Sexual minority workers deserve access to secure, well-paid work with benefits where they can foster connection and be free from discrimination.

Consumers' Preferences towards Bread Characteristics Based on Food-Related Lifestyles: Insights from Slovenia.

Consumers' recognition and understanding of food characteristics can have an important role when making purchase decisions. The current study analysed consumer preferences for bread, an important food in the diets of Central European countries. The study included a conjoint experiment on a representative sample of 547 adult consumers in Slovenia. The following bread attributes: functional ingredients (chia seeds, linseed, quinoa, and Tartary buckwheat); nutritional claims (low salt, high fibre, and high protein); and other claims (organic, free from additives, flour from Slovenia, and wholegrain) were studied. The results showed the strongest relative importance for functional ingredients (a mean relative importance of 83.9%). In addition, a deeper insight into consumer preference was investigated by a recently developed modular instrument for food-related lifestyles. Latent class cluster analysis (LCA) enabled the identification of four consumer segments (uninvolved, conservative, health-conscious, and moderate) with different preferences toward selected functional ingredients, nutrition, and other claims. The results provide insights that allow for a better understanding of consumer preferences for functional ingredients and claims, and new perspectives for bread marketing to different consumer segments based on food-related lifestyles. Identifying the drivers that affect bread purchasing and consumption can support reformulation activities and product promotion in the direction of reinforcing healthier food choices.

Persona Design Methodology for Work-Commute Travel Behaviour Using Latent Class Cluster Analysis

The present study proposes a new methodology that combines quantitative and qualitative data for the generation of representative personas for commuters. The profiles can be used to better understand their travel behaviour and mode choices. The research is based on the example of the region of Agder in Norway and aims to overcome the persona development shortcomings identified by previous researchers. Data from a regional travel behaviour survey (N= 1 849) is analysed using latent class cluster analysis (LCCA), and enriched with qualitative input from 32 interviews, and information provided by an expert panel. This results in a set of 20 representative persona profiles for the case study region. The proposed methodology is easily replicable in other urban networks and has the potential to provide insight into the mobility behaviour and needs of specific groups of people in order to adapt the transport services and encourage climate-friendly behaviour.

Association Between Diabetes and Personality Traits Among the Elderly in China: A Latent Class Analysis.

The present study aimed to identify individuals with different personalities using latent class analysis and further distinguish those with a high risk of diabetes among different clusters. Data were utilized from a large-scale, cross-sectional epidemiological survey conducted in 2018 across 23 provinces in China, employing a multi-stage, stratified sampling technique. Latent class cluster analysis was performed to identify distinct personality clusters based on a series of variables concerning life attitudes. Logistic regression was used to calculate adjusted odds ratios (AORs) after controlling for potential confounding variables, including age, gender, body mass index, smoking status, alcohol consumption, hypertension, and physical activity levels, to determine the association between these groups and diabetes. Four distinct personality clusters were identified, namely the energy-poor (2.0%), self-domination (61.3%), optimistic (21.3%), and irritable (15.4%) groups. The prevalence of diabetes in these groups was 14.6%, 9.7%, 9.3%, and 11.6%, respectively. After adjusting for potential confounders, the "energy-poor group" exhibited more odds of having diabetes as compared to the "optimistic group" (AOR 1.683, 95%CI: 1.052-2.693; P=0.030). This study identified an energy-poor group of individuals with a high risk of diabetes. Targeted interventions should consider the emotional and personality characteristics of the elderly.

Segmenting citizens according to their self-sufficiency: A tool for local government

Identifying subgroups of citizens with varying levels of self-sufficiency in a large local or regional population provides local government with essential input for providing matching services and well-grounded spending of health and well-being expenditures. This paper identifies self-sufficiency levels of citizens by segmenting a broad adult population. We used data from a citizen survey based on a randomly selected response group containing questions on a wide range of topics, including finances, health and living conditions, and complemented these data with registration data, including information on housing type and household composition. We conducted a latent class cluster analysis using six indicators: perception of making ends meet, perceived health, quality of life, self-efficacy, access to socialsupport and social network. High scores on the indicators translate to high levels of self-sufficiency. We used a biased-adjusted, three-step approach to characterise the segments. Six meaningful segments were identified and labelled as ‘highly self-sufficient,’ ‘self-sufficient – medium access to social support,’ ‘self-sufficient – medium self-efficacy,’ ‘moderately self-sufficient – low self-efficacy & high social network,’ ‘moderately self-sufficient – low access to social support/social network & high perceived health’ and ‘not self-sufficient.’ At a macro level, perception of making ends meet and quality of life have discriminating value in assessing self-sufficiency. For a more detailed differentiation between groups with similar levels of self-sufficiency, perceived health, self-efficacy, access to socialsupport, and social network are valuable indicators. Overall, this study introduces a comprehensive tool to assess self-sufficiency in larger groups of citizens by using a parsimonious number of indicators. Local and regional governments can apply this tool to effectively assess the self-sufficiency levels of their population and signal potentially vulnerable groups. In this way, the tool makes the identification of self-sufficiency levels of larger populations more feasible and more efficient and can be widely adopted in different contexts.

Food-related consumer behaviours in times of crisis: Changes in the wake of the Ukraine war, rising prices and the aftermath of the COVID-19 pandemic

When the COVID-19 pandemic subsided, the war in Ukraine led to further disruptions in consumers’ daily behaviours, with rising prices for food and energy. We conducted a survey study on self-reported changes in food-related consumer behaviour in ten European countries and compared the results to a similar study conducted two years ago. A latent class cluster analysis distinguished five clusters and showed that different types of consumers can be distinguished based on how they react to the crisis as regards their eating habits. 19% of survey participants reported no major changes, and 32% reported changes mostly in terms of more price sensitivity. Among those that reported changes beyond reacting to higher prices, there are indications of more mindful eating and more deliberate choices. The changes already found earlier in response to the COVID-19 pandemic therefore seem to have been strengthened and supplemented by reactions to price increases. The results present a challenge to the food industry in terms of supplying healthy and sustainable food at affordable prices.

Australian ethical consumer segmentation of food and agricultural products and export opportunities for the Pacific Islands

Abstract Ethical consumerism is one of this century’s retail success stories and has continued to grow despite the global pandemic and economic contraction. Consumers worldwide are increasingly putting their money where their values are in purchases of food and drink, clothing, housing, and financial and other services. Understanding these consumers is key to competing successfully in markets for ethically sourced products and presents opportunities for reducing poverty in developing countries where food and agricultural production is the dominant economic activity, such as small island developing states (SIDS) in the Pacific. While ethical consumerism is an expanding research field, findings rarely are analysed in detail for market development. Rarer still is segmentation in terms of product attributes aimed explicitly at poverty reduction in slow-growth, emerging economies like those of the Pacific. Here we report a latent class cluster analysis of survey responses by ethically conscious Australian consumers (n = 637) associated with Oxfam, the development charity. Four distinct consumer segments were identified with implications for Pacific producers. Product origin and ‘the story behind the product’ were particularly important to these consumers. The findings provide practical insights to agricultural stakeholders in the Pacific and those similarly situated around the globe that are in need of new economic opportunities.

Segmenting and predicting prosocial behaviours among tourists: a latent class approach

ABSTRACT In building sustainable post-pandemic destinations, it is critical to understand the typologies of tourists’ prosocial behaviours. Consequently, this study innovatively applied a latent class cluster analysis to segment the prosocial behaviours of 403 Macau tourists. Three ordered discrete segments were derived based on consistent tourists’ probabilities of performing prosocial behaviours on the trip namely: the Self-centred, the Intermediate, and the Philanthropist. The associated ordered logistic regression predicting the segments revealed that relative to the Self-centred, the Intermediate and the Philanthropist are more likely to face death terror, are sociable – seek vacation friends – and believe in cultural and heritage conservation. Not only does this research expand the theoretical application of Terror Management Theory, the Scrooge effect, and the self-esteem concept, it contributes to prosocial alternative tourism with novel destination management implications for marketing and promoting prosocial tourism performance.

Identifying profiles of ride-sourcing users in the Metro Vancouver Region for a better understanding of ride-sourcing behaviour

Ride-sourcing has the potential to attract demand from more sustainable modes, alter activity patterns, and worsen congestion; however, it can also help increase the mobility and accessibility of those who cannot drive. Understanding how and why these services are used can help inform policies that address the negative externalities of these services while taking advantage of their benefits. This study examines the use of ride-sourcing services, which were introduced just before the COVID-19 pandemic, in the Metro Vancouver Region. Using data from a web-based survey, this study applies latent class cluster analysis to identify profiles of ride-sourcing users based on the purposes for which they use ride-sourcing. To the authors’ knowledge, this is only the second study to identify latent segments of ride-sourcing users and the first to do so based on the purposes of their ride-sourcing trips. Additionally, the decision to use ride-sourcing and the profile to which an individual belongs are jointly modelled using a two-stage multinomial logistic regression model. The results suggest the impacts of ride-sourcing use on other modes and the frequency of ride-sourcing use may differ based on the profile to which the user belongs. Besides, the findings of this study highlight the influence of socio-demographic attributes, the use of other shared mobility services, and attitudinal factors on the utilization of ride-sourcing. This study aims to contribute to the literature by offering further insights into the heterogeneity among ride-sourcing users, which can help inform a more targeted approach to addressing the negative externalities of these services.

Heterogeneous attitudes and behaviors in relation to participation in the ecotourism: Do customer segments play a role?

The potential of ecotourism is expanding year on year. Given the broad nature of the tourism market, it is necessary to identify smaller groups of individuals that have common traits. This study aims to explore if there are any segments of ecotourist based on attitude and behavior towards ecotourism, profiles the segments according to their sociodemographic characteristics, and analyzes the extent to which they differ by country of origin. The study used a sample of 809 ecotourists from Spain and Chile, while latent class cluster analysis was employed as the segmentation method. The ecotourists were divided into three segments (Basic ecotourists—ecotourists of thought—, Average ecotourists, and Conscientious and active ecotourists) and the bias-adjusted three-step methodology was used to examine the differences in the country of origin. In sum, the largest portion of tourists falls into the category of the average ecotourist segment, which lies between the two aforementioned segments and exhibits moderate attitudes and intentions to engage in ecotourism and pay a premium for it. This segment is more prevalent in Spain compared to Chile. This research adds to the body of knowledge on the marketing of ecotourism behavior by modeling the heterogeneity in the ecotourists’ attitudes and behaviors. This study suggests that ecotourism organizations, whether they are accommodation providers, tourism operators, or companies that organize activities in nature, should be aware of the unique qualities of their customers, as well as their preferences, in order to create successful marketing strategies.

Exploring the heterogeneities of factors affecting e-bicyclist severities in delivery e-scooter crashes

Delivery e-bicyclists are vulnerable traffic participants, and delivery e-scooter crashes annually cause great losses. There are limited empirical analyses performed to characterize the endogenous and exogenous influences of potential factors on delivery e-scooter crash severities. This study focused on identifying the heterogeneous impacts of contributing factors. Based on 3064 police-reported delivery e-scooter crashes, different potential factors selected from seven perspectives were examined. Combining the latent class clustering analysis and the partial proportional odds model, an unsupervised learning approach was developed to modeling the heterogeneities as well as the corresponding marginal effects. According to the AIC, BIC, and entropy-based approach, the most appropriate number of clusters was identified as three. The model results revealed that the top six factors significantly influencing the injury severities were running red light, reverse riding, speeding, bike lane, changing lane, and e-bicyclist age. Their maximum absolute values of the marginal effects were more than 29%. The sub-models owned a better goodness of fit in characterizing heterogeneities and avoiding the model bias. It was easy to find the remarkable heterogeneities within each factor and across different clusters. The findings could provide insightful guidance and effective recommendations for alleviating the crash losses and promoting rider safety.