Latent Class Analysis: A Guide to Best Practice

Weller, Bridget E.; Bowen, Natasha K.; Faubert, Sarah J.

doi:10.1177/0095798420930932

Cited by 756 publications

(724 citation statements)

References 52 publications

Supporting

Mentioning

697

Contrasting

Unclassified

Order By: Relevance

“…Para verificar a existência de grupos de consumidores com diferentes comportamentos em relação a COVID-19, os dados relativos às atitudes dos respondentes foram submetidos à Análise de Classes Latentes (Hagenaars e McCutcheon, 2002;Vermunt, 2010;Weller, Bowen e Faubert, 2020), utilizando-se a análise de mistura finita (FIMIX), disponível no pacote SmartPLS 2.0 M3.…”

Section: Resultsunclassified

Comportamento dos consumidores durante a pandemia de COVID-19: análise de classes latentes sobre atitudes de enfrentamento e hábitos de compra

et al. 2021

View full text Add to dashboard Cite

A COVID-19 transformou a realidade mundial impondo restrições às formas de viver, trabalhar e consumir. Poucos estudos anteriores a junho de 2020 abordaram seus impactos no comportamento dos consumidores. Esta pesquisa objetivou verificar a existência de grupos heterogêneos nas atitudes frente à pandemia e seu efeito no comportamento de compra. A abordagem foi quantitativa, utilizando escalas testadas nos contextos da SARS e H1N1, adaptadas e validadas para o contexto brasileiro. Aplicou-se a Modelagem de Equações Estruturais. Foram identificados três segmentos: ‘Céticos’ (36,7%), ‘Preocupados’ (50,1%) e ‘Indiferentes’ (13,22%). Os resultados apontam para uma diferenciação dos consumidores pelas atitudes em situações de risco percebido, papel das crenças e a consequente mudança no comportamento de compra, com implicações para o gerenciamento da saúde pública e empresarial.

show abstract

Section: Resultsunclassified

Comportamento dos consumidores durante a pandemia de COVID-19: análise de classes latentes sobre atitudes de enfrentamento e hábitos de compra

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Second, model fit indices were used to evaluate the best model: the Akaike information criterion (AIC), Bayesian information criterion (BIC), sample-adjusted Bayesian information criterion (SABIC), and consistent Akaike information criterion (CAIC). A low value for any of these criteria indicated a better model [24]. The adjusted likelihood ratio and entropy were also used to define the best model.…”

Section: Discussionmentioning

confidence: 99%

“…Indeed, the likelihood ratio provided a p-value that allowed us to determine whether one model was statistically better than another. Entropy, on the other hand, indicates how precisely the model defines the classes [24]. The "poLCA" package version 1.4.1 in R was used for latent class analysis.…”

Section: Discussionmentioning

confidence: 99%

Using Latent Class Analysis to Identify Health Lifestyle Profiles and Their Association with Suicidality among Adolescents in Benin

Hoogstoel

Fassinou

Samadoulougou

et al. 2021

IJERPH

View full text Add to dashboard Cite

Youth suicidality is considerably prevalent in low- and middle-income countries, including Benin. Factors such as psychosocial distress, socio-environmental factors, and health risk behaviors are associated with suicidality. However, little is known about how these factors co-occur in these countries. An analysis of these factors taken together would help to identify the profiles most at risk and better target prevention policies. Our study aimed to identify profiles related to these factors and their association with suicidality among adolescents in Benin. Data from the 2016 Global School-Based Student Health Survey were used, and factors related to lifestyle (tobacco and alcohol consumption and physical activity), physical violence, parental support, and psychological distress were studied. Latent class analysis was used to identify the profiles, and a modified Poisson regression with generalized estimating equations, adjusted for sociodemographic characteristics, was performed to assess the association between these profiles and suicidality. The survey results show that globally, 13.8% of the adolescents (n = 2536) aged 11 to 18 had thought about suicide, 15.6% had planned suicide, and 15.6% had attempted suicide. Four profiles were identified: a low-risk group, one with psychological distress problems, a group with violence problems, and one with alcohol, tobacco, and violence problems. The risk of suicidality, in terms of ideation, planning, or attempting, was higher for adolescents in Profiles 2, 3, and 4 than those in Profile 1 (p < 0.05). Adolescents in Profile 2 were particularly affected by this increased risk (prevalence ratio (PR) for ideation = 1.13, 95% CI = 1.03–1.23; PR for planning = 1.12, 95% CI = 1.04–1.22; PR for attempting = 1.09, 95% CI = 1.01–1.17). This study highlights the typical profiles that may be linked with suicidality among adolescents in Benin. A holistic consideration of these factors could help in planning better preventive measures to reduce suicidality among adolescents in Benin.

show abstract

“…To make the outcome significant, it is suggested to include at least 300 individuals in a Latent Class Analysis (LCA). 13 We expect the number of included individuals in our study to exceed 300 considering an inclusion period of 5 years and a high prevalence of MDD.…”

Section: Sample Size Calculationmentioning

confidence: 99%

Trajectories of somatic drug utilization patterns in Major Depressive Disorder: A Study protocol for a Danish nationwide register study using Latent Class Analysis

et al. 2021

View full text Add to dashboard Cite

Background: Major Depressive disorder (MDD) is a heterogeneous, multi-etiological disorder that is associated with chronic medical conditions and a high somatic treatment burden. A better understanding of the somatic diseases and treatment burden in MDD can be provided through a mapping of the somatic drug utilization patterns over time. The objective of this study is therefore to characterize the somatic drug profiles and their transitions over time (i.e. trajectories) among MDD patients. Methods: This descriptive study will be a nationwide register-based study including all Danish patients with an incident MDD diagnosis between 2011 and 2015. Using Latent Class Analysis, we will identify homogenous MDD patient subgroups according to somatic drug utilization (i.e. drug profiles). The development in somatic drug profiles will be depicted in four different time intervals from three years prior to the MDD diagnosis to three years after the diagnosis. Patients will be assigned to the latent class (drug profile) to which they have the highest probability of belonging using modal assignment. The treatment trajectories will be performed by cross tabulating these assignments. Discussion: Profiles and trajectories of somatic drug use will provide a new perspective on patterns of somatic drug burden in MDD patients. Moreover, identifying homogenous subgroups of MDD patients regarding somatic drug use can contribute to a deeper understanding of MDD etiology. In the future, this knowledge could help optimizing MDD treatment by studying if different antidepressants will show different efficacy and safety depending on the profiles and trajectories of somatic diseases.

show abstract

Latent Class Analysis: A Guide to Best Practice

Cited by 756 publications

References 52 publications

Comportamento dos consumidores durante a pandemia de COVID-19: análise de classes latentes sobre atitudes de enfrentamento e hábitos de compra

Comportamento dos consumidores durante a pandemia de COVID-19: análise de classes latentes sobre atitudes de enfrentamento e hábitos de compra

Using Latent Class Analysis to Identify Health Lifestyle Profiles and Their Association with Suicidality among Adolescents in Benin

Trajectories of somatic drug utilization patterns in Major Depressive Disorder: A Study protocol for a Danish nationwide register study using Latent Class Analysis

Contact Info

Product

Resources

About