• Industrial Engineering and Management Systems
• /
• v.11 no.4
• /
• pp.371-384
• /
• 2012

This study proposes a multi-level data analysis approach to identify both superficial and latent relationships among variables in the data set obtained from a vocational rehabilitation (VR) services program of people with significant disabilities. At the first layer, data mining and statistical predictive models are used to extract the superficial relationships between dependent and independent variables. To supplement the findings and relationships from the analysis at the first layer, association rule mining algorithms at the second layer are employed to extract additional sets of interesting associative relationships among variables. Finally, nonlinear nonparametric canonical correlation analysis (NLCCA) along with clustering algorithm is employed to identify latent nonlinear relationships. Experimental outputs validate the usefulness of the proposed approach. In particular, the identified latent relationship indicates that disability types (i.e., physical and mental) and severity (i.e., severe, most severe, not severe) have a significant impact on the levels of self-esteem and self-confidence of people with disabilities. The identified superficial and latent relationships can be used to train education program designers and policy developers to maximize the outcomes of VR training programs.

• Journal of the Korean Statistical Society
• /
• v.23 no.2
• /
• pp.463-482
• /
• 1994

In Bayesian analysis of randomized response models, the likelihood function does not combine tractably with typical priors for the parameters of interest, causing computational difficulties in posterior analysis of the parameters of interest. In this article, the difficulties are solved by introducing appropriate latent variables to the model and using the Gibbs sampling algorithm.

• The Korean Journal of Applied Statistics
• /
• v.28 no.1
• /
• pp.61-74
• /
• 2015

Structural equation modeling (SEM) is a widely used in social sciences such as education, business administration, and psychology. In SEM, the latent variable score is the estimate of the latent variable which cannot be observed directly. This study uses stagewise structural equation modeling(stagewise SEM; SSEM) by partitioning the whole model into several stages. The traditional estimation method minimizes the discrepancy function using the variance-covariance of all observed variables. This method can lead to inappropriate situations where exogenous latent variables may be affected by endogenous latent variables. The SSEM approach can avoid such situations and reduce the complexity of the whole SEM in estimating parameters.

• Communications for Statistical Applications and Methods
• /
• v.8 no.3
• /
• pp.777-786
• /
• 2001

The autoregressive models have been used to describe a wade variety of time series. Then the problem of determining the order in the times series model is very important in data analysis. We consider the Bayesian approach for finding the order of autoregressive(AR) error models using the latent variable which is motivated by Tanner and Wong(1987). The latent variables are combined with the coefficient parameters and the sequential steps are proposed to set up the prior of the latent variables. Markov chain Monte Carlo method(Gibbs sampler and Metropolis-Hasting algorithm) is used in order to overcome the difficulties of Bayesian computations. Three examples including AR(3) error model are presented to illustrate our proposed methodology.

• Survey Research
• /
• v.12 no.2
• /
• pp.113-144
• /
• 2011

The purposes of the present study were (1) to identify the latent classes depending on youth employees' trajectories in job satisfaction and turnover intention and (2) to test the effects of person-job fit(major fit, education level fit, skill level fit) on job satisfaction and turnover intention using Youth Panel 2001. In order to estimate latent classes of job satisfaction and turnover intention changes simultaneously and study probabilities linking latent class membership in trajectory across the two variables, we applied dual trajectory model, an extension of semi-parametric group-based approach, Results showed that four latent classes were identified for job satisfaction, which were defined, based on the trajectory patterns, as increasing group, decreasing group, medium-level group, and high-level group. And, three latent classes estimated for turnover intention were defined as low-level group, maintaining group, and rapidly decreasing group. To test the effects of person-job fit variables, we added the variables as time-dependant variables to the unconditional latent class model. The effect of education level fit and skill level fit were found significant in the groups which are low in job satisfaction and have high in turnover intention. Findings from this study suggest the need to consider trajectory heterogeneity in the study of youth employees' job satisfaction and turnover intention to capture the dynamic dimension of overlap between the two constructs.

• Environmental and Resource Economics Review
• /
• v.16 no.3
• /
• pp.701-721
• /
• 2007

A key issue in the development and application of stated preference nonmarket valuation is the incorporation of unobserved heterogeneity in utility models. Two approaches to this task have dominated. The first is to include individual-specific characteristics into the estimated indirect utility functions. These characteristics are usually socioeconomic or demographic variables. The second employs generalized models such as random parameter logit or probit models to allow model parameters to vary across individuals. This paper examines a third approach: the inclusion of psychological or 'latent' variables such as general attitudes and behaviour-specific attitudes to account for heterogeneity in models of stated preferences. Attitudinal indicators are used as explanatory variables and as segmentation criteria in a choice modelling application. Results show that both the model significance and parameter estimates are influenced by the inclusion of the latent variables, and that attitudinal variables are significant factors for WTP estimates.

• Korean Chemical Engineering Research
• /
• v.48 no.4
• /
• pp.475-482
• /
• 2010

In this work, soft sensor based on image anlaysis is proposed for quantitatively estimating the visual appearance of manufactured products and is applied to quality monitoring. The methodology consists of three steps; (1) textural feature extraction from product images using wavelet transform, (2) numerical estimation of the product appearance through projection of the textural features on subspace, and (3) use of latent variables of textural features (i.e., numerical estimates of product appearance). The focus of this approach is on the consistent and quantitative estimation of continuous variations in visual appearance rather than on classification into discrete classes. This approach is illustrated through the application to the estimation and monitoring of the appearance of engineered stone countertops.

• Journal of rehabilitation welfare engineering & assistive technology
• /
• v.6 no.1
• /
• pp.83-87
• /
• 2012

This study is aimed to identify latent classes which are based the change patterns in assistive technology device use among worker with disabilities and to test the effects of independent variables(gender, education, disability type, disability density, activity and participation of ICF: ICF, subjective socioeconomic status: SES, job satisfaction, life satisfaction) on determining latents classes. This study applied Nagin's(1999) semi-parametric group based approach to the panel survey of employment for the disabled. Because dependant variable has dichotomous scale, logit model was used. The results identified three latent classes, which could be defined based on the patterns as follows; assistive device continued use group, assistive device mid-level use group, assistive device sharp decline use group. The effects of the independent variables on the latent classes was tested by multinomial logit analysis. The results showed that education, disability type, ICF, SES, and life satisfaction were significant determinants of the latent classes. Finally, the implications based on analysis results were suggested.

• Survey Research
• /
• v.10 no.3
• /
• pp.37-58
• /
• 2009

The purpose of this study were to identify the subgroups (i.e., latent classes) depending on early adolescents' change patterns in aggression and to test the effects of individual-background variables on determining the latent classes. For these goals, we applied Nagin's(1999) semi-parametric group-based approach to the Korean Youth Panel Study. Results showed that four latent classes were identified, which could be defined based on the patterns as low-level group, increasing group, intermediate-level group, and high-level group. By adding gender, self-control, parent attachment, teacher attachment, and the number of delinquent friends to the unconditional latent class model, we tested the effects of the variables on the latent classes. Multinomial logit analysis showed that gender, self-control, teacher attachment, and the number of delinquent friends were significant determinants of the latent classes. Findings from this study suggest the need to consider heterogeneity in the study of early adolescents' aggression to facilitate more refined targeting of intervention program.

• Journal of the Korean Society of Safety
• /
• v.37 no.2
• /
• pp.18-27
• /
• 2022

Construction accidents are difficult to prevent because several different types of activities occur simultaneously. The current method of accident analysis only indicates the number of occurrences for one or two variables and accidents have not reduced as a result of safety measures that focus solely on individual variables. Even if accident data is analyzed to establish appropriate safety measures, it is difficult to derive significant results due to a large number of data variables, elements, and qualitative records. In this study, in order to simplify the analysis and approach this complex problem logically, data preprocessing techniques, such as latent class cluster analysis (LCCA) and predictor importance were used to discover the most influential variables. Finally, the correlation was analyzed using an alluvial flow diagram consisting of seven variables and fourteen elements based on accident data. The alluvial diagram analysis using reduced variables and elements enabled the identification of accident trends into four categories. The findings of this study demonstrate that complex and diverse construction accident data can yield relevant analysis results, assisting in the prevention of accidents.