• Genomics & Informatics
• /
• v.20 no.3
• /
• pp.34.1-34.11
• /
• 2022

Multilevel analysis is an appropriate and powerful tool for analyzing hierarchical structure data widely applied from public health to genomic data. In practice, however, we may lose the information on multiple nesting levels in the multilevel analysis since data may fail to capture all levels of hierarchy, or the top or intermediate levels of hierarchy are ignored in the analysis. In this study, we consider a multilevel linear mixed effect model (LMM) with single imputation that can involve all data hierarchy levels in the presence of missing top or intermediate-level clusters. We evaluate and compare the performance of a multilevel LMM with single imputation with other models ignoring the data hierarchy or missing intermediate-level clusters. To this end, we applied a multilevel LMM with single imputation and other models to hierarchically structured cohort data with some intermediate levels missing and to simulated data with various cluster sizes and missing rates of intermediate-level clusters. A thorough simulation study demonstrated that an LMM with single imputation estimates fixed coefficients and variance components of a multilevel model more accurately than other models ignoring data hierarchy or missing clusters in terms of mean squared error and coverage probability. In particular, when models ignoring data hierarchy or missing clusters were applied, the variance components of random effects were overestimated. We observed similar results from the analysis of hierarchically structured cohort data.

• The Korean Journal of Applied Statistics
• /
• v.27 no.1
• /
• pp.89-102
• /
• 2014

In this study, we analyze the effect on the smoking status in the Seoul Metropolitan area using a multilevel logistic model with Community Health Survey data from the Korea Centers for Disease Control and Prevention. Intraclass correlation coefficient (ICC), profiling analysis and two types of predicted value were used to determine the appropriate multilevel analysis level. Sensitivity, specificity, percentage of correctly classified observations (PCC) and ROC curve evaluated model performance. We showed the applicability for multilevel analysis allowed for the possibility that different factors contribute to within group and between group variability using survey data.

• Communications for Statistical Applications and Methods
• /
• v.13 no.3
• /
• pp.701-718
• /
• 2006

In cancer microarray experiments, the experimenter or patient which is nested in each experimenter often shows quite heterogeneous error variability, which should be estimated for identifying a source of variation. Our study describes a Bayesian method which utilizes clinical information for identifying a set of DE genes for the class of subtypes as well as assesses and examines the experimenter effect and patient effect which is nested in each experimenter as a source of variation. We propose a Bayesian multilevel mixed effect model based on analysis of covariance (ANACOVA). The Bayesian multilevel mixed effect model is a combination of the multilevel mixed effect model and the Bayesian hierarchical model, which provides a flexible way of defining a suitable correlation structure among genes.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.31 no.3D
• /
• pp.469-476
• /
• 2011

The purposes of research is to understand the need of and the effectiveness of multilevel model to evaluate community support program in watershed areas. If the properties of policy target have hierarchical characteristics, the multilevel analysis is an adequate method to evaluate and test the effectiveness of policy. Also, the technique of multilevel modeling is extended to testing the relevance between performance appraisal and policy effectiveness. The case study of watershed region's community support program was estimated using satisfaction and economic aid level of policy target. This research has three results. First, the multilevel analysis should be used in nested data structure to estimate the effect of policy intervention. Second, the indexes of multilevel modeling should be used complementally to that of the traditional index approach. Third, the spatial hierarchical structure should be considered as the hierarchical structure in policy evaluation.

• Korean Journal of Health Education and Promotion
• /
• v.30 no.1
• /
• pp.53-64
• /
• 2013

Objectives: In this study, a multilevel analysis model has been designed to investigate the effect of personal characteristics associated with smoking cessation on anti-smoking determinants with a goal of finding out the factors which have influence on smoking cessation among the entrants of smoking cessation clinic in a public health center. Methods: A total of 253,136 male smokers who received smoking cessation services for more than six(6) months in a smoking cessation clinic of public health center from July 16, 2007 to July 15, 2008 were examined. For technical analysis, SPSS Version 2.0 has been used. For multilevel analysis on smoking cessation determinants, in addition, HLM 7.0 has been adopted. Results: According to the unconditional model of multilevel analysis, the success rates of smoking cessation among the entrants of a smoking cessation clinic were 47.3%. In an unconditional slope model test to which regional variables were added, a negative effect was observed in average smoking amount, total smoking period, nicotine dependence and services while a positive effect was found in age, stress and type of social security in terms of the log of the odds of smoking cessation. In a conditional model test, a positive effect was observed in Non-Smoking Campaigns (NSC) and Frequency of Counseling (FC) in terms of the log of the odds of smoking cessation in regional variables. Conclusions: It is important to approach smokers individually and, at the same time, build healthy environment for a local community to increase smoking cessation rates among the entrants of smoking cessation clinic in a public health center.

• Korea Journal of Hospital Management
• /
• v.20 no.1
• /
• pp.53-64
• /
• 2015

Our objective was to evaluate differences in linear regression versus multilevel(cross-level interaction model) modeling for affecting factors lumbar discectomy. The data were used in 2011 patients with HIRA sample data. Total number of analysis is 3,641 patients and 248 hospitals. The results of research model showed that the type and location of the hospital-level factors were significant. However, all factors of patient-level were similar in the two models. Therefore, it requires the selection of an appropriate model for a more accurate analysis of the influencing factors in the daily medical charge.

• Science of Emotion and Sensibility
• /
• v.19 no.4
• /
• pp.95-110
• /
• 2016

The purpose of this study was to propose a proper method for the multilevel mediation analysis, for which the hierarchical method should be utilized, then MLM (multilevel modeling) approach as a hierarchical method has been popularly utilized until MSEM (multilevel structural equation modeling) approach was not proposed. This purpose was covered by three research questions about statistical methods, analytic procedure, and real example. First, MSEM statistical method was preferred to MLM method for its estimation accuracy and analytic flexibility. Second, the four-step procedures of model building, assumption examination, model comparison, and coefficient testing were proposed for the multilevel mediation analysis. Third, the real data of 2695 students of elementary and secondary schools and 89 teachers were analyzed in the multilevel directions of $2{\rightarrow}2{\rightarrow}1$ and $1{\rightarrow}1{\rightarrow}2$. Out of these directions of $2{\rightarrow}2{\rightarrow}1$, and $1{\rightarrow}1{\rightarrow}2$ model, only the coefficient of $2{\rightarrow}2{\rightarrow}1$ model was significant at the 95% CI. Mplus programs used for the real example are attached on the Appendix. Based on the results, significance and limitations of this study, were discussed in detail.

• Research in Community and Public Health Nursing
• /
• v.28 no.4
• /
• pp.482-493
• /
• 2017

Purpose: This study aims to investigate the individual and environmental factors related to emotional/behavioral problems to early adolescence in Korea by applying multilevel modeling. Methods: From the database of the 2014 Korean Child and Youth Panel Survey (KCYPS), the researchers selected 1,977 adolescents who are in the second year of middle school. Multilevel model analysis was performed to estimate the impact of relevant factors at the individual and environmental levels. Results: At the individual level, the significant factors associated with emotional/behavioral problems included BMI and study tendency in boys, and drinking, study tendency and economic levels in girls. At the environmental level, the significant factor associated with emotional/behavioral problems included relationship with the teacher. Conclusion: The emotional/behavioral problems of early adolescence are influenced not only by the individual factors but also by the environment factor. Therefore, the environment surrounding the adolescents should also be considered to prevent emotional/behavioral problems.

• The Journal of Economics, Marketing and Management
• /
• v.10 no.5
• /
• pp.1-6
• /
• 2022

Purpose: To overcome the question that depends too much on expert's subjective judgment in traditional risk identification, this paper structure the multilevel generalized fuzzy comprehensive evaluation mathematics model of the risk identification of project, to research the risk identification of the project. Research design, data and methodology: This paper constructs the multilevel generalized fuzzy comprehensive evaluation mathematics model. Through iterative algorithm of AHP analysis, make sure the important degree of the sub project in risk analysis, then combine expert's subjective judgment with objective quantitative analysis, and distinguish the risk through identification models. Meanwhile, the concrete method of multilevel generalized fuzzy comprehensive evaluation is probed. Using the index weights to analyse project risks is discussed in detail. Results: The improved fuzzy comprehensive evaluation algorithm is proposed in the paper, at first the method of fuzzy sets core is used to optimize the fuzzy relation matrix. It improves the capability of the algorithm. Then, the method of entropy weight is used to establish weight vectors. This makes the computation process fair and open. And thereby, the uncertainty of the evaluation result brought by the subjectivity can be avoided effectively and the evaluation result becomes more objective and more reasonable. Conclusions: In this paper, we use an improved fuzzy comprehensive evaluation method to evaluate a railroad engineering project risk. It can give a more reliable result for a reference of decision making.

• The Korean Journal of Health Service Management
• /
• v.9 no.2
• /
• pp.81-88
• /
• 2015

The purpose of this study was to compare the risk-adjusted in-hospital mortality for craniotomies between logistic regression and multilevel analysis. By using patient sample data from the Health Insurance Review & Assessment Service, in-patients with a craniotomy were selected as the survey target. The sample data were collected from a total number of 2,335 patients from 90 hospitals. The sample data were analyzed with SAS 9.3. From the results of the existing logistic regression analysis and multilevel analysis, the values from the multilevel analysis represented a better model than that of logistic regression. The intra-class correlation (ICC) was 18.0%. It was found that risk-adjusted in-hospital mortality for craniotomies may vary in every hospital. The agreement by kappa coefficient between the two methods was good for the risk-adjusted in-hospital mortality for craniotomies, but the factors influencing the outcome for that were different.