• Communications for Statistical Applications and Methods
• /
• v.9 no.3
• /
• pp.703-713
• /
• 2002

In longitudinal studies, outcomes are repeatedly measured over time for each subject. It is common to have missing values or dropouts for longitudinal data. In this study time trend in longitudinal data with dropouts is of concern. The confounding of time trend with dropout process is investigated through simulation studies. Some simulation results are reported for binary responses as well as continuous responses with patterns of dropouts varying. It has been found that time trend is not confounded with random dropout process for binary responses when it is estimated using GEE.

• Journal of the Korean Statistical Society
• /
• v.26 no.3
• /
• pp.319-332
• /
• 1997

The application of multivariate linear rank statistics to data with item nonresponse is considered. Only a modest extension of the complete data techniques is required when the missing data may be thought of as a random sample, and an appropriate modification of the covariances is derived. A proof of the asymptotic multivariate normality is given. A review of some related results in the literature is presented and applications including longitudinal and repeated measures designs are discussed.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.2
• /
• pp.385-392
• /
• 2012

Logistic regression is a well known binary classification method in the field of statistical learning. Mixed-effect regression models are widely used for the analysis of correlated data such as those found in longitudinal studies. We consider kernel extensions with semiparametric fixed effects and parametric random effects for the logistic regression. The estimation is performed through the penalized likelihood method based on kernel trick, and our focus is on the efficient computation and the effective hyperparameter selection. For the selection of optimal hyperparameters, cross-validation techniques are employed. Numerical results are then presented to indicate the performance of the proposed procedure.

• Journal of the Korean Data and Information Science Society
• /
• v.25 no.6
• /
• pp.1263-1272
• /
• 2014

Longitudinal categorical data commonly occur from medical, health, and social sciences. In these data, the correlation of repeated outcomes is taken into account to explain the effects of covariates exactly. In this paper, we introduce marginalized random effects models that are used for the estimation of the population-averaged effects of covariates. We also review how these models have been developed. Real data analysis is presented using the marginalized random effects.

• Journal of Child Welfare and Development
• /
• v.16 no.2
• /
• pp.109-135
• /
• 2018

The purpose of this study was to analyze the longitudinal causal relationship between caregiver relations and peer relations of children in out-of-home care. We analyzed the three years(2011-2013) of longitudinal data from the Panel Study on Korean Children in Out-of-Home Care. The autoregressive cross-lagged model (ARCL) was used to measure the longitudinal causal relationship between caregiver relations and peer relations. As a result, first, caregiver relations and peer relations showed stability over time. In other words, the results of the measurement at three time points showed that the caregiver relations and peer relations at the previous time had a significant effect on the caregiver relations and peer relations at the later time point. Second, the previous caregiver relations had a significant effect on the subsequent peer relations over time. Third, the previous peer relations had a significant effect on the subsequent caregiver relations over time. This study confirmed the interrelationships of caregiver relations and peer relations of children in care by examining the longitudinal data using the longitudinal analysis method.

• The Korean Journal of Applied Statistics
• /
• v.33 no.3
• /
• pp.281-296
• /
• 2020

Repeated outcomes from the same subjects are referred to as longitudinal data. Analysis of the data requires different methods unlike cross-sectional data analysis. It is important to model the covariance matrix because the correlation between the repeated outcomes must be considered when estimating the effects of covariates on the mean response. However, the modeling of the covariance matrix is tricky because there are many parameters to be estimated, and the estimated covariance matrix should be positive definite. In this paper, we consider analysis of multivariate longitudinal data via two modeling methodologies for the covariance matrix for multivariate longitudinal data. Both methods describe serial correlations of multivariate longitudinal outcomes using a modified Cholesky decomposition. However, the two methods consider different decompositions to explain the correlation between simultaneous responses. The first method uses enhanced linear covariance models so that the covariance matrix satisfies a positive definiteness condition; in addition, and principal component analysis and maximization-minimization algorithm (MM algorithm) were used to estimate model parameters. The second method considers variance-correlation decomposition and hypersphere decomposition to model covariance matrix. Simulations are used to compare the performance of the two methodologies.

• The Mathematical Education
• /
• v.55 no.4
• /
• pp.447-465
• /
• 2016

The purpose of this study is to analyze longitudinal trends of students' mathematical affective domain by use of the data mining method. For this purpose, we used the Korea education longitudinal study(KELS 2005) which was the survey data for students' achievement test, affective domain test, teachers' evaluation, and parents' evaluation from $7^{th}$ grader in the year of 2005 to $11^{th}$ grader in the year of 2010. Subjects of this study is a total of 5040 students who answered to the mathematical affective domain survey in KELS 2005. The result findings are as follows. First, students' affective domain had changed negatively as they went up to higher grade. Second, if students' affective domain had built at a certain level in $7^{th}$ grade, the level did not change easily until $11^{th}$ grade. Third, major factors of students' affective domain were shown to be self-efficacy, intrinsic motivation, efforts and patient, and time management.

• Journal of Auto-vehicle Safety Association
• /
• v.10 no.1
• /
• pp.38-44
• /
• 2018

This paper presents an analysis on driving safety in lane change situation based on road driving data. Autonomous driving is a global trend in vehicle industry. LKAS technologies are already applied in commercial vehicle and researches about lane change maneuver have been actively studied. In autonomous vehicle, not only safety control issue but also imitating human driving maneuver is important. Driving data analysis in lane change situation has been usually dealt with ego vehicle information such as longitudinal acceleration, yaw rate, and steering angle. For this reason, developing safety index according to surrounding vehicle information based on human driving data is needed. In this research, driving data is collected from perception module using LIDAR, radar and RT-GPS sensors. By analyzing human driving pattern in lane change maneuver, safety index that considers both ego vehicle and surrounding vehicle state by using relative velocity and longitudinal clearance has been designed.

• The Journal of Bigdata
• /
• v.5 no.1
• /
• pp.11-27
• /
• 2020

In this paper, we developed a big data platform to store, process, and analyze effectively on such education longitudinal study data. And it was applied to the Seoul Education Longitudinal Study(SELS) to confirm its usefulness. The developed platform consists of data preprocessing unit and data analysis unit. The data preprocessing unit 1) masking, 2) converts each item into a factor 3) normalizes / creates dummy variables 4) data derivation, and 5) data warehousing. The data analysis unit consists of OLAP and data mining(DM). In the multidimensional analysis, OLAP is performed after selecting a measure and designing a schema. The DM process involves variable selection, research model selection, data modification, parameter tuning, model training, model evaluation, and interpretation of the results. The data warehouse created through the preprocessing process on this platform can be shared by various researchers, and the continuous accumulation of data sets makes further analysis easier for subsequent researchers. In addition, policy-makers can access the SELS data warehouse directly and analyze it online through multi-dimensional analysis, enabling scientific decision making. To prove the usefulness of the developed platform, SELS data was built on the platform and OLAP and DM were performed by selecting the mathematics academic achievement as a measure, and various factors affecting the measurements were analyzed using DM techniques. This enabled us to quickly and effectively derive implications for data-based education policies.

• The Korean Journal of Applied Statistics
• /
• v.35 no.6
• /
• pp.703-724
• /
• 2022

Although longitudinal studies mainly produce multivariate longitudinal data, most of existing statistical models analyze univariate longitudinal data and there is a limitation to explain complex correlations properly. Therefore, this paper describes various methods of modeling the covariance matrix to explain the complex correlations. Among them, modified Cholesky decomposition, modified Cholesky block decomposition, and hypersphere decomposition are reviewed. In this paper, we review these methods and analyze Korean children and youth panel (KCYP) data are analyzed using the Bayesian method. The KCYP data are multivariate longitudinal data that have response variables: School adaptation, academic achievement, and dependence on mobile phones. Assuming that the correlation structure and the innovation standard deviation structure are different, several models are compared. For the most suitable model, all explanatory variables are significant for school adaptation, and academic achievement and only household income appears as insignificant variables when cell phone dependence is a response variable.