Earticle

Purpose : This study aimed to examine the three-year (2022~2024) period in job stress among workers in a manufacturing company and to identify factors influencing these changes using a linear mixed model. Methods : The study participants were 597 workers whose job stress was assessed annually from 2022 to 2024. Job stress was measured using the Korean Occupational Stress Scale (KOSS). The data were analyzed with SPSS version 25.0, applying linear mixed modeling to account for repeated measures. Results : Job stress did not show a significant change over time. However, significant predicting factors of changes in job stress included gender, age, work experience, and employment type. Conclusion : When developing and implementing workplace interventions to manage job stress among industrial workers, demographic characteristics (i.e., gender, age) and occupational factors (i.e., work experience, employment type) identified as significant factors of job stress should carefully be considered. The study is noteworthy in that it examined longitudinal trends and influencing factors of job stress over a three-year period.

Purpose: This study was conducted to determine factors influencing changes in the self-esteem of multicultural adolescents over time. Methods: A longitudinal design was adopted to analyze secondary data. This study used the 2nd, 4th, and 6th year data of Multicultural Adolescents Panel Study (MAPS) collected by the Korea Youth Policy Institute. The linear mixed model was utilized to analyze self-esteem changes over time. Results: The factors of sex and age of the students, the mother’s Korean writing ability, and the Filipino or Japanese mothers had a significant effect on the self-esteem change of multi-cultural adolescents over time. Specifically, it was found that female students’ self-esteem was decreased by 0.04 (t=-2.39, p=.017) more than male students over time. As the subject’s age and mother’s Korean writing ability increased, self-esteem was decreased by 0.01 (t=-2.53, p=.011) and 0.03 (t=-2.52, p=.012) over time. In addition, compared to that of those who had mothers from Korea, self-esteem of those who had Filipino or Japanese mothers was decreased by 0.04 significantly. Conclusion: Therefore, when developing a self-esteem promotion program, it is necessary to develop content by reflecting the characteristics of mothers as well as the gender and age of students. Furthermore, the significance of this study is in that it established a foundation for a positive self-identity of multicultural adolescents whose number is increasing recently by identifying factors affecting their self-esteem.

본 논문은 반복측정 변량분석의 대안으로서 교차 무선효과를 가진 선형 혼합효과 모형을 소개한다. 두 종류의 선형 혼합효과 모형의 설정, 추정방법, 모형 비교방법, 추론방법에 관해 R패키지 중 하나인 lme4 패키지의 lmer 함수를 가지고 설명한다. 예제데이터를 통해 선형 혼합효과 모형의 용례를 보인다. 또한, 몬테카를로 시뮬레이션 실험을 통해 특정 상황에서 선형 혼합효과 모형과 변량분석 사이에서 가설검정 수행을 비교한다.

This paper introduces a linear mixed-effects model with crossed random effects as an alternative to repeated measures analysis of variance (RM-ANOVA). With lmer function in the lme4 package, one of the R packages, two kinds of the linear mixed-effects model are described regarding the model specification, an estimation method, model comparison criteria, and an inference method. The use of the linear mixed-effects model is illustrated through an empirical example data. The performance of hypothesis testing is compared via a Monte Carlo simulation study between the mixed-effects model and variance of analysis framework.

심리학 실험자료 분석을 위한 선형 혼합효과 모형에서 실험자료의 특성에 따라서 모형설정 방법이 실험조건의 효과를 탐지하는데 미치는 영향을 몬테카를로 시뮬레이션 방법을 통해 조 사하였다. 구체적으로 자료생성모형의 무선효과의 복잡성을 변화시키면서 모형을 선택하는 전략에 따른 1종 오류비율과 검정력을 비교하였다. 그 결과, 최대모형 접근(또는 무선기울기 모형)은 자료의 무선효과 구조가 단순한 조건에서 실험조건 효과를 지나치게 엄격하게 검증 하여 상대적으로 낮은 검정력을 보였다. 반면, 모형비교 접근을 사용했을 때는 본 연구의 모 든 시뮬레이션 조건에서 1종 오류비율이 유의수준에 근접하면서도 최대모형 접근에서보다 우 수하거나 그에 상당하는 검정력을 나타냈다. 실험 연구자가 선형 혼합효과 모형을 분석도구 로 사용할 때 고려해야 할 점을 논의하였다.

In a linear mixed-effects model for psychological experimental data analysis, the effect of the model selection procedure on detecting the experimental condition effect was investigated through a Monte Carlo simulation study. Specifically, while changing the complexity of the random effect structure of the data-generating model, the type I error rate and power were compared between the model selection strategies. As a result, the maximal model approach (or the random slope model) showed relatively low statistical power under the condition that the structure of the random component of the data was simple. On the other hand, when the model comparison approach was used, the Type I error rate approached the significance level, and the power was superior to or equivalent to that of the maximal model approach in all simulation conditions of this study. Finally, we discussed the points that experimental researchers should consider when using the linear mixed-effects model as an analysis tool.

본 연구는 선형 혼합 모형을 활용하여 쓰기 능력이 장기적으로 어떻게 발전되어 가는지를 통사적 복잡성 (syntactic complexity) 측면에서 4명의 사례 연구자들이 작성한 에세이를 활용하여 분석하였다. 사례 연구자들은 자 발적으로 Criterion 이라는 자동 쓰기 평가 프로그램의 피드백을 받아 초안과 수정본을 교실 밖에서 매달 한 개 씩 일 년 동안 작성했다. 총 48개 초안 에세이가 본 연구의 분석 대상이고, 통사적 복잡성은 Syntactic Complexity Analyzer의 14개 지표 중에서 추출하여 분석하였다. 선형 혼합 모형 분석 결과를 살펴보면 다음과 같다. 시간과 T-unit의 평균길이 간에는 통계적으로 유의미한 선형 관계가 있었으며, 이는 연구 참여자들이 시간이 지날수록 길이 가 긴 T-uni을 포함한 에세이를 작성했음을 보여준다. 또한 시간과 의존절과 독립절 간의 비율 간에도 통계적으로 유의미한 선형 관계가 있었는데, 이는 연구 참여자들이 일 년 동안 높은 비율의 의존절이 포함된 에세이를 작성했음 을 시사한다.

This study investigates longitudinal writing development in terms of syntactic complexity using linear mixed effects (LME) model. This study employs essays written by four case study participants. Participants voluntarily wrote essays outside of the classroom and submitted the first and second drafts, after reflecting on the automated writing evaluation feedback (i.e., Criterion) every month over one year. A total of 48 first drafts were analyzed and syntactic complexity features were selected from Syntactic Complexity Analyzer. Results of LME showed that there was a significant positive linear relationship between time and mean length of T-unit and also between time and the ratio of dependent clauses to independent clauses, indicating that case study participants wrote longer T-units and also a higher proportion of dependent clauses over one year.

The estimation of pest density is a prime concern of Integrated Pest Management (IPM) because the success of artificial intervention such as spraying pestcides or natural enemies depends on pest density. Also, the spatial pattern of pest population within plants or plots has been studies in various ways. In this study, we applied generalized linear mixed model to Tetranychus urticae Koch , two-spotted spider mite count in glasshouse grown roses. For this analysis, the subject-specific as well as pupulation-averaged approaches are used.

일반화된 선형 혼합 모형(GLMM)은 자료가 계수의 형태로 나타나는 범주형 자료의 경우, 혹은 집락의 형태나 과산포된 비정규 자료, 또는 비선형 모형에 따르는 자료를 다루기 위한 모형 설정에 사용된다. 본 연구에서는 이에 대한 개요와 더불어, 이 모형의 적합을 위해 제시된 통계적 기법들중 의사가능도(quasi-likelihood: QL)를 이용한 추정 방법 및 Monte-Carlo 기법을 이용한 추정 방법들에 대해 조사하였다. 또한 GLMM에 대한 현재의 연구 방향 및 앞으로의 연구 가능 주제들에 대해서도 언급하였다.

The generalized linear mixed model framework is for handling count-type categorical data as well as for clustered or overdispersed non-Gaussian data, or for non-linear model data. In this study, we review its general formulation and estimation methods, based on quasi-likelihood and Monte-Carlo techniques. The current research areas and topics for further development are also mentioned.

A diagnostic tool for testing homogeneity for random effects is proposed in unbalanced linear mixed model based on score statistic. The finite sample behavior of the test statistic is examined using Monte Carlo experiments examine the chi-square approximation of the test statistic under the null hypothesis.

In this paper, we study the local influence analysis of LIU type estimator in the linear mixed models. Using the method proposed by Shi (1997), the local influence of LIU type estimator in three disturbance models are investigated respectively. Furthermore, we give the generalized Cook's distance to assess the influence, and illustrate the efficiency of the proposed method by example.

Breaks in presence (BIP) are those moments during virtual environment (VE) exposure in which participants become aware of their real world setting and their sense of presence in the VE becomes disrupted. In this study, we investigate participants' experience when they encounter technical anomalies during game play. We induced four technical anomalies and compared the BIP responses of a navigation mode game to that of a combat mode game. In our analysis, we applied a linear mixed model (LMM) and compared the results with those of a conventional regression model. Results indicate that participants felt varied levels of impact and recovery when experiencing the various technical anomalies. The impact of BIPs was clearly affected by the game mode, whereas recovery appears to be independent of game mode. The results obtained using the LMM did not differ significantly from those obtained using the general regression model; however, it was shown that treatment effects could be improved by consideration of random effects in the regression model.

This study developed a Poisson generalized linear mixed model and a procedure to estimate genetic parameters for count traits. The method derived from a frequentist perspective was based on hierarchical likelihood, and the maximum adjusted profile hierarchical likelihood was employed to estimate dispersion parameters of genetic random effects. Current approach is a generalization of Henderson's method to non-normal data, and was applied to simulated data. Underestimation was observed in the genetic variance component estimates for the data simulated with large heritability by using the Poisson generalized linear mixed model and the corresponding maximum adjusted profile hierarchical likelihood. However, the current method fitted the data generated with small heritability better than those generated with large heritability.

OBJECTIVES: To evaluate malaria transmission in relation to insecticide-treated net (ITN) coverage in Nigeria. METHODS: We used an exploratory analysis approach to evaluate variation in malaria transmission in relation to ITN distribution in 1,325 Demographic and Health Survey clusters in Nigeria. A Bayesian spatial generalized linear mixed model with a Leroux conditional autoregressive prior for the random effects was used to model the spatial and contextual variation in malaria prevalence and ITN distribution after adjusting for environmental variables. RESULTS: Spatial smoothed maps showed the nationwide distribution of malaria and ITN. The distribution of ITN varied significantly across the 6 geopolitical zones (p< 0.05). The North-East had the least ITN distribution (0.196± 0.071), while ITN distribution was highest in the South-South (0.309± 0.075). ITN coverage was also higher in rural areas (0.281± 0.074) than in urban areas (0.240± 0.096, p< 0.05). The Bayesian hierarchical regression results showed a non-significant negative relationship between malaria prevalence and ITN coverage, but a significant spatial structured random effect and unstructured random effect. The correlates of malaria transmission included rainfall, maximum temperature, and proximity to water. CONCLUSIONS: Reduction in malaria transmission was not significantly related to ITN coverage, although much could be achieved in attempts to curtail malaria transmission through enhanced ITN coverage. A multifaceted and integrated approach to malaria control is strongly advocated.

Mixed linear models have been widely used in various correlated data including multivariate survival data. In this paper we extend hierarchical-likelihood(h-likelihood) approach for mixed linear models with right censored data to that for left censored data. We also allow a general random-effect structure and propose the estimation procedure. The proposed method is illustrated using a numerical data set and is also compared with marginal likelihood method.

We consider the following semi-parametric non-linear mixed effect regression model : y\ulcorner=f($\chi$\ulcorner;$\beta$)+$\sigma$$\mu$($\chi$\ulcorner)+$\sigma$$\varepsilon$\ulcorner,i=1,…,n,y*=f($\chi$;$\beta$)+$\sigma$$\mu$($\chi$) where y'=(y\ulcorner,…,y\ulcorner) is a vector of n observations, y* is an unobserved new random variable of interest, f($\chi$;$\beta$) represents fixed effect of known functional form containing unknown parameter vector $\beta$\ulcorner=($\beta$$_1$,…,$\beta$\ulcorner), $\mu$($\chi$) is a random function of mean zero and the known covariance function r(.,.), $\varepsilon$'=($\varepsilon$$_1$,…,$\varepsilon$\ulcorner) is the set of uncorrelated measurement errors with zero mean and unit variance and $\sigma$ is an unknown dispersion(scale) parameter. On the basis of finite-sample, small-dispersion asymptotic framework, we derive an absolute lower bound for the asymptotic mean squared errors of prediction(AMSEP) of the regular-consistent non-linear predictors of the new random variable of interest y*. Then we construct an optimal predictor of y* which attains the lower bound irrespective of types of distributions of random effect $\mu$(.) and measurement errors $\varepsilon$.

본 연구에서는 부유식 해상풍력발전의 운영 및 유지보수에 필요한 체계적인 정비계획 수립을 위해 최적화기법을 활용한 수리 모형을 제안하고자 한다. 주간 단위로 선박과 기술자를 운용하는 계획정비와 고장정비 작업의 배정에 혼합정수계획법(Mixed Integer Linear Programming, MILP)을 도입하였다. 본 연구의 최적화 모델을 활용한 사례연구에서는 선박과 기술자의 투입 규모가 유지정비 비용에 미치는 영향을 확인하였으며 1년간 정비계획 수립에서 더 나아가 정비작업별 상세 스케줄링까지 연계되는 단계적 최적화 방법론을 함께 제시하였다. 세부적으로는 기상 데이터와 정비 데이터를 활용한 발전량 손실을 비가동 비용으로 반영하여 정비 우선순위를 선정하였으며, 이를 통해 국내 실정에 맞는 해상풍력단지의 유지보수 전략을 제시할 수 있을 것으로 기대한다.

In this paper, we propose operations and maintenance (O&M) planning approach for floating offshore wind farm using the mathematical optimization. To be specific, we present a MILP (Mixed Integer Linear Programming that suggests the composition of vessels, technicians, and maintenance works on a weekly basis. We reflect accessibility to wind turbines based on weather data and loss of power generation using the Jensen wake model to identify downtime cost that vary from time to time. This paper also includes a description of two-stage approach for maintenance planning & detailed scheduling and numeric analysis of the number of vessels and technicians on the O&M cost. Finally, the MILP model could be utilized in order to establish the suitable and effective maintenance planning reflecting domestic situation.

순서반응 집락자료에 대해 일반화선형혼합모형을 고려하고 모수추정 방법에 대해 논의한다. 우도함수가 변량효과의 분포함수에 관한 적분으로 표현되기 때문에 그 계산이 매우 복잡하여 그동안 여러 가지 근사적분이 제안되었다. 라플라스 변환, 가우스-에르미트 구적 및 몬테칼로 중요표집 근사적분 등이 대표적 방법들이다. 변량효과의 분포함수를 통상 다변량정규분포로 가정하며, 대부분의 통계패키지는 정규분포에 국한하고 있다. 변량효과의 분포함수에 대한 정규성의 가정에 대해 근사적분들의 로버스트성을 검토하고자 한다. 순서반응을 갖는 실제 집락자료 데이터에 대해 누적로짓 GLMM 적합 결과를 비교하고, 변량효과에 대해 몇 가지 주요 분포를 가정하여 모의실험을 수행함으로써 앞에서 언급한 근사적분의 로버스트성을 고찰한다.

We consider the generalized linear mixed model(GLMM) for the ordinal response clustered data. And the estimation techniques will be discussed. Until now several approximate integration methods have been suggested in order to maximize the likelihood function which is represented as the integral with respect to the distribution of random effects. The Laplace transformation, the Gauss-Hermite quadrature and the Monte Carlo importance sampling are mostly used. The distribution of random effects is usually assumed to be multivariate normal, and most statistical packages are confined to the normal distribution. We compare the results of cumulative logit GLMMs for a real dataset of ordinal response and we also check the robustness of approximate integration method against the normality assumption of the random effects through a Monte Carlo study.

생동성 시험과 같은 임상약리학분야의 연구는 일반적으로 한 개체 내에서 반복하여 측정된 자료구조를 사용하므로 선형혼합모형을 이용하여 분석하는 것이 보편적이다. 이러한 모형에서 랜덤효과는 개체 내 관측 자료 사이의 상관관계를 설명하고, 공분산행렬은 개체-내 변동을 설명한다. 생동성 분석은 두 약물의 약동학적 변수인 Cmax와 AUC의 기하평균비에 대한 90% 신뢰구간이 동등성 한계인 [0.8, 1.25] 범위에 드는지 알아보는 분석으로, 고정효과에는 시기, 순서군, 치료효과를, 랜덤효과에는 개체효과를 가지는 선형혼합모형을 이용하여 분석한다. 이러한 분석이 적용된 실제 예를 살펴보기 위하여 레보플록사신 연구의 자료를 활용하였다.

Linear mixed models are commonly used in the clinical pharmaceutical studies to analyze repeated measures such as the crossover study data of bioequivalence studies. In these models, random effects describe the correlation between repeated outcomes and variance-covariance matrix explain within-subject variabilities. Bioequivalence analysis verifies whether a 90% confidence interval for geometric mean ratio of Cmax and AUC between reference drug and test drug is included in the bioequivalence margin [0.8, 1.25] performed using linear mixed models with period, sequence and treatment effects as fixed and sequence nested subject effects as random. A Levofloxacin study is referred to for an example of real data analysis.

경시적 자료는 같은 개체에서 시간에 따라 반복 측정된 자료이다. 따라서 반복 측정된 자료는 상관관계가 존재하며 이것을 설명하면서 공변량의 반응변수의 효과를 추정해야 한다. 경시적 순서형 자료분석에서는 잠재변수의 조건부 누적확률을 로짓 연결함수 또는 프로빗 연결함수를 이용한 선형혼합 모형을 이용하여 공변량의 효과를 추정한다. 본 논문에서는 경시적 순서형 자료분석을 위한 두 가지 형태의 연결함수를 가지는 일반화선형혼합모형 및 주변화모형을 고찰한다. 그리고 최근에 제안된 프로빗 연결함수를 가지는 베이지안 누적 프로빗 선형혼합모형을 이용하여 경시적 순서형자료인 사업체 패널조사자료를 분석한다. 이 모형은 잠재변수의 조건부 상관계수 행렬의 모형화에 초구분해를 고려하여 고차원이며 양정치성을 만족하는 상관계수를 추정하는 방법이다. 사업체 패널 조사자료는 반응변수로 순서형 자료인 사업체의 교육훈련참여율을 고려하였고, 상관계수 행렬은 자기상관구조를 가정한 여러 모형을 비교하고 가장 적합한 모형을 제시한다. 그리고 그 모형을 이용하여 연도별 효과와 성과배분제도 실시여부, 1인당 연평균 교육시간, 노동조합여부가 유의미한 것을 찾았다.

Longitudinal data are measured repeatedly over time from the same subject. Therefore, the repeated outcomes have correlations, and it is necessary to estimate the covariate effect on the response variable while explaining the correlations. In longitudinal ordinal data analysis, the covariate effect is estimated using generalized linear mixed models using a logit link function or a probit link function. In this paper, we review the generalized linear mixed models and marginalized models with the two types of link functions for longitudinal ordinal data analysis. Specifically, a Bayesian cumulative probit linear mixed model with the probit link function is used to analyze Korean workplace panel survey (WPS) data, which is longitudinal ordinal data. In the model, the correlation matrix is high-dimensional and positive definite, and it is estimated using the hypersphere decomposition. In the WPS data, corporate training participation rate is considered as a response variable. Assuming different correlation structures, several models are compared. For the most suitable model, some explanatory variables, the annual effect, profit sharing schemes status, average annual training hours per person, and labor union status, have effects on corporate training participation rate.