비균질 네트워크 상에서 성장한 표면의 거칠기에 대한 연결 밀도의 영향

유형하 인하대학교 대학원 2016 국내석사

The effects of network structure on functional characteristics of diverse complex systems have been of great interest recently. In this thesis, we investigate the effect of adding or subtracting links on the synchronizability of elements by studying the Family model for the surface growth with relaxation, in which particles are deposited at randomly selected sites and diffuse locally to the lowest neighbors, representing the load balancing in distributed computing. The roughness of the surface in the stationary state exhibits interesting dependence on the link density and the heterogeneity of connectivity. First we find in scale-free(SF) networks with the degree exponent γ larger than 3 and Erdős-Rényi(ER) networks(γ→∞) that the stationary-state roughness becomes finite, implying that the system is synchronized, if the link density exceeds a threshold. This transition behavior is observed also in the Edwards-Wilkinson(EW) model of surface fluctuation, in which the roughness diverges with system size only when the spectral dimension of the underlying network is not larger than 2. We find that the spectral dimension reaches 2 at the threshold link density, suggesting that the spectral dimension is an important characteristic for the system’s synchronizability. For SF networks with 2<γ<3, the surface roughness is not reduced significantly with increasing link density, but it keeps diverging with system size for a wide range of link although the spectral dimension of such heterogeneous networks as 2<γ<3 is quite large even with a small link density. We investigate local fluctuations depending on node degree to find that the anomalous behaviors of roughness for 2<γ<3 can originate in large fluctuations at abundant hub nodes.

일반병동 간호사와 호스피스병동 간호사의 회복탄력성과 전문직 삶의 질 비교

유형하 이화여자대학교 임상바이오헬스대학원 2023 국내석사

This study is a descriptive comparative study conducted to compare the resilience and professional quality of life between general ward nurses and hospice ward nurses. The data collection period for this study was from May 4 to May 25, 2023. The study included 100 general ward nurses from comprehensive hospitals located in Seoul, Gyeonggi, and Gyeongsangnam-do for the general ward nurses, and 100 hospice ward nurses from inpatient hospice specialized institutions located in Seoul, Gyeonggi, Gangwon, and Jeonbuk for the hospice ward nurses. The research tools used were the Korean version of the Connor-Davidson Resilience Scale (CD-RISC 25) developed by Connor and Davidson (2003) and the Professional Quality of Life: ProQOL version 5 measurement tool developed and modified by Stamm (2010) and translated by Kim Shin et al. (2010). The collected data were analyzed using descriptive statistics, t-test, ANOVA, ANCOVA, Scheffé test, Pearson's correlation coefficient, linear regression analysis, and multiple regression using SPSS WIN 27.0 software. The results of this study are as follows: 1. The compassion satisfaction, which is a sub-factor of the quality of professional life, was 32.03±5.83 for general ward nurses and 35.17±5.94 for hospice ward nurses. The burnout score was 28.93±4.22 for general ward nurses and 26.56±4.81 for hospice ward nurses. The secondary traumatic stress score was 26.30±5.57 for general ward nurses and 25.63±5.25 for hospice ward nurses. The resilience score was 2.41±0.42 for general ward nurses and 2.59±0.46 for hospice ward nurses. After adjusting for variables that showed significant differences between the two groups, such as age, marital status, religion, education level, clinical experience, and department transfer experience, there were no significant differences in the average scores of compassion satisfaction, burnout, secondary traumatic stress, and resilience between general ward nurses and hospice ward nurses. 2. When examining the quality of professional life (compassion satisfaction, burnout, secondary traumatic stress) and resilience according to the general characteristics of general ward nurses, there was a significant difference in secondary traumatic stress based on department transfer experience (t=2.30, p=.024), and there was a significant difference in resilience based on department transfer experience (t=-2.04, p=.044). When examining the quality of professional life (compassion satisfaction, burnout, secondary traumatic stress) and resilience according to the general characteristics of hospice ward nurses, compassion satisfaction showed significant differences based on age (F=7.02, p=.001), marital status (t=3.48, p=.001), education level (F=7.93, p=.001), and total clinical experience (F=6.22, p=.003). Burnout showed significant differences based on age (F=5.97, p=.004), marital status (t=-3.12, p=.002), education level (F=6.32, p=.003), and total clinical experience (F=3.85, p=.025). Secondary traumatic stress showed significant differences based on total clinical experience (F=3.46, p=.035). Resilience showed significant differences based on age (F=3.81, p=.026) and marital status (t=2.18, p=.032). 3. In both general ward nurses (r=.55, p<.001) and hospice ward nurses (r=.61, p<.001), there was a significant positive correlation between compassion satisfaction and resilience, which are sub-factors of the quality of professional life. In both general ward nurses (r=-.58, p<.001) and hospice ward nurses (r=-.57, p<.001), there was a significant negative correlation between burnout and resilience, which are sub-factors of the quality of professional life. 4. In the linear regression model of this study, the variable that influenced empathy satisfaction, a sub-factor of the quality of professional life, in general ward nurses was resilience (β=.55, p<.001), and the variable that influenced burnout was also resilience (β=-.58, p<.001), with explanatory powers of 30% and 33%, respectively. In the multiple regression model of this study, the variable that influenced empathy satisfaction, a sub-factor of the quality of professional life, in hospice ward nurses was resilience (β=.54, p<.001), and the variable that influenced burnout was also resilience (β=-.52, p<.001), with explanatory powers of 48.3% and 42.1%, respectively. Based on these research findings, although no significant differences were found in the levels of resilience, compassion satisfaction, burnout, and secondary traumatic stress between general ward nurses and hospice ward nurses, it is suggested that nursing interventions aimed at enhancing resilience to reduce burnout and secondary traumatic stress among general ward nurses and hospice ward nurses could improve their professional quality of life. 본 연구는 일반병동 간호사와 호스피스병동 간호사를 대상으로 회복탄력성과 전문직 삶의 질을 비교하기 위해 시행된 서술적 비교연구이다. 본 연구의 자료 수집 기간은 2023년 5월 4일부터 5월 25일이었고, 연구 대상자는 종합병원급 이상 기관의 일반병동 간호사 100명과 입원형 호스피스전문기관 간호사 100명이었다. 연구 도구는 Stamm (2010)이 개발하고 발전시킨 전문직 삶의 질 측정도구 5 (Professional Quality of Life: ProQOL version 5)를 김신 등(2010)이 번역한 도구와 Connor와 Davidson (2003)이 개발한 한국형 코너-데이비슨 회복력 척도(Connor-Davidson Resilience Scale; CD-RISC 25)를 사용하였다. 수집된 자료는 SPSS WIN 27.0 프로그램을 이용하여 기술통계, t-test, ANOVA, ANCOVA, Scheffé test, Pearson’s correlation coefficients, Linear regression analysis, Multiple regression으로 분석하였다. 본 연구의 결과는 다음과 같다. 1. 회복탄력성은 일반병동 간호사 2.41±0.42점, 호스피스병동 간호사 2.59±.46점이었다. 전문직 삶의 질 하위요인인 공감 만족은 일반병동 간호사 32.03±5.83점, 호스피스병동 간호사 35.17±5.94점이었고, 소진은 일반병동 간호사 28.93±4.22점, 호스피스병동 간호사 26.56±4.81점, 이차 외상성 스트레스는 일반병동 간호사 26.30±5.57점, 호스피스병동 간호사 25.63±5.25점이었다. 두 집단 간에 연령, 결혼, 종교, 최종 학력, 임상 경력, 부서 이동 경험에서 유의한 차이가 있어 차이가 나는 변수에 대한 값을 공변량으로 하여 수정 점수를 제시한 결과, 호스피스병동 간호사가 일반병동 간호사보다 공감 만족과 회복탄력성의 평균값이 더 높고, 소진과 이차 외상성 스트레스의 평균값이 더 낮았지만 유의한 차이를 나타내지 않았다. 2. 일반병동 간호사의 일반적 특성에 따른 회복탄력성과 전문직 삶의 질(공감 만족, 소진, 이차 외상성 스트레스)을 살펴보면, 이차 외상성 스트레스는 부서 이동 경험(t=2.30, p=.024)에서 유의한 차이가 있었고, 회복탄력성은 부서 이동 경험(t=-2.04, p=.044)에서 유의한 차이가 있었다. 호스피스병동 간호사의 일반적 특성에 따른 회복탄력성과 전문직 삶의 질(공감 만족, 소진, 이차 외상성 스트레스)을 살펴보면, 회복탄력성은 연령(F=3.81, p=.026), 결혼 여부(t=2.18, p=.032)에 따라 유의한 차이가 있었고, 공감 만족은 연령(F=7.02, p=.001), 결혼 여부(t=3.48, p=.001), 학력(F=7.93, p=.001), 총 임상 경력(F=6.22, p=.003)에 따라, 소진은 연령(F=5.97, p=.004), 결혼 여부(t=-3.12, p=.002), 학력(F=6.32, p=.003), 총 임상 경력(F=3.85, p=.025)에 따라, 이차 외상성 스트레스는 총 임상 경력(F=3.46, p=.035)에 따라 유의한 차이를 나타냈다. 3. 일반병동 간호사(r=.55, p<.001)와 호스피스병동 간호사(r=.61, p<.001) 모두에서 회복탄력성과 전문직 삶의 질의 하위요인인 공감 만족은 유의한 양적 상관관계가 있었고, 일반병동 간호사(r=-.58, p<.001)와 호스피스병동 간호사(r=-.57, p<.001) 모두에서 회복탄력성과 전문직 삶의 질의 하위요인인 소진은 유의한 부적 상관관계가 있었다. 4. 본 연구의 선형회귀모형에서 일반병동 간호사의 전문직 삶의 질의 하위요인인 공감 만족에 영향을 미치는 변수는 회복탄력성(β=.55, p<.001), 소진에 영향을 미치는 변수는 회복탄력성(β=-.58, p<.001)으로, 각 변수에 대한 설명력은 30%, 33%였다. 본 연구의 다중회귀모형에서 호스피스병동 간호사의 전문직 삶의 질 하위요인인 공감 만족에 영향을 미치는 변수는 회복탄력성(β=.54, p<.001)이었고, 소진에 영향을 미치는 변수도 회복탄력성(β=-.52, p<.001)으로, 각 변수에 대한 설명력은 48.3%, 42.1%였다. 이상의 연구 결과를 통해 일반병동 간호사와 호스피스병동 간호사 간의 회복탄력성과 전문직 삶의 질 하위요인인 공감 만족, 소진 및 이차 외상성 스트레스 정도에 유의한 차이를 나타내지 않았지만, 두 병동 간호사의 회복탄력성이 공감 만족과 소진에 유의한 영향요인으로 나타났다. 따라서 회복탄력성이 높을수록 공감 만족이 높아지고 소진이 감소되었으므로 일반병동 간호사와 호스피스병동 간호사의 전문직 삶의 질을 높이기 위해 회복탄력성을 강화하는 프로그램을 적용한 간호 중재가 필요하다.

Fluctuations on complex networks

유형하 인하대학교 대학원 2020 국내박사

As technology advances, the world has become more and more wired than ever before and complexities of those various wired systems exceed the boundaries of conventional scientific methods and have made us feel almost hopeless in accessing them until re- cently. In the dawn of the 21st century, however, a brand new field, ’Network Science’ rose and has helped us accessing complex systems. Since the birth of the network sci- ence, studying complex patterns of interactions observed in real-world system by means of running dynamics on complex networks mimicking the real-world complex systems has been extensively used and has brought much progress so far. In this thesis, we are going to adopt that methodology too. Among many problems with regard to complex systems, we are specifically interested in the synchronization of loads distributed among nodes in a networked system. When a system functions stably, it shows good synchro- nization. If each unit in a system sits on regular lattice structure, we can easily find the synchronization condition. However, in many real networked systems, their connection topology is not as simple as that of regular lattices; it is random and heterogeneous, in which case, the previous studies about synchronization are useless. Therefore studying synchronization on complex networks is what we believe a necessary and essential job for us living in the age of avalanche of complex systems. In this thesis, we measure the degree of synchronization in terms of fluctuation of loads distributed among nodes in model networks. The connection topology in model networks shows multifaceted as- pects depending on control parameters of model networks. For comprehensive study, we investigate fluctuation on various model networks being generated with different control parameters, thereby having different connection topology. Many large-scaled, real-world systems such as World Wide Web etc. show sparse connectivity. The more links you add to the system, the better synchronization you get. However, adding more links is usually very costly. Hence, we investigate fluctuation with changing the link density of model networks. The importance of the minimum degree can be another important factor. The fluctuation on networks with the minimum degree larger than 1 scales dif- ferently compared to the result performed on static model networks, urging us to study fluctuation on configuration model networks with variable minimum degree. It turns out that networks’ dimension actually determines the scaling behavior of fluctuation if the given network is less heterogeneous while networks’ heterogeneity governs the scaling behavior of the fluctuation if the given network is strongly heterogeneous. The effect of minimum degree comes as the inhibiting factor of the heterogeneity-driven fluctuation; It mitigate fluctuation by giving a hub node more routes to other hub nodes to level heights between them. Meanwhile, it turns out that the action of neighbor nodes at hubs is significant for moderately heterogeneous networks. Throughout the study presented in this thesis, we find working mechanism behind fluctuation on heterogeneous complex networks. We believe this result can be applied to engineer and optimize the real-world complex systems.