Multilevel modelling

The use of ability grouping is frequently justified on the grounds that it is an effective means of raising attainment. Little large-scale quantitative research has been undertaken since the introduction of the National Curriculum in England and Wales. The aim of this article is to examine the effects of setting on students' achievement in English, mathematics and science General Certificate of Secondary Education (GCSE) examinations. Data were collected from a cohort of over 6000 Year 9 students in 45 mixed secondary comprehensive schools who were followed up in Year 11 when they sat GCSE examinations. Multilevel modelling was used to estimate the effect of setting on GCSE attainment, taking account of prior attainment, social disadvantage, gender and attendance. There were no significant effects of setting in English, mathematics or science. Effects on higher and lower attaining students were not consistent in the three subjects. Socially disadvantaged students achieved significantly lower grades and girls achieved higher grades than boys, especially in English. In all three subjects, students of similar ability achieved higher GCSE grades when they were placed in higher sets. Theoretical and practical implications are discussed.

This study analyses the relationships between self-rated health and both individual and mean national social trust, focusing on a variant of Wilkinson's hypothesis that individuals will be less healthy the greater the lack of social cohesion in a country. It employs multilevel modelling on World Values Survey data across 69 countries with a total sample of 160,436 individuals. The results show that self-rated health are positively linked to social trust at both country and individual levels after controlling for individual socio-demographic and income variables plus individual social trust; increased trust is associated with better health. Moreover, this analysis of social trust gives some insight into distinctive results for the former Soviet Bloc countries, which have high reported levels of poor health, alongside the Scandinavian countries which have high levels of trust and better health situations. Our results support and extend the Wilkinson hypothesis that the level of trust, an indicator of social cohesion, is predictive of individuals' health.

Researchers have paid little attention to non-adult child supervision and the prevalence and factors influencing this practice in low-income countries. A better understanding of this phenomenon is needed to inform the development and implementation of policies and interventions to enhance child supervision in those settings. This study explores the prevalence and factors associated with young children being home alone or under the care of another young child in Lao People's Democratic Republic. Using the 2011-2012 Lao Social Indicator Survey (N = 10,740 for the subsample of 'child was home alone' and N = 10,539 for the subsample of 'child cared by another child < 10 years of age'), multi-level Poisson regressions were performed to determine the number of days children under five years of age were home alone or under the care of another child younger than 10 years of age. Large discrepancies across provinces and between urban and rural populations within each province were found. Children living in rural areas were more than five times more likely to be unsupervised than children living in urban settings (incidence rate ratio, IRR 5.2; 95% CI: 1.8-15.2), and children living in rural areas were nearly twice more likely to be under the care of another child than children living in urban settings (IRR 1.9; 95% CI: 1.3-2.8). Age was also a common factor in explaining variation in both dependent variables. Policies aimed at facilitating adequate child care and supervision should consider rurality to increase children's protection.

Een Pedagogisch-Didactische KernStructuur® (PDKS®) bestaat uit een diagnostische, met name op kerndoelen en curriculumspecificaties gebaseerde systematiek voor het ontwerpen en realiseren van ‘Optimaliserend Onderwijs’ en de bijbehorende speel-/leerprocessen voor elk kind of elke leerling. In dit document wordt een nieuw ontwerp van een praktijkgerichte PDKS®-systematiek uitgewerkt die is ingebed in een webapplicatie ten behoeve van gebruik in instellingen voor vóór- en vroegschoolse educatie en in scholen voor PO en VO. Doel van dit digitale curriculumbeheersysteem is om het voor genoemde instellingen, scholen en leerkrachten / docenten gemakkelijker te maken het onderwijs en de speel-/leerprocessen verantwoord te individualiseren. Dit wil zeggen: deze onderwijs- en leerprocessen adequaat voor te bereiden en in te zetten ter ondersteuning van de ontwikkeling en de leerbehoefte, het leerniveau en het leertempo van individuele leerlingen en groepjes leerlingen in een flexibel groeps- of klasverband. De groeps- of klasorganisatie van leerlingen wordt onafhankelijk(er) van leeftijd en leerlingen kunnen verantwoord, individueel of samen, de eigen speel-/leerprocessen méér zelf reguleren. Via de hier ontwikkelde, informatietechnologisch ingebedde PDKS® kunnen essentiële onderwijs- en leerprocessen worden geïntegreerd. Bijvoorbeeld diagnostiek gebaseerd op individuele én landelijke indicatoren (dubbele diagnostiek), passende instructie en speel-/leerprocessen, vrije curriculumactiviteiten, curriculuminformatie en -beheer, en systematische evaluatie op verschillende onderwijsniveaus (leerling, groepje, groep/klas, school, bovenschools, landelijk). Samenwerking met jeugdzorg wordt eenduidiger, gemakkelijker en méér preventief. De combinatie van resultaten van landelijk genormeerde toetsing en de schoolgebonden evaluatie van leerarrangementen levert tevens concreet zicht op de feitelijke ‘schoolkwaliteit’.

In human mate choice, sexually dimorphic faces and voices comprise hormone-mediated cues that purportedly develop as an indicator of mate quality or the ability to compete with samesex rivals. If preferences for faces communicate the same biologically relevant information as do voices, then ratings of these cues should correlate. Sixty participants (30 male and 30 female) rated a series of opposite-sex faces, voices, and faces together with voices for attractiveness in a repeated measures computer-based experiment. The effects of face and voice attractiveness on face-voice compound stimuli were analyzed using a multilevel model. Faces contributed proportionally more than voices to ratings of face-voice compound attractiveness. Faces and voices positively and independently contributed to the attractiveness of male compound stimuli although there was no significant correlation between their rated attractiveness. A positive interaction and correlation between attractiveness was shown for faces and voices in relation to the attractiveness of female compound stimuli. Rather than providing a better estimate of a single characteristic, male faces and voices may instead communicate independent information that, in turn, provides a female with a better assessment of overall mate quality. Conversely, female faces and voices together provide males with a more accurate assessment of a single dimension of mate quality.

We explored links between social capital and self-rated health and life satisfaction in a diverse sample of rich and developing countries. A four-factor measure of social capital was developed using data on 69,725 adults in 50 countries that were collected in the World Values Survey. Multilevel analyses showed links between country social capital and health and life satisfaction. However, cross-level interactions indicated that the benefits of social capital were greater in women than men, in older adults and in more trusting, affiliated individuals. Social inequalities in the contributions of social capital to population health are worthy of further study.

This article empirically assesses the validity of current theoretical models of attitudes toward immigrants and immigration policy in the Asia-Pacific region. This paper takes representative data from the World Value Survey and implements a multilevel model to test five of the main theories in the literature: the human capital theory, the social capital theory, the political orientation theory, the contact/group threat theory and the economic competition theory. The results from the analysis lend credence to the important effects of human capital, social capital and political-ideological variables on respondents’ attitudes toward immigrants and immigration policy. However, results provide mixed and weak evidence for the contact/group threat theory and the role of economic determinants. Importantly, economic determinants tend to be more powerful in shaping people’s attitudes toward immigration policy than shaping attitudes toward immigrants as people. Altogether, this article sheds new light on the validity of current theoretical models based on western countries for other areas of the world. Finally, the results from the paper also support the usefulness of non-economic, as opposed to purely economic, models in the understanding of individuals’ attitudes toward immigrants and immigration policy in the Asia-Pacific region.

As part of a larger longitudinal study of Western Australian high schools, 212 teachers were surveyed about their perceptions of the school in which they teach and the management of that school. In particular, teachers responded to questions about the school environment, their morale and the organisational health of the school, and they were asked some self-concept and teaching efficacy questions. A multilevel model was used to investigate the effects of these and other school characteristics on teacher morale (the dependent variable), while accounting for school level differences. This study demonstrated that teacher morale varied between teachers and between schools, with school environment explaining 54% of the variability in morale. That is, when the school environment was positive, teacher morale was higher. The implications of this finding point to the importance of both improving the school environment and teacher morale in order to enhance the health of schools. This is only achieved when teachers believe that their school is improving and has a positive climate.

While urban form affects building energy consumption, the pathways, direction and magnitude of the effect are disputed in the literature. This paper uses a unique dataset to examine the effect of urban form on residential electricity consumption in Ningbo, China. Using survey and utility bill data of 534 households in 46 neighborhoods in the city, we model the electricity use of households using a multi-level regression model. We find that neighborhood street configuration and tree shade are important in controlling residential electricity consumption and, consequently, greenhouse gas emissions. Our results suggest that seasonality and dwelling type condition the effect of neighborhood densities on electricity consumption. Neighborhood density is associated with household electricity consumption in summer months, while there is no such association in winter months. As neighborhood density increases, households in slab and tower apartments in dense urban neighborhoods consume more electricity in summer months, which can be partly explained by exacerbated heat island effect. Interestingly, the neighborhood density is negatively associated with summer electricity consumption for single-family houses, suggesting that the effect of neighborhood density is different for different types of dwelling units.

La confiance est le ciment de la société. Elle indique jusqu'à quel point la population se fie au fait que, dans des circonstances prévisibles, l'interaction qu'elle aura avec une institution sera également prévisible. La recherche présentée visait le développement de nouvelles méthodes permettant de mieux visualiser, analyser et comprendre l'évolution dans le temps de la confiance envers les institutions, ceci dans quatre régions du monde soit l'Asie, l'Amérique latine, l'Afrique de même que l'Asie de l'Ouest et l'Afrique du Nord (WANA). Deux types de procédures sont utilisées, la régression locale pour la visualisation de l'évolution dans le temps et l'analyse à quatre niveaux pour estimer la répartition des effets du contexte, du temps et des caractéristiques des individus sur la confiance. La base de données comprend des informations provenant de 635 sondages conduits dans 98 pays de 1995 à 2014 auprès de 827,131 répondants ayant répondu à près de 10 millions de questions sur la confiance. Les analyses montrent que la confiance est généralement stable dans le temps sauf en Afrique du Nord et au Moyen Orient (évolution quadratique), qu'elle est moins élevée en Amérique latine que dans les autres régions, qu'elle est moins élevée pour les partis politiques et les syndicats et plus élevée pour l'Église et les leaders religieux, quoiqu'en diminution. On peut également constater une absence de différence significative selon le genre et une diminution avec le temps de la différence entre les moins de 30 ans et les personnes d'âge moyen. Cet article permet de montrer le potentiel de ces types d'analyse pour analyser de façon comparative la confiance ou tout autre série de mesures présentes dans un nombre important de bases de données.

It is well-established that individual innovativeness is an indispensable quality for employees working in a variety of workplace environments. However, the interaction between the psychological and organizational factors influencing innovativeness remains unclear. This study seeks to address that research gap by examining a model comprising a mix of psychological factors (implicit theory and goal orientation) and organizational culture. Data were collected from 315 staff members working in 34 different departments/ schools at Tampere University, Finland. The study employed a Bayesian multilevel path analysis that matched the hierarchical structure of the data to test the hypotheses. The results suggest that psychological factors reflecting goal orientation are the most important for interpreting individual innovativeness. Specifically, mastery goal orientation was shown to be a positive predictor and performance-approach goal orientation a negative predictor of innovativeness. Unexpectedly, departmental culture had neither a direct effect on innovativeness nor a moderation effect on the relationships between the psychological variables and innovativeness. Plausible explanations for these results and implications for future research are discussed.

Tras las reformas estatutarias comenzadas en 2007, la situación sobre
la territorialidad en España se ha visto alterada. Por ello, a través de un
análisis multinivel, en este texto se pretende comprobar cómo es la
perspectiva ciudadana ante la organización del Estado y si el reto
catalán ha supuesto un cambio en la actitud e influencia de las
principales variables de la cultura política y, por ende, en la
configuración del perfil de los apoyos hacia la descentralización.
Gracias al análisis se ha podido comprobar cómo, efectivamente, hay
un cambio en la influencia de algunas de las variables, siendo lo más
destacable la diferenciación existente en cada una de las comunidades
autónomas del Estado.

Официальный перевод статьи, вышедшей в Social Compass December 2015, на английском языке. Полгода в открытом доступе.

В данной статье рассматривается недавно выявленный парадокс, состоящий в том, что
современные мусульмане демонстрируют более высокий уровень протестантской
трудовой этики, чем современные протестанты. Автор проверяет, подтверждается ли этот
феномен данными Всемирного исследования ценностей (WVS). В соответствии с теорией
Р. Инглхарта о постматериалистическом сдвиге, трудовая этика должна быть сильнее в
развивающихся странах, где наблюдается нехватка экзистенциальной безопасности. В
статье также проверяется, простираются ли эффекты протестантской трудовой этики за
границы религиозного населения протестантских стран. В многоуровневых моделях,
построенных по 25 437 респондентам из 55 стран, не выявлены значимые различия в
уровне трудовой этики мусульман и протестантов. Если респондент живет в исторически
протестантском обществе, этот фактор не усиливает его трудовую этику, однако если этот
респондент религиозен, то его трудовая этика будет сильнее. При переходе стран от
развивающихся к развитым трудовая этика ослабевает. Это ведет к дальнейшим вопросам
об универсальных чертах религиозной этики и нерелигиозных факторах, объясняющих
экономический прогресс, обычно связываемый с протестантской трудовой этикой.

Perry, N. S., Baucom, K. J. W., Bourne, S., Butner, J., Crenshaw, A. O., Hogan, J. N., Imel, Z. E., Wiltshire, T. J., & Baucom, B. R. W.

Researchers commonly use repeated-measures actor–partner interdependence models (RM-APIM) to understand how romantic partners change in relation to one another over time. However, traditional interpretations of the results of these models do not fully or correctly capture the dyadic temporal patterns estimated in RM-APIM. Interpretation of results from these models largely focuses on the meaning of single-parameter estimates in isolation from all the others. However, considering individual coefficients separately impedes the understanding of how these associations combine to produce an interdependent pattern that emerges over time. Additionally, positive within-person, or actor, effects are commonly misinterpreted as indicating growth from one time point to the next when they actually represent decline. We suggest that change-as-outcome RM-APIMs and vector field diagrams (VFDs) can be used to improve the understanding and presentation of dyadic patterns of association described by standard RM-APIMs. The current article briefly reviews the conceptual foundations of RM-APIMs, demonstrates how change-as-outcome RM-APIMs and VFDs can aid interpretation of standard RM-APIMs, and provides a tutorial in making VFDs using multilevel modeling.

As U.S. correctional systems continue to rollout evidence-based programs, the utility of " complimentary " programs that do not address recidivism reduction remains in question. Many U.S. prisons have a variety of prison-based animal programs, yet the outcomes are largely unexplored. This research addresses a literature gap by evaluating the intermediate outcomes associated with a statewide prison-based dog handler program. Using propensity score matching, we compared 1,001 inmates in a pretest, postentry design, aimed at measuring change across four outcomes. Results indicate that dog handler program inmates experienced significant improvement in three of four areas. Implications and further research needs are explored.

Disagreements are integral to fruitful team collaboration but have rarely been studied within actual team interactions. We develop a temporal account of how disagreement episodes begin and are resolved during team interactions, testing explanatory factors at multiple levels: team context (team conflict states and team productivity), individual characteristics and perceptions (individual status and perceptions of team viability), and behavioral patterns (problem solving versus off‐task communication) with a statistical discourse analysis of 32,448 turns of talk by 259 employees during 43 team meetings. As hypothesized, problem‐solving behaviors (e.g., describing problems and proposing solutions) ignited content disagreements, often by participants who perceived greater team viability. In contrast, after off‐task behaviors or talk by higher status team members, participants started fewer content disagreements. Moreover, content disagreements started by higher status individuals were more likely than those started by others to be resolved with agreements, especially via agreements with higher status individuals. Also, problem‐solving behaviors facilitated the resolution of disagreement episodes with agreement, whereas off‐task behaviors hindered them. Contrary to our hypotheses, team conflict states and productivity were not linked to starting or ending disagreements. We discuss the conceptual and methodological importance of capturing team interaction dynamics at work and derive practical implications for managing content disagreement.

Nordic countries are the most gender equal countries in the world, but at the same time they have disproportionally high prevalence rates of intimate partner violence (IPV) against women. High prevalence of IPV against women, and high levels of gender equality would appear contradictory, but these apparently opposite statements appear to be true in Nordic countries, producing what could be called the ‘Nordic paradox’. Despite this paradox being one of the most puzzling issues in the field, this is a research question rarely asked, and one that remains unanswered. This paper explores a number of theoretical and methodological issues that may help to understand this paradox. Efforts to understand the Nordic paradox may provide an avenue to guide new research on IPV and to respond to this major public health problem in a more effective way.

This study aims to propose and empirically test a multilevel model of cross-level interactions between authentic leadership and innovation at the team level, and perception of support for innovation and creativity at the individual level. We use data from 23 team leaders and 289 team members in a Slovenian manufacturing and processing firm engaged in producing innovative products and customer solutions and conduct a multilevel analysis using hierarchical linear modelling (HLM). The results indicate that whereas perceived team leaders’ authentic leadership directly influences team members’ individual creativity and team innovation, the impact of self-ascribed team leaders’ authentic leadership was not significant. In addition to that, the relationship between team leaders’ authenticity and creativity is mediated by perception of support for innovation. Using a multilevel approach, this is the first study to our knowledge to quantitatively examine the relationship between authentic leadership and creativity and innovation. In addition, unlike previous research on related topics that relied solely on one source of information, we examine authentic leadership with empirical data gathered from both team leaders and their employees.

Multilevel structural equation modeling (MSEM) is gaining popularity in the social sciences as a framework for estimating latent variable models in the presence of hierarchical data. In addition, we believe that MSEMs are quite helpful to psychological and educational inquiries. Many research papers have been published on technical developments in MSEM; however, we are not aware of any tutorials on how to properly implement these complex models. In addition, applied researchers may be unfamiliar with how to implement a Bayesian estimation approach to MSEM, despite the fact that it has distinct advantages in this modeling context (Depaoli and Clifton, 2015; Hox, van de Schoot, and Matthijsse, 2012). To that end, this paper serves as a nontechnical tutorial on the implementation and application of MSEM using both frequentist and Bayesian estimation methods. We demonstrate the implementation of MSEM with three motivating examples using data from the Program for International Student Assessment (PISA). In addition, we present a small Monte Carlo study to show how the estimation of MSEM is impacted by different types of priors for data structures similar to those presented in the PISA examples. We conclude with recommendations for applied researchers based on findings from the empirical examples and the simulation study. We also outline future methodological considerations in the context of Bayesian MSEM.

The objectives of this article are twofold: (a) to outline the basic concepts associated with the linear mixed model and (b) to illustrate how this model can be used to analyse systematic interindividual differences in intraindividual change, this being achieved through a longitudinal study of a cohort of children living in Cordoba (Argentina). These objectives will be met by using the PROC MIXED statement of the SAS software. This software fits a wide variety of linear mixed models to longitudinal data, thus enabling valid statistical inferences to be made. Since the choice of covariance structure may influence the values obtained in significance tests for fixed effects, we focus our attention on this aspect. The most common covariance structures for modelling longitudinal data are described and guidelines are proposed for choosing the structure which enables more powerful and more efficient regression parameter estimates to be made.

In een reguliere groep of klas zijn steeds ‘gemiddeld ontwikkelde’ leerlingen en ‘risicoleerlingen’. De risicoleerlingen wijken óf naar beneden óf naar boven af van het gemiddelde. Zij hebben minder onderwijskansen en méér schoolproblemen dan de gemiddelde leerlingen. In het reguliere jaarsysteem bieden ‘passend onderwijs voor leerlingen met beperkingen’ en ‘verrijking voor hoogbegaafde leerlingen’ geen (echte) oplossingen. Nodig is preventieve individualisering zoals ontworpen in ‘Optimaliserend Onderwijs’. De kenmerken hiervan worden behandeld en uitgewerkt in actuele onderwijsinhouden, onderwijsinstrumenten en methoden. Verhelderd wordt hoe Optimaliserend Onderwijs gelijkere onderwijskansen en minder schoolproblemen kan bevorderen bij vooral de risicoleerlingen. Landelijk onderwijsbeleid en schoolbesturen kunnen met deze informatie het risicovolle reguliere leerstofjaarsysteem transformeren in Optimaliserend Onderwijs voor elke leerling.

Introduction: Despite the current enthusiasms in applying the concept of microfinance as a poverty alleviation tool in many countries, the risk management aspects of micro-financing should not be overlooked. Ethiopia is also one of the countries where microfinance has been given due consideration as a safety net for the poor to help them overcome the adversities of poverty.
Objectives: The objective of this study is to analysis a credit risk of Microfinance loan repayment on the Microcredit and Savings Institution in Gambella town, Ethiopia.
Methodology: A study was conducted on 457 people who took loans from microfinance institutions in Gambella town, Gambella region, Ethiopia in 60 months starting January 2014 and ending December 2018. The data used in this study was the secondary data from GMCSI. We have applied Cox regression model and weibull parametric regression model after purposeful selection covariates, checking for the linearity of covariates in the model and assessment of model adequacy to diagnosis the model from semi parametric model for analyzing the loan repayment performance data on eighteen explanatory variables that would be included in the study. Loan repayment status is dependent variable, while eighteen borrower’s characteristics, institutional characteristics and firm characteristics are considered as explanatory variables. In this case the value of the dependent variable (loan repayment performance) is 0 and 1, if borrowers defaulted it takes 1 and otherwise 0.
Results: So, the outcome revealed that depicts from the total of 457 borrowers 256 (56.0%) of them are defaulted (unpaid on due) and 201 (44.0%) of them repaid their loan on due. The median of survival time is 41 months, which mean the survival curve reaches the 50% line before the end of the study by probability 0.499 which is approximation to 0.5. In fact, both Cox PH model and Weibull parametric regression model have considered but the latter model was the most appropriate for analyses of a credit risk of microfinance loan repayment in Gambella region. The significant factors identified through the relevant model are; age, residence, occupation, business sectors, repayment period, suitability of installment period and frequency of loan collection are the risk factor that cause the defaulted of loan.
Conclusion: Hence, governmental and non-governmental stakeholders should pay attention to give training on the risk factors identified on the current paper so as to optimize individual’s knowledge and awareness so that unpaid on due minimally reduced.

Key words: Credit risk; Loan repayment; Microcredit; Cox PH model ; parametric regression model.

The major pedagogical features of the book are threefold:
1. The emphasis on concepts and methods through both real data and realistic-type applications throughout.
2. Each chapter has an extensive collection of exercises, including end-of-section exercises that emphasize the material in the section. Many of these exercises are from published sources, including past examination questions from King Saud University, in Saudi Arabia, and Methodist University College Ghana. Answers to all the exercises are given at the end of the book.
3. References cited in each chapter are listed at the end of the chapter.

This paper demonstrates an area eect on health and suggests improvements in research practice for work in this ®eld. We use an area classi®cation which can be related clearly to factors that in¯uence the health of individuals and take account of people's diering propensities to draw in¯uence from their area. Multilevel analysis demonstrates that the degree of deindustrialisation which an area experienced in Britain, in the 1980s, has an independent association with the health of resident individuals. A signi®cant relationship between a person's attitude to their community and their health is shown to be independent of individual and area characteristics. We conclude that both individual and area characteristics in¯uence health. 7

With the release of Stata 14 came the mestreg command to fit multilevel mixed effects parametric survival models, assuming normally distributed random effects, estimated with maximum likelihood utilising Gaussian quadrature. In this article, I present the user written stmixed command, which serves as both an alternative and a complimentary program for the fitting of multilevel parametric survival models, to mestreg. The key extensions include incorporation of the flexible para-metric Royston-Parmar survival model, and the ability to fit multilevel relative survival models. The methods are illustrated with a commonly used dataset of patients with kidney disease suffering recurrent infections, and a simulated example, illustrating a simple approach to simulating clustered survival data using survsim (Crowther and Lambert (2012, 2013)).

Schule wird neben der Vermittlung von Wissen und Kompetenzen auch die Aufgabe der Demokratieerziehung übertragen. Ob das soziale Klima sowie die Ermöglichung von Partizipation in der Schule einen Einfluss auf die allgemeine Bereitschaft Jugendlicher hat, sich (gesellschafts-)politisch zu engagieren, wird in dieser Studie anhand der Daten eines Jugendsurveys mit 756 thüringischen Schülerinnen und Schülern im Alter von 12 – 16 Jahren untersucht. Die potentiellen Veränderungen der Intentionen zu politischem Engagement in einem Einjahreszeitraum werden mit Mehrebenenanalysen untersucht. Die
Ergebnisse zeigen, dass die Partizipationsmöglichkeiten auf Individualebene ein signifikanter Prädiktor für die politische Partizipationsbereitschaft im Folgejahr sind, nicht jedoch die interindividuell geteilte Wahrnehmung der Partizipationsmöglichkeiten. Der Einfluss des sozialen Klimas konnte nur zum Teil auf Individualebene nachgewiesen werden und fiel insgesamt deutlich geringer aus als erwartet. Eine Interaktion zwischen beiden Schulkontextaspekten in deren Wirkung auf politische Partizipationsbereitschaft zu T2 konnte nicht bestätigt werden. Insgesamt wurde in dieser Studie Schule als relevanter Kontext der politischen Sozialisation Jugendlicher bestätigt. Dass extracurriculare Faktoren, d.h. Aspekte des allgemeinen schulischen Miteinanders, also des Schulklimas, Relevanz für die politische Entwicklung Jugendlicher haben, konnte ebenso gezeigt
werden. Die Effekte fallen jedoch vergleichsweise gering aus.

Background: Muay Thai is a combat sport growing in international popularity. Previous research has highlighted marked jurisdictional differences in the judging systems employed, but no studies have compared techniques used by fighters across geographic regions or have been on how these might be a function of the different judging systems employed. This paper aimed to address this issue by examining differences in technique selection and application between Thai and UK Muay Thai fighters using notational analysis. Method: The winners of thirty-two fights involving 16 Thai and 16 UK fighters were analysed. Three multilevel Poisson regression models were used to estimate differences in technique frequency and key performance indicators between Thai and UK fighters. Results: Thai fighters used more attacking and defensive techniques than UK fighters, particularly knees (p < 0.001), roundkicks to the body (p < 0.001), and push kicks (p < 0.001). Thai fighters also tended to catch an opponent's leg significantly more often than UK fighters (p < 0.001), but UK fighters were significantly more likely to use other defensive techniques. There were also statistically significant interactions between nationality and a range of quality indicators, including delivering techniques at an appropriate distance (p < 0.001), the effectiveness of techniques used (p < 0.001), and returning to a balanced stance (p < 0.001). Conclusions: The results suggested that Thai fighters using better distancing, were more effective and more balanced. The practical implications of findings and their implications for the sport and future research are discussed.

This study aimed to assess the contextual factors associated with teenage pregnancy in sub-Saharan Africa. Using DHS data we modelled females aged 15-19 with multilevel logistic regression to establish the independent influence of social disadvantage on teenage pregnancy in West, East and Southern Africa with aid of the MLwiN programme. Results showed teenage pregnancy decreased in East Africa, plateaued in West Africa and increased slightly in Southern Africa between 1992 and 2011. Multilevel multivariate regression revealed teenage pregnancy was associated with family disruption (0.39; 0.40: P<0.05), community-levels of female unemployment (1.01; 0.99: P<0.05) and community poverty (1.01; 1.02: P<0.05) in Southern and East Africa, while only community poverty (1.01; P<0.05) independently predicted the outcome in West Africa. Our findings emphasise the necessity of creating regional-specific interventions and prevention campaigns to address multilevel factors such as family disruption as well as the need for governments to address issues of unemployment, poverty and inequality. (Afr J Reprod Health 2016 (Special Edition);20[3]: 94-107). Résumé Cette étude visait à évaluer les facteurs contextuels associés à la grossesse chez les adolescentes en Afrique sub-saharienne. En nous servons des données EDS, nous avons modélisé les femmes âgées de 15-19 ans avec la régression logistique multiniveaux pour établir l'influence indépendante de désavantage social sur la grossesse chez les adolescentes de l'Afrique : de l'Ouest, de l'Est et de l'Afrique du Sud à l'aide du programme MLwiN. Les résultats ont montré que la grossesse chez les adolescentes a diminué en Afrique Orientale, a atteint un palier en Afrique Occidentale et a légèrement augmenté en Afrique du Sud entre 1992 et 2011. La régression multivariée à plusieurs niveaux a révélé que la grossesse chez les adolescentes a été associé à une perturbation de la famille (0,39; 0,40: P <0,05), les niveaux communautaires du chômage au féminin (1,01; 0,99: P <0,05) et la pauvreté de la communauté (1.01; 1.02: P <0,05) en Afrique australe et orientale, tandis que la pauvreté communautaire (1,01; P <0,05) a prédit indépendamment le résultat en Afrique de l'Ouest. Nos résultats soulignent la nécessité de créer des interventions spécifiques régionales et des campagnes de prévention pour traiter les facteurs à plusieurs niveaux tels que la perturbation de la famille, ainsi que la nécessité pour les gouvernements de résoudre les problèmes du chômage, de la pauvreté et de l'inégalité. (Afr J Reprod Health 2016 (Edition Spéciale); 20[3]: 94-107).

Recognizing the multilevel nature of social networks and incorporating this perspective into our theorizing and analytics can help us in specifying and testing more accurate models and generate a better understanding of the integrated influence of social networks at different levels. Because organizational networks are a multilevel phenomenon and ignoring their multilevel nature presents important limitations to understanding them, they require a multilevel theory. This chapter addresses these limitations by drawing on and extending the nascent research on multilevel networks. It briefly introduces foundational social network concepts and then defines multilevel social networks, clarifying what constitutes multilevel social networks. The chapter then presents multilevel network constructs, with special emphasis on understanding constructs at higher levels than the individual. Next, it presents a framework of multilevel network models and explains how social networks at different levels influence one another, and how networks at different levels jointly influence other phenomena.

Aim. To investigate impacts of practice environment factors and burnout at the nursing unit level on job outcomes and nurseassessed quality of care in acute hospital nurses. Background. Prior research has consistently demonstrated correlations between nurse practice environments and nurses' job satisfaction and health at work, but somewhat less evidence connects practice environments with patient outcomes. The relationship has also been more extensively documented using hospital-wide measures of environments as opposed to measures at the nursing unit level. Design. Survey. Method. Data from a sample of 546 staff nurses from 42 units in four Belgian hospitals were analysed using a two-level (nursing unit and nurse) random intercept model. Linear and generalised linear mixed effects models were fitted including nurse practice environment dimensions measured with the Revised Nursing Work Index and burnout dimensions of the Maslach Burnout Inventory as independent variables and job outcome and nurse-assessed quality of care variables as dependent variables. Results. Significant unit-level associations were found between nurse practice environment and burnout dimensions and job satisfaction, turnover intentions and nurse-reported quality of care. Emotional exhaustion is a predictor of job satisfaction, nurse turnover intentions and assessed quality of care as well besides various nurse work practice environment dimensions. Nurses 'ratings of unit-level management and hospital-level management and organisational support had effects in opposite directions on assessments of quality of care at the unit; this suggests that nurses' perceptions of conditions on their nursing units relative to their perceptions of their institutions at large are potentially influential in their overall job experience. Conclusion. Nursing unit variation of the nurse practice environment and feelings of burnout predicts job outcome and nursereported quality of care variables. Relevance to clinical practice. The team and environmental contexts of nursing practice play critical roles in the recruitment and retention of nurses, and as well as in the quality of care delivered. Widespread burnout as a nursing unit characteristic, reflecting a response to chronic organisational stressors, merits special attention from staff nurses, physicians, managers and leaders.