Tourists' information processing is a dynamic process in that their information use depends on the surrounding context. From tourists' personal characteristics (e.g., age, gender, and travel experience), nature of tourism products (e.g., intangibility and variability), to the development of information technology (e.g., the prevalent usage of mobile devices for information search), a variety of contextual factors are involved when tourists process information for decision-making. Given the importance of online reviews in the hospitality and tourism field as information sources, this dissertation aims to understand the contextual effects of online reviews on tourists' decision-making. By selecting several contextual factors, three independent and interrelated essays examine how tourists' cognitive or behavioral responses to online reviews are affected by those factors. Considering that local search (e.g., looking for nearby restaurants by using "restaurants near me" as a search query) becomes an important context for using online reviews, both Study 1 and 2 focus on the local search context. Study 1 investigates the role of online reviews in the local search context; specifically, how online reviews are used as ranking factors by local search platforms (LSPs), is examined with an analytical approach. Study 2 investigates tourists' processing of online reviews in the local search context; specifically, how online reviews are differently processed in the local search context (e.g., searching for a restaurant that can be visited immediately) compared with the non-local context (e.g., searching for a restaurant that can be visited in a month), is examined by conducting an experiment. Building on Study 2, Study 3 investigates how tourists' processing of online reviews is affected by another contextual factor, the nature of tourism products; specifically, how the variability of tourism products (i.e., their change in quality over time) influences the way tourists process online reviews, is examined through social media analytics. Results of the three essays provide empirical support for the underlying argument of this dissertation: understanding tourists' responses to online reviews depends on factors that transcend their information characteristics. As a whole, the findings of this dissertation suggest the need for considering the surrounding context to further understand how online reviews affect tourists' decision-making. As practical implications, this dissertation discusses the importance of leveraging various types of information about tourists' context (e.g., location accessed from smartphones, and physiological condition accessed through smartwatches). / Doctor of Philosophy / Tourists use online reviews within specific situations. The effects of such reviews on tourists' decision-making are difficult to explain without considering the surrounding contexts. Depending on when (e.g., before or during the trip), where (e.g., at home or destination), or for which products (e.g., restaurants, attractions, or hotels) they use online reviews, even the same online review can be differently perceived by tourists (e.g., how helpful it is). Therefore, the reviews have an increased or reduced influence on their product choices. This dissertation aims to understand the context-dependence of tourist's use of online reviews. The three essays in this dissertation examine how online reviews are used or processed by tourists under certain context: how online reviews affect tourist's decision-making in the local search context (e.g., searching for "restaurants near me" using smartphones during the trip) (Study 1); how tourists process online reviews while relying on reviews for immediately choosing places to visit (Study 2); and how tourists perceive online reviews when they are recently posted (Study 3). The findings confirm the dynamic nature of tourist's use of online reviews and offer several insights for tourism businesses to hone their strategies on marketing online reviews.

In contrast to the decline of New Public Management (NPM) in many countries, reports on executive agency policy in Korea point to its success. To explore why it has been successful, this study investigates the relationship between multi-level culture and work motivation (i.e., job satisfaction and organizational commitment). This study develops a multi-level framework and examines direct, moderating, and varying contextual effects in the relationships among organizational cultures (clan, adhocracy, market, and hierarchy), individual perceptions of organizational culture, and work motivation, using survey data from 1,535 public employees in 46 executive agencies in Korea and hierarchical linear modeling (HLM). Key findings include, first, clan, adhocracy, market, and hierarchy organizational cultures all have positive associations with organizational commitment; however, individual perceptions of organizational market and hierarchy cultures are negatively related to organizational commitment and job satisfaction. Second, market and hierarchy cultures have positive varying contextual effects in relationships with organizational commitment. Third, organizational culture has more explanatory power than individual perceptions of organizational culture. Finally, efforts to develop and sustain balanced organizational culture evidently led to the success of the executive agency system in Korea. The findings support the argument that NPM reforms need to be examined through the lens of a multi-level culture and balanced culture perspective. / Ph. D.

This dissertation has multiple aims in studying hierarchical linear models in biomedical data analysis. In Chapter 1, the novel idea of studying the durations of adolescent growth spurts as a predictor of adulthood obesity is defined, established, and illustrated. The concept of contextual effects modeling is introduced in this first section as we study secular trend of adulthood obesity and how this trend is mitigated by the durations of individual adolescent growth spurts and the secular average length of adolescent growth spurts. It is found that individuals with longer periods of fast height growth in adolescence are more prone to having favorable BMI profiles in adulthood. In Chapter 2 we study the estimation of contextual effects in a hierarchical generalized linear model (HGLM). We simulate data and study the effects using the higher level group sample mean as the estimate for the true mean versus using an Empirical Bayes (EB) approach (Shin and Raudenbush 2010). We study this comparison for logistic, probit, log-linear, ordinal and nominal regression models. We find that in general the EB estimate lends a parameter estimate much closer to the true value, except for cases with very small variability in the upper level, where it is a more complicated situation and there is likely no need for contextual effects analysis. In Chapter 3 the HGLM studies are made clearer with large-scale simulations. These large scale simulations are shown for logistic regression and probit regression models for binary outcome data. With repetition we are able to establish coverage percentages of the confidence intervals of the true contextual effect. Coverage percentages show the percentage of simulations that have confidence intervals containing the true parameter values. Results confirm observations from the preliminary simulations in the previous section of this paper, and an accompanying example of adulthood hypertension shows how these results can be used in an application.

The aim of this study is to describe and explain (individual level) public political participation, with particular focus on the significance of the local, geographical context. Studies of political participation have traditionally focused on individual level explanations. Here, however, the question of the significance of place, is also raised i.e. does place have an effect on the probability of the individual to take political action? Such causal relationships are known as contextual (or neighbourhood) effects. These occur when contextual factors affect individual behaviour so that it varies systematically between different contexts, even after controlling for individual level predictors. Although empirical research has been lacking, there is a widespread assumption that place of residence can have both positive and negative effects on outcomes at the level of the individual. This is the case especially with regard to urban residential segregation, which is believed to cause self-generating, negative effects on the political engagement of citizens. My line of argument is that contextual effects cannot be taken for granted; rather they must be tested empirically in a systematic way, using individual level data and appropriate techniques. Political participation is operationalised in terms of: voting in local elections, contacting local officials, and participating in manifestations. The local, geographical context is operationalised in two ways; as Swedish municipalities and as city districts. The latter is done using case studies of two Swedish cities; Umeå, a medium-sized town with moderate socioeconomic segregation, and Göteborg, a large city with extensive polarisation. Survey data is used and analysed by means of multilevel analysis, a technique developed especially for hierarchical data and contextual analysis of individual level outcomes. The results do not provide strong support for the hypothesis of contextual effects on public political participation. There are, in several cases, strong, bivariate relationships between socioeconomic composition and political participation at the aggregate level. However, this is not confirmed in analyses of individual level data. The variation between individuals residing in different places is significant in only one case; when the context is operationalised as municipalities and the dependent variable is participation in manifestations. This variation cannot, however, be explained neither by individual level SES/political engagement nor by socioeconomic composition at the municipal level. An analysis of crosslevel interactions shows that employed persons residing in affluent districts of Göteborg have a higher probability to vote and to participate in contacting than employed persons living in poor neighbourhoods. Similarly, individuals with an immigrant background living in affluent districts in Göteborg are more likely to vote than those living in poor areas. These results give some support for the hypothesis of contextual effects on political participation. However, as the number of observations in this particular analysis is very small, the results are not robust and, consequently, must be interpreted with caution. In order to identify relevant individual level predictors, the SES and CV-models are applied. The results indicate that socioeconomic variables such as employment status and education are important predictors of voting. However, when it comes to contacting officials and participating in manifestations, socio-political resources such as political engagement and organisational membership are better as predictors of political participation.

Il existe aujourd'hui d'énormes inégalités d'accès aux filières d'élite du système éducatif français entre enfants d'origine sociale et géographique différente. une toute petite partie des familles, concentrées dans une minorité de lycées, s'accapare l'essentiel des positions scolaires donnant accès aux élites sociales et intellectuelles. au-delà de ces constats aujourd'hui relativement bien établis, se pose la question des politiques publiques susceptibles de réduire ces inégalités et d'élargir la base sociale et géographique de recrutement des élites en france. autant la littérature abonde de constats alarmants sur la fermeture sociale des élites françaises, autant elle est pauvre en diagnostic empiriquement étayée sur le type de politique qui serait susceptible d'en ouvrir l'accès. est-il à portée de politiques publiques réalistes d'améliorer la fluidité sociale dans un pays comme la france ou bien la reproduction des élites représente-t-elle une contradiction au coeur des sociétés démocratiques modernes ?sans s'arrêter au constat fataliste du maintien voire de l'aggravation de la reproduction sociale des élites en france, cette thèse propose grâce à la collecte de données empiriques inédites de repérer scientifiquement les facteurs institutionnels et sociaux qui, à différents niveaux du système éducatif, déterminent les inégalités sociales d'accès aux filières d'élite. la richesse de ce matériel empirique nous permettra par la même occasion de concevoir les politiques publiques qui peuvent modifier le cours des choses de manière réaliste, en utilisant la marge d'action que possède encore le système éducatif pour diminuer la reproduction sociale des élites malgré les inégalités économiques qui caractérisent notre société. / Educational inequalities result not only from individual but also environmental factors.School institutions themselves create a large diversity of contexts, whose causes and effects have raised a deep interest in social sciences. In contrast to family factors, schooling conditions (schools, classes, tracks and majors, teachers, etc.) stem from administrative rules that are controlled by public authorities. As such, understanding their role in students' outcomes may help designing efficient policy levers for mitigating school inequalities. Using very unique datasets, I implement natural experiment approaches to shed new light on debates regarding the role of educational contexts on students' achievement and school inequalities, and to formulate insightful policy recommendations. Three independent research papers are contained in this thesis. The first paper investigates the link between how male-dominated a field is, and gender bias against women in this field. Contrary to expectations, I show that evaluation is biased in favor of females in more male-dominated subjects (e.g. math, philosophy) and in favor of males in more female-dominated subjects (e.g. literature, biology). The second paper examines the effect of classmates' characteristics on students' achievement in high school, and reveals the importance of keeping some peers during school transitions. The last paper evaluates the effects of new public school openings and emphasizes the role of public school proximity in the choice between the private and the public sector, a major driving force behind the large inequalities of schooling conditions.

La poursuite est un mouvement oculaire qui permet de suivre un objet qui se déplace de manière lente et continue dans notre environnement. Les recherches ont démontré que ce mouvement est sous-tendu par la contribution simultanée de signaux liés à l'image de la cible et du contexte visuel sur la rétine, de signaux extra rétiniens reflétant l'implication de mécanismes cognitifs et de la copie d'efférence, ainsi que de la mise en jeu de processus d'inhibition et de suppression de certains signaux liés au contexte visuel. Cette combinaison dynamique permet au système de s'adapter et de maintenir le mouvement de poursuite dans une grande variété de contextes. Si la présence et la sélection d'un signal de mouvement dans l'environnement est considérée comme une condition sine qua none au déclenchement et au maintien de la poursuite, certaines observations depuis les années 70 nuancent ce postulat. Afin d'approfondir la compréhension des mécanismes sous-jacent au maintien de la poursuite après disparition du signal de mouvement et de questionner la nature des signaux sensoriels conduisant à la génération de poursuite, mon travail de thèse a porté sur l'étude de l'influence de deux contextes spécifiques. Ces deux contextes, l'un visuel l'autre auditif, partagent la propriété d'être corrélés au mouvement oculaire généré par l'individu. De fait, ces contextes fournissent un signal nouveau porteur d'une information sur le mouvement en cours. Dans différentes configurations de visibilité de la cible, nous avons étudiés l'influence de ces contextes sur le maintien de la poursuite, et sur la génération volontaire de mouvements oculaires lisses et continus. / Pursuit eye movements allow us to track a target which moves continuously and slowly in our visual environment. Studies have shown this movement is based on the simultaneous contribution of retinal signals linked to the retinal image of the visual target and to the context, on extra retinal signals underlying cognitive process and the efference copy, and some inhibition and suppression processes related to the visual context. This dynamical combination allows the pursuit system to adapt in a wide range of contexts. If the presence of a motion signal in the visual environment is thought as a prerequisite to initiate and then to maintain the pursuit, some results since the 70’s moderate this view. In order to investigate the mechanisms underlying the maintenance of pursuit eye movements after target disappearance and the nature of signals leading to generate pursuit, we investigate the effects of two different contexts. These contexts, a visual one, and an auditory one, share the same property which is to be dependent on eye movements. As a result of this dependence, the contexts yield a new signal, visual or auditory, which carry an information about the ongoing eye movement. We study the effects of these information induced by the contexts in procedure involving the generation and the maintenance of smooth pursuit eye movements, and the generation of smooth and continuous eye movements without a moving target.

The aim of the study is to analyse the significance of surroundings for an individual's socio-economic career in the form of education, occupational status and income. Guided by a social theoretical perspective the study reveals places of good fortune and places of few opportunities. The approach uses two empirical data sets. The first is a register database with in-dividuals in the Swedish municipalities of Gävle, Jönköping and Västerås. A cluster analysis was conducted and forms the physical and socio-demographic context of the 248 areas. The survey cohort consists of individuals born 1970, who lived at least five years in the same area during their adolescence. The most important finding is that the socio-demographic and physical context of the residential area of adolescence affects the subsequent socio-economical career. The second data set includes data from an interview study carried out with residents in the three municipalities. Inter-viewees speak to contextual effects on their socio-economic career and emphasised the importance of surroundings, especially for children. More specifically, the quantitative multi-level analyses showed that the individ-ual’s socio-demographic context during 1985-89 was significantly associated with his/her education in 1995. The individuals who lived in dwellings built during the Million Programme era had fewer years of education than others did and lower chances of being employed. Favourable places for a socio-economic career are those dominated by middle-class families where contextual effects towards higher educa-tion are evident. The income of an individual though can not be clearly considered as effected from the context in the residential area. The political implications of the study should be considered together with the nu-merous area-based programs in Sweden directed to improve conditions such as de-gree of occupation, education and income. Further studies on contextual effects can help to guide allocation of resources to schools and education in certain residential areas.

Contextual effects have previously been shown to be related to political attitudes and behaviour. Focusing on the contextual effect of social exclusion, this paper evaluates whether individuals living in areas with higher levels of social exclusion tend to be less trusting of political institutions. Regression analysis was used based on data from the Swedish Citizen Survey 2003 and Small Areas for Market Statistics. The results showed no evidence for a relationship between social exclusion and political trust.

Contextual effects have previously been shown to be related to political attitudes and behaviour. Focusing on the contextual effect of social exclusion, this paper evaluates whether individuals living in areas with higher levels of social exclusion tend to be less trusting of political institutions. Regression analysis was used based on data from the Swedish Citizen Survey 2003 and Small Areas for Market Statistics. The results showed no evidence for a relationship between social exclusion and political trust.

The present study shows how students, eleven years old, solve problems in mathematics when they work together in groups. The main question raised is about the difficulties students experience in finding the relationship between mathematics and everyday discourse and vice versa. Two empirical studies about students' problemsolving in mathematics divided into three different articles are presented in this study. One is about how students discuss, while they are trying to solve what the area of the triangle is. The other study is about how different solutions in problemsolving have various meanings. The main finding ofthis work concems communication. Depending on the social contexts, different kinds oflanguages are developed in. When children are in their everyday contexts, they use one kind oflanguage, that is they use everyday concepts. In school, children have to leam the language ofmathematics. Leaming mathematics is about getting students to use the language of mathematics to mediate events and phenomens in the world around. One can describe learning as assimilating communicative and technical tools. Which are used as mediating tools in social practices. Another main finding of this study is to show how a teacher can get students attentive to how to change between different types of discourse and how to use special conc~pts for a special context. The most important aspect is that a teacher has pedagogical conversations with students on how to move between different communicative contexts. Often, students have difficulties in understanding in which communicative context they are involved.