This dissertation addresses statistical issues related to endogenous peer selection in the context of social networks, social interaction models and snowball sampling methods. The first chapter studies the peer effects of friends, studymates, and seatmates on academic performance. We use data on social networks, personality traits, and cognitive ability measures collected through a unique survey conducted in three schools in Hong Kong. We estimate a social interaction model which accounts for endogenous network formation and correlation between multiple networks with a Bayesian approach. Our results show that the cognitive ability of studymates and the conscientiousness of friends positively affect a student's mathematics exam score while the conscientiousness of studymates and the cognitive ability of friends have no effect. The second chapter proposes a novel identification strategy for social interactions in a model with endogenously formed social networks. The network endogeneity arises from the correlation between the links of the network and the unobservables that determine the outcome of interest. We show that the eigenvectors of the adjacency matrix that defines the social network are control variables for network endogeneity without imposing any parametric assumption. We propose an information criterion to select the number of eigenvectors to be included as control variables. We apply the proposed method to the same empirical application as the first chapter and compare the results. The third chapter studies a snowball sampling method for social networks with endogenous peer selection. Snowball sampling is a sampling design which preserves the dependence structure of the network. It sequentially collects the information of vertices linked to the vertices collected in the previous iteration. The snowball samples suffer from a sample selection problem because of the endogenous peer selection. We propose a new estimation method that uses the relationship between samples in different iterations to correct selection. We use the snowball samples collected from Facebook to estimate the proportion of users who support the Umbrella Movement in Hong Kong.

The work emphasizes that the roots of inequality find fertile breeding grounds on the educational systems and focuses on classroom aiming to understand possible sources of inequality among mates because it is an environment where students interact, sharing much time together. In detail, I investigate how hierarchies, networks of friends, and classroom peers influence students’ motivations, aspirations, academic competences, behaviors, and educational choices. In chapter I of this work, I outline a theoretical framework arguing that classroom inequality is a result of varying characteristics of the interacting actors such as their gender, age, ethnic origin, socioeconomic background as well as academic competencies. The classroom sorting of students with specific characteristics broadly depends on formal and informal institutional rules. To shed light on these patterns of educational systems, I rely on three distinct concepts, such as inequality, diversity, and sorting. In chapter II of this work, I test whether teachers’ grading is an inequality-enhancing factor in Italy. Previous contributions suggest that teacher’s grading is biased by preferences and stereotypes. My idea is that teachers’ grading standards might produce a hierarchy among students, even among equally able students. This hierarchy, in turn, could have a pervasive influence on students’ perception of their own competencies, thereby influencing their academic achievement, motivation, and self-stigma. In chapter III, I investigate the extent to which extent smoking and drinking friends lead to emulate the same behavior in a critical age like the adolescence. Unhealthy habits dramatically affect life expectancy, above all, when rooted in the early stage of individual development. In addition, I analyze if non-reciprocal friendship matters more or not as a driver of the behavior emulation because adolescents desire to be accepted. In final chapter IV, I test to what extent the presence of students with a migration background affects several outcomes in classrooms, including students’ attitudes and anti-social behavior. Italy is dealing with a dramatic increase of immigrant students since the late ‘80s, but a series of data suggest that the school is not well equipped for this challenge. Overall, the thesis aims to contribute to important theoretical debates in the sociology and economics of education, such as the role of relative positions in the social environment (chapter II), peer effects in critical developmental stages (chapter III), and the social integration in heterogeneous contexts (chapter IV). However, it aims also to inform policymakers on possible side effects of current widespread educational practices such as grading on a curve (chapter II), the actual role of peers in the spreading of unhealthy behaviors among adolescents (chapter III), and the need of imposing interventions devoted to optimizing classrooms compositions (chapter IV).

Thesis advisor: Ronnie Sadka / In the first essay, In this paper, I examine how the connections between activist hedge funds and other institutional investors affect the activist campaigns. I identify a positive causal effect of long-term relationships with other investors on the short-run and long-run performance of activists' target companies. Overall, my results highlight that connections to other institutional investors benefit institutional asset managers. In the second essay, we show that firms in the same board-interlock networks tend to have similar corporate governance practices. We utilize a novel instrument based on staggered adoptions of universal demand laws across states to identify causal peer effects in firms' decisions to adopt various governance provisions. The impact of universal demand laws on the incentives faced by directors as they seek to maximize their career outcomes is a likely mechanism explaining these effects. In the third essay, I investigate whether hedge funds employ short sales to mask their exiting intention when they engage in shareholder activism. Using a hand-collected sample, I find that the probability of a spike in short interest before exit announcements is higher in firms targeted by activists who have a history of short interest increase in their previous targets. According to my findings, the hypothesis is that these hedge funds are more likely to use short sales since they are more concerned about locking their profit and not taking the risk of exit announcements. Overall, this paper provides new evidence of a possible exiting strategy: Silent Exiting via short selling. / Thesis (PhD) — Boston College, 2017. / Submitted to: Boston College. Carroll School of Management. / Discipline: Finance.

Submitted by Felipe Sousa (sousa.felipefelix@gmail.com) on 2015-03-06T18:09:54Z No. of bitstreams: 1 Tese EESP_Felipe Félix.pdf: 1321305 bytes, checksum: 7c11fc274b427fc083b4457d9fd818cc (MD5) / Approved for entry into archive by Renata de Souza Nascimento (renata.souza@fgv.br) on 2015-03-06T18:11:44Z (GMT) No. of bitstreams: 1 Tese EESP_Felipe Félix.pdf: 1321305 bytes, checksum: 7c11fc274b427fc083b4457d9fd818cc (MD5) / Made available in DSpace on 2015-03-06T18:14:34Z (GMT). No. of bitstreams: 1 Tese EESP_Felipe Félix.pdf: 1321305 bytes, checksum: 7c11fc274b427fc083b4457d9fd818cc (MD5) Previous issue date: 2015-02-09 / Brazil has been experiencing an increase in demand for credit cards, especially in the lower economic classes. However, the population of lower income and less qualified represents greater risks to the operation, this fact is observed by the high rates of default. Hence, companies use debt renegotiation strategies in an attempt to recover some of the investment. However, few studies have studied the long-term consequence of these strategies. Using experiments performed by a credit card company, whose renegotiation campaigns varied from month to month, this study is looking for evidences that the renegotiation of debt offerings can affect the reputation of the firm. We conclude that increased discount on the negotiations has a significant effect on the costumer’s incentive to honor its obligations to the company. Therefore, as a consequence of peer effects, clients that had previously been in good standing also become delinquent, show us moral hazard evidence in their incentive to pay. / O Brasil vem vivenciando um aumento na demanda por cartões de crédito, principalmente nas classes baixas. Entretanto, a população de menor renda e menor qualificação representa maior riscos para a operação. Este fato é evidenciado pelas altas taxas de inadimplência. Exposto isso, empresas se utilizam de estratégias de renegociação de dívida na tentativa de recuperar parte do investimento realizado. Entretanto, poucos foram os estudos acerca da consequência no longo prazo destas estratégias. Utilizando os experimentos realizados por uma empresa de cartão de crédito, cujas campanhas de renegociação variavam mês a mês, este estudo, procurou evidências de que as ofertas de renegociação de dívidas podem afetar a reputação da firma, fazendo com que clientes da rede mesma rede social deste que recebeu a oferta de renegociação também fiquem inadimplentes. Concluímos que o aumento do desconto nas negociações tem um efeito significativo sobre o incentivo do cliente em honrar suas obrigações junto a empresa, ou seja, o aumento de 0,01 p.p. no desconto dado aos clientes aumenta em 0,05 sua probabilidade em atrasar sua fatura no próximo período.

In Chapter 2, I employ the educational production function to identify the different effects of making a friend of the same gender and the opposite gender in a school network. Unlike other gender peer effects literature that only quantifies the causal effects of the proportion of girls in an aggregated level, such as other students in the same class, grade, or dorm, I study the gender of the five best friends nominated by the student. I address the endogeneity of friendship composition by employing a novel set of instrumental variables for the number of same-gender and opposite-gender friends. We find that having more friends, especially in the early accumulation stage, lowers the test scores. We also explore the mechanisms. In Chapter 3, I investigate the role of social learning in enrollment decisions for a public pension scheme. All else equal, if a qualified rural resident moves from a community where no other co- villagers participate in the new pension scheme to a community that is fully covered by the pension scheme, the probability of an individual enrolling by 0.541 percentage point. We use robustness checks to illustrate that the estimated peer effects are not driven by the common unobserved factors, but by social interactions. In Chapter 4, we use the survey data on Chinese middle students and the instrumental variables method to explore the different effects of making friends with the same gender and the opposite gender in a school network on mental health. The empirical results find that having a larger number of same-gender friends improves mental health but having a larger number of opposite-gender friends hurts mental health. / Doctor of Philosophy / We need human connections. Along with other assets, such as money and skills, networks and relationships are resources that could help with economic outcomes in our daily lives. The rapid development of the Internet and the intelligentization of digital devices such as mobile phones have made it easier to establish relationships with others. They also generate much more data nowadays that makes it possible to study social relationships. In this dissertation, we mainly discuss two aspects of social networks. First, we use popularity as a measure of social capital and study how social capital influence middle school students' academic outcomes and mental health outcomes using Chinese data. Given that middle school students are in the embryonic stage of personal emotional development, we distinguish friends by of the same gender or not. We find that popularity with the same gender and the opposite gender differently impacts the outcomes. Second, it is intuitive that under the influence or pressure of a group, an individual tends to make his or her speech and behavior consistent with the group. Therefore, we are interested in if an individual's choice will be driven by other people's choices in the same group. We consider the adoption of a newly introduced pension program for rural residents in China. Besides demographic characteristics, a person's decision is also influenced by those around them. If a higher proportion of his or her co-villagers choose to join the pension plan, he or she is more likely to join.

Yes / This study investigates peer effects in the online peer-to-peer (P2P) lending market using data from a Chinese online lending platform, Renrendai. The empirical results indicate that both the borrowers' success rate in obtaining loans and the default rate after loans are deemed non-coercive among their peers, referred to as the peer effects of lending and peer effects of default, respectively. The peer effect of lending is more pronounced in high-risk cities, whereas the peer effect of defaulting is more pronounced for borrowers with more difficulty obtaining loans, indicating ex-ante selection and ex-post learning mechanisms, respectively. The peer effects of lending promote P2P lending market efficiency, and the peer effects of defaulting inhibit market efficiency. Collectively, our results suggest that both lenders and borrowers follow peer effects to reduce information asymmetry in P2P lending markets. / The full-text of this article will be released for public view at the end of the publisher embargo on 22 June 2025.

Denna kvalitativa intervjustudie sökte svar på frågan om hur tidigare kraftigt omotiverade elever som exponerats för allehanda insatser i syfte att stimulera studiemotivationen beskriver betydelsen av de olika påverkansfaktorer som ledde fram till att eleven i sin motivationsutveckling blev så studieinriktad att måluppfyllelsen räckte, eller såg ut att räcka, till minst en komplett gymnasieexamen. Resultaten gav för handen att endast en av skolans insatser beskrevs som betydelsefull: etablering av en förtroendefull relation med någon personal, oaktat vilken, men endast om den också utmynnade i en studieplan som eleven kunde tro på och som med fördel kontinuerligt följdes upp. Vårdnadshavares betydelse beskrevs som nästintill obefintlig. Studiens viktigaste bidrag utgörs av den mycket stora betydelsen av elevens egna beslutsamhet att såväl i som utanför klassrummet radikalt reducera umgänge med kompisar som inte sköter skolan och istället utveckla en nära vänskapsrelation till en annan allierad vän som tillsammans med eleven vill ta skolan på allvar. Samtliga informanters berättelser var därutöver oförenliga med diskursen om inre motivation och det lustfyllda lärandet. Istället var det berättelsen om nödvändigheten av att oaktat intresse och lust tvinga sig studera som framträdde.

本文透過健保資料精神疾病歸人檔，擷取1997～2004年門診中之精神分裂症患者，針對其中701位精神科專科醫師對患者的精神分裂症一、二代藥物之處方行為進行研究。迴歸結果顯示醫師性別對處方選擇沒有特別的影響，同儕效果在相當程度上對醫師產生蠻大的改變。藥價調整前，教學醫院的醫師會受到周遭醫師開藥行為影響，其受影響程度大於非教學醫院。藥價調整後，教學醫院醫師幾乎不受同儕影響，相對的，非教學醫院比起藥價變動前，更易受到同僚行為而改變其行為。不過大體而言當周遭醫師容易開新藥（舊藥）時，醫師本身也會較容易開第二代（第一代）藥物。

Despite a century of educational reforms, no matter how achievement is measured, learning and opportunity gaps can still be predicted by race and socioeconomic status. Teachers and schools are blamed for functioning to reproduce social inequality. This study investigated teacher agency and transformative potentials. It considered how teachers modified their pedagogical practices when teaching low-income and high-poverty students. In order to capture teacher beliefs and logic, a qualitative approach was used involving in-depth interviews of a small number of participants. The research used the context of the dislocation of students from high-poverty Orleans Parish schools in the year following Hurricane Katrina and their absorption into often higher income schools to understand middle-class teachers’ perspectives on their new students’ learning needs and how they adjusted their practice. Participants were middle-school mathematics teachers ranging in experience and orientation. Evacuees had weaker mathematics backgrounds (often two years below grade level). In all cases, evacuees were in classes with non-evacuees. Teachers made different pedagogical choices: continuing to use diverse methods aimed at higher-order understanding, or moving to direct instructional strategies; remediating or accelerating students with below-grade-level mathematics skills; and whether or not to help students acculturate (code-switch) from one set of classroom norms and etiquettes to another. Key factors influencing choices included: socioeconomic makeup of their classes; teachers’ level of mathematics expertise; emphasis on test scores; teachers’ views of students’ culture; and teachers’ peer environments. The study provides insights into teacher and classroom mechanisms that contributed to Katrina evacuee multi-year achievement gains.

The following study was conducted at an upper secondary school in Sweden and attempts to explore the question of what influences male pupils’ reading habits. Many quantitative international studies, including PISA, PIRLS and IEA Reading Literacy, have sought to answer this question, but only partially succeeded due to the limitations of their methods. Therefore, this study seeks to explore this question in more depth using qualitative methods, including interviews and classroom observations, but also minor tests. Two facts which the previously mentioned international studies have found is that boys and particularly immigrant boys tend to have worse reading results than their counterparts. It is therefore the aim of this study to study four male students in upper secondary school; of which two are native Swedes and the other two are unaccompanied refugee children; one from Afghanistan and the other from Morocco. The findings of this study are as follows. Firstly, necessity was found to be the single most important factor for the reading habits of these four pupils; especially the two refugees. Both refugees learnt to read under harsh circumstances in madrassas in their respective home countries. Moreover, the Moroccan pupil learnt to speak and read Spanish fluently during his seven years as a homeless child. Furthermore, in the absence of necessity, interest was found to be decisive in determining the pupils’ reading habits. In addition to this, the study theorizes that an interest in reading generally arises before the ability to read and not vice versa. However, teachers can in fact affect their pupils’ reading habits even in upper secondary school.