I cover three topics in empirical microeconomics. In the first chapter, titled Investor Attention to Firm versus Market-wide Information Shocks: Evidence from North Korean Missile Tests, I study whether attention towards salient political events leads to underutilization of firm-specific information in the South Korean stock market. I find that companies with earnings surprises in the top quartile experience a 1.6% increase in the abnormal return on the announcement day, but a same-day missile test takes away 70% of the positive response. In the second chapter, titled Does Cultural Proximity Mitigate the Effect of Immigration on Electoral Outcomes? (with Gerard Domènech), we study the effect of immigration on electoral outcomes using individual-level administrative data in Spain. In a multiple instrumentations framework, we find that recent immigrants who arrived within two years are associated with an increase in the vote share of the extremist parties. Such an effect persists for additional two years but dissipates in the long-term. When split by regions of origin, African immigrants have the greatest impact, followed by Latin American immigrants. European immigrants do not affect the extremist vote shares. An analysis of the unemployment rate and the number of children suggests that immigrants tend to assimilate over time. The findings are consistent with the hypothesis that cultural proximity mitigates the political reaction to immigrants. In the third chapter, titled The Effect of Daddy Quota on Gender Labor Market Outcomes (with Petra Niedermeyerova), we study the impact of a father-specific parental leave policy on labor market outcomes in Quebec, Canada. Using a province-level difference-in-difference approach, we find that the so-called daddy quota increases the probability of employment for women and decreases the wage of younger men. The results suggest that the daddy quota promoted equal opportunities for women in the labor market. In a theoretical framework, we show that policy-driven changes in gender norms are consistent with our findings.

The frequency of illicit use of prescription stimulants (IUPS) was examined in relation to known risk factors. A hierarchical linear regression (N = 903) was used to evaluate self-reported past-year IUPS and demographic variables. Gender (p < .05), race (p < .05), and GPA (p < .01) were significantly associated with the frequency of past-year illicit use. However, Greek organization membership status, perceived ADHD symptoms and access to prescription stimulant medication were not significant individual correlates of past-year IUPS. Additional analyses were conducted on the user subgroup (n = 101) to determine whether these relationships were similar compared to the general sample. Among users, ASRS Inattention and Hyperactivity-Impulsivity Subscale scores were significantly associated with the frequency of past-year IUPS (p < .05), even after accounting for gender, race, and GPA. Interpretations of these findings and implications for future research are discussed.

La présente recherche s’intéresse à la négligence des corrélations en contexte d’investissement. À l’aide d’un questionnaire réalisé auprès de 88 planificateurs financiers agréés (Pl. Fin.) en collaboration avec l’Institut québécois de la planification financière (IQPF), nos résultats suggèrent que le décideur moyen utilise plutôt le profil de tolérance au risque d’un client fictif afin d’orienter sa prise de décision (motif principal préféré par 60% des répondants). Entre deux fonds d’investissement à rendements égaux, nous constatons une tendance modale de l’individu à conseiller le fond présentant la plus petite variance individuelle indépendamment de sa corrélation. Non seulement le sujet représentatif ne favorise pas une mesure du risque agrégée basée sur le portefeuille global de son client (i.e., variance de portefeuille), mais celui-ci ne semble pas non plus concevoir l’existence d’une alternative moins risquée dans un contexte systémique. À partir de six formats de présentation différents dans un design inter-sujet, nos résultats suggèrent également qu’une surcharge d’information nuit à l’attention limitée du participant et exacerbe sa propension à intégrer l’information telle que vue et présentée (heuristique de valeur nominale). / This research seeks to highlight the underlying mechanisms that might cause correlation neglect among a group of financial experts in an investment context. Via an experimental questionnaire administered to 88 financial planners, we ask what would be their advice to twelve fictitious clients who wish to invest in a new fund according to a fixed set of information (six frames, between-subjects design). The planner had to consider the previously owned fund “to give an optimal advice that maximizes the portfolio diversification”. Between two alternative funds to invest with equal average returns, our results suggest that the representative expert emphasizes mainly the client’s risk tolerance profile to make a selection (first motive designated by 60% of subjects). We thus identify a pattern from which the modal subject recommend the low (high) individual variance fund with low (high) risk tolerance profile of a client. This observed behaviour not only neglects to account for the full portfolio variance, which depends crucially on the correlation between the funds, but also suggests a difficulty to conceptualize the impacts of aggregating two or more assets in the same system. We suggest that our results may be explained by a “face-value” heuristic hypothesis that posits that the subject neglects to consider the underlying generating process of the final outputs showed display and that this is exacerbated by an information overload when present.

Par leur utilité, les sondages sont devenus des éléments importants de toute campagne électorale. Ils ont d’ailleurs beaucoup évolué au fil du temps et on les retrouve maintenant sous plus d’une forme. En effet, inspirés par les marchés de prédiction, les sondeurs ont développé un type de questionnement complémentaire aux sondages classiques interrogeant les individus sur leurs propres préférences. Cette autre méthode invite plutôt les participants à faire part de leurs anticipations sur le comportement agrégé de la population. Néanmoins, aucune de ces deux approches n’est parfaite. Leur pouvoir prédictif peut être affecté par plusieurs facteurs. Parmi ceux-ci, on compte l’inattention envers la corrélation, un biais cognitif susceptible d’affecter les réponses des individus sondés. À la lumière de qui précède, ce mémoire a pour premier objectif de comparer le pouvoir prédictif de ces deux types de méthode afin de distinguer la plus performante. Le second objectif est de mesurer l’effet potentiel du biais d’inattention envers la corrélation sur les réponses de participants à des sondages. Pour y arriver, des sujets ont été invités à participer à une expérience en laboratoire où ils ont occupé à la fois le rôle de sondé et de preneur de décision. Les résultats de l’expérience montrent que l’inattention envers la corrélation affecte significativement les réponses de participants à des sondages. De plus, certains facteurs personnels semblent être associés à une probabilité plus élevée d’être victime du biais.

Attention-deficit/hyperactivity disorder (ADHD) is a common neurodevelopmental disorder that affects academic, social, and health functioning. The psychometric properties of measures commonly used in diagnostic settings to assess the constructs of attention, hyperactivity, and executive functioning, including concurrent and incremental validity with other commonly used measures, are not well-established. Additionally, these specific psychometric properties within ecological samples of children and adolescents presenting to community-based clinics is infrequent. The purpose of this dissertation was to describe a sample of children and adolescents presenting to a community-based mental health clinic and to test the concurrent and incremental validity of parent- and teacher-report questionnaires and neuropsychological measures of inattention, hyperactivity, and executive functioning. Participants included a consecutive sample of 597 youth aged six to eighteen years presenting to a community-based mental health clinic for a psychological assessment between 2010 and 2019. Measures included the BASC (2nd and 3rd editions), Conners 3 (long and short forms), BRIEF (1st and 2nd editions), D-KEFS, NEPSY-II, and CPT (2nd and 3rd editions). Approximately 50% of individuals were diagnosed with ADHD. Correlations between attention, hyperactivity, and executive function constructs on questionnaires were significantly and moderately-to-strongly correlated (Spearman’s  = .27-.82) to each other (e.g., BASC Hyperactivity and Conners Hyperactivity/Impulsivity subscales) and across parent and teacher forms (e.g., Conners Parent Inattention and Conners Teacher Inattention subscales). Constructs measured by neuropsychological tests were not significantly correlated to similar constructs measured by other neuropsychological measures or questionnaires (e.g., inhibition measured by NEPSY-II Response Set and BRIEF Inhibition subscale; Spearman’s  = .02-.38). Logistic regression analyses suggested that measures such as the BASC, BRIEF, NEPSY-II, and CPT do not clinically significantly increase prediction of ADHD diagnosis above and beyond the Conners 3. Results indicate convergence of information with potential redundancy in the assessment and diagnosis of ADHD across questionnaire measures and informants. Questionnaires and neuropsychological measures were not correlated, suggesting they measure different constructs or different aspects of the same construct. Results indicate that clinicians may sufficiently rely on questionnaires and reduce the number of other neuropsychological measures administered during a diagnostic assessment without significantly reducing diagnostic accuracy. However, the value of evidence of symptoms in multiple settings and converging information should still be considered.

The intent of this thesis is to explore the mediating effects of ADHD Inattentive symptoms on domestic parent-child interaction problems and working memory. Inattentive symptoms in children with ADHD are known to cause forgetfulness, slow processing speed, and negative parent-child interactions. Working memory deficits in phonological short term memory and the central executive are also well-established in children with ADHD. However, it is currently unknown to what extent inattentive symptoms are responsible for home behavior problems in conjunction with phonological working memory deficits. The aims are tested using two validated, common clinical questionnaires: The Teacher Report Form and the Home Situations Questionnaire. Additionally, working memory and executive function are tested using a phonological letter-number sequencing task and a visuospatial dot-in-the-box task. Results show that inattentive symptoms mediate the relationship between working memory and parent-child interaction problems. Teacher reports of inattention affect the degree to which the child experiences behavior problems at home, and also affects the phonological working memory system implicated in this behavior. Future directions include using a more diverse sample, investigating a wider range of ADHD symptoms, investigating effects across multiple settings, and exploring possibilities of additional executive functioning mediators.

No description available.

Tableau d'honneur de la Faculté des études supérieures et postdoctorales, 2017-2018 / Des études économiques récentes ont identifié la difficulté des personnes à prendre des décisions optimales lorsqu’il existe une corrélation entre différentes variables d’état (aléatoires), que l’on appelle maintenant dans la littérature «inattention envers la corrélation». Dans cet article, nous supposons que l’inattention envers la corrélation est un trait individuel d’une personne et nous proposons différentes mesures de cette caractéristique. Nous comparons différentes mesures en termes de corrélation à partir des résultats d’expériences de laboratoire. Nous présentons les applications de ces mesures dans des domaines précis. Mots clés : heuristiques et biais, inattention envers la corrélation, mesure. / Recent economic studies identified the difficulty of persons to make optimal decisions when there is correlation between (random) state variables, now referred to in the literatures as “correlation neglect.” In this article, we presume correlation neglect to be an individual trait of a person and propose different measures of this characteristic. We compare different measures in terms of their correlation based on results from laboratory experiments. We present applications of the measures in the field. Keywords : heuristics and biases, correlation neglect, measurement.

Dans le cadre de ce mémoire, nous proposons d’étudier l’effet de deux biais sur les investisseurs aux marchés financiers. Plus particulièrement, nous tentons de déterminer si, conformément à la théorie, l’aversion à la perte entraîne des préférences pour de longs horizons d’investissement et de faibles niveaux de rétroaction, et si l’aversion à l’ambiguïté engendre pour sa part des préférences contraires. Les données expérimentales montrent que l’effet de l’aversion à la perte sur les préférences est minime. Toutefois, l’aversion à l’ambiguïté aurait un effet manifeste sur celles-ci, menant les participants à préférer de courts horizons d’investissement, ainsi qu’une grande fréquence de rétroaction. Ces préférences peuvent vraisemblablement être expliquées par le désir des participants de mettre à jour plus rapidement leurs anticipations afin de sortir du contexte d’ambiguïté dans lequel ils se trouvent.

Objetivo: Este estudo investigou se fatores ambientais familiares estão associados com o transtorno de déficit de atenção/hiperatividade, tipo predominantemente desatento (TDAH-D). Método: Estudo de caso-controle. A amostra foi composta de 100 crianças e adolescentes com TDAH-D e 100 controles sem o diagnóstico de transtorno de déficit de atenção/hiperatividade (TDAH). A amostra foi composta de pacientes da comunidade, e todos foram avaliados de maneira sistemática, incluindo o diagnóstico feito por entrevista semi-estruturada, e revisão por psiquiatra de infância e adolescência. Os fatores familiares avaliados foram: índice de adversidade psicossocial de Rutter (discórdia marital, baixa classe social, tamanho familiar grande, criminalidade paterna, transtorno psiquiátrico materno), Family Environment Scale – FES (escores de coesão, expressividade e conflito) e Family Relationship Index – FRI (baseado nos escores acima). Resultados: Após o ajuste dos fatores confundidores (fobia social e presença de TDAH materno), a razão de chances (RC) para TDAH-D aumentou com o aumento progressivo da presença dos indicadores de adversidades de Rutter. A RC para TDAH-D foi 3,9 vezes maior em pacientes tendo 1 indicador de Rutter, quando comparado a pacientes que não tinham nenhum indicador (p=0,035; intervalo de confiança [IC] = 1.1-14). A chance de ter TDAH-D foi 7,9 vezes (p=0,002; IC=2.1- 28.9) e 8,9 vezes (p=0,006; IC=1,9-43) maior em crianças e adolescentes que tinham 2 e 3 indicadores, respectivamente. Famílias de crianças com menores escores de coesão apresentaram maior RC para TDAH-D (RC 1,24; 95% IC 1,05- 1,45). Valores menores do FRI, um índice geral do relacionamento familiar, também estiveram relacionados com uma chance maior de TDAH-D (RC 1,11; 95% IC 1,03- 1,21). Conclusões: Nós concluímos que adversidade familiar (em geral), e baixa coesão familiar, além de um baixo índice de relacionamento familiar (em particular), são associados com um aumento do risco para TDAH-D. / Objective: This study investigated whether family-environment risk factors are associated with attention-deficit/hyperactivity disorder, inattentive type (ADHD-I). Method: In a case-control study, we assessed a nonreferred sample of 100 children and adolescents with ADHD-I and 100 non-ADHD controls (6-18 years old). They were systematically evaluated through structured diagnostic interviews. The following family adversity measures were used: Rutter’s family adversity index (marital discord, low social class, large family size, paternal criminality, maternal mental disorder), Family Environment Scale - FES (subscores of cohesion, expressiveness and conflict) and Family Relationship Index - FRI (based on the subscores above). Results: After adjusting for confounding factors (social phobia and maternal history of ADHD), the odds ratio (OD) for ADHD-I increased as the number of Rutter’s indicators increased. More specifically, the odds of having ADHD-I were 3.9 times greater in patients having one indicator compared with patients having none of Rutter’s indicators (p = .035; confidence interval [CI] = 1.1 – 14). The odds of having ADHD-I were 7.9 (p = .002; CI = 2.1 – 28.9) and 8.9 (p = .006; CI = 1.9 – 43) times greater in children and adolescents with 2 and 3 indicators, respectively. Families of children with lower FES cohesion subscores presented higher OR for ADHD-I (OR 1.24; 95% confidence interval 1.05-1.45). Lower levels of FRI, a general index of family relationship, also were related to higher risk of ADHD-I (OR 1.11; 95% confidence interval 1.03-1.21). Conclusions: We conclude that family adversity (in general), and low family cohesion and low family relationship index (in particular), are associated with an increase in the risk for ADHD-I.