Although the 14th Amendment of the U.S. Constitution grants American citizenship to any person born or naturalized in this country, subjective perceptions of who belongs in the country are driven by default assumptions that the prototypical American is White. This belief that Whites are somehow more American than members of other ethnic groups lies in sharp contrast to the widespread endorsement of multiculturalism in everyday life. Two studies provide evidence that these race-based beliefs about the prototypical American can produce discriminatory behavior against ethnic minorities in domains where patriotism is relevant, but not in domains where patriotism is irrelevant. Study 1 demonstrated that the more participants believe that the prototypical American is White, the less willing they are to hire highly qualified Asian Americans in national security jobs where patriotism is essential. Additionally, this effect was partially mediated by doubts about Asian Americans’ loyalty to thee country. Study 2 replicated and extended these findings by demonstrating that the more participants believe that the prototypical American is White, the less willing they are to hire highly qualified Asian Americans in national security jobs, but not in private business jobs where patriotism is irrelevant. Together, these studies demonstrate how race-based beliefs about the prototypical American can lead to discriminatory behavior against ethnic minorities, particularly in domains where national loyalty is important.

Les effets d’amorçage comportemental ont fait l’objet de nombreuses recherches. Ces travaux ont révélé que l’activation incidente de concepts pouvait orienter le comportement des individus et que cette influence pouvait avoir lieu en dehors de la conscience. Toutefois, des critiques ont émergé à propos de ce champ de recherche, remettant en question la robustesse de certains types d’effets d’amorçage et la véracité de ces effets. L’identification des processus sous-jacents aux effets d’amorçage comportemental s’est alors imposée comme indispensable afin de mieux comprendre sous quelles conditions ils apparaissent et ainsi pourquoi ils font l’objet de difficultés de réplication. Ce travail de thèse a pour objectif d’apporter une contribution à l’identification des processus sous-jacents aux effets d’amorçage. A travers neuf expériences, nous avons testé d’une part, l’idée selon laquelle des caractéristiques méthodologiques (e.g., la durée d’exposition aux amorces) étaient propices à l’émergence de ces effets, d’autre part le rôle modérateur de certains facteurs (i.e., le concept de soi) dont une partie de la littérature suppose qu’ils sont impliqués dans ces effets. Les résultats montrent que le concept de soi pourrait être un déterminant de l’ampleur et de la direction des effets d’amorçage comportemental, selon l’implication de la conscience de soi ou de l’estime de soi. De plus, le niveau de perception des amorces semble être un candidat privilégié des modérateurs de l’amorçage comportemental. Dans l’ensemble, ces résultats fournissent des éléments afin de poursuivre l’investigation des mécanismes sous-jacents aux effets d’amorçage comportemental. / The prime-to-behavior effects have been widely studied. Research revealed that incidental concept activation could guide subsequent behavior and that this influence could be unconscious. However, some critics have emerged about these effects, questioning the robustness and the veracity of the prime-to-behavior effects. Identifying the underlying processes is essential to better understand under which conditions these effects appear and why some researchers have failed to replicate them. This work aimed at contributing to the identification of underlying mechanisms of these effects. Through nine experiments, we tested at one hand some methodological characteristics supposed to promote the emergence of these effects and, on the other hand some supposed theoretical moderators of the prime-to-behavior effects. Results prove the self-concept could determine the magnitude and the direction of these effects, depending on the self-consciousness or the self-esteem to be involved. Moreover, the level of conscious perception of the primes seems to be a potential moderator of the prime-to-behavior effects. Together, these results provide elements to continue the investigation of the underlying mechanisms of the prime- to-behavior effects.

[pt] Neuroticismo é um fator de personalidade que engloba quatro subfatores: depressão, vulnerabilidade, passividade e instabilidade emocional. Esta pesquisa teve como objetivos construir e buscar evidências de validade de medidas implícitas para aferir as facetas de neuroticismo. Para tanto, foram realizados dois estudos: (1) elaboração dos itens; (2) teste empírico das medidas. O primeiro estudo consistiu em um procedimento empírico para selecionar itens e buscar evidências de validade baseadas no conteúdo das medidas construídas. Participaram 408 pessoas, sendo que todas eram estudantes universitárias. Esse procedimento resultou em listas de palavras representativas do indivíduo e das facetas de neuroticismo. No segundo estudo, as palavras selecionadas previamente foram implementadas em testes de associação implícita para que fossem buscadas mais evidências de validade e obtidos indicadores de fidedignidade. Participaram 417 pessoas com diferentes níveis de escolaridade, sendo que 72,2 por cento tinham ensino superior incompleto e 61,4 por cento eram mulheres. Os resultados indicaram que as medidas de depressão, insegurança e passividade apresentaram evidências de validade satisfatórias, incluindo uma estrutura consonante com as definições dos construtos e relações esperadas com outras variáveis. Por exemplo, encontraram-se correlações positivas entre as medidas construídas e neuroticismo e correlações negativas com satisfação de vida. A medida de instabilidade emocional, por outro lado, se mostrou mais problemática no que diz respeito às evidências de validade. Ainda, os índices de consistência interna de todas as medidas se mostraram adequados e os coeficientes de correlação teste-reteste foram superiores aos valores que têm sido encontrados na literatura. Discute-se, entretanto, que mais evidências de validade são necessárias para que essas medidas possam ser usadas em contextos aplicados. / [en] Neuroticism is a personality factor that encompasses four subfactors: depression, vulnerability, passivity and emotional instability. This research aimed to build and search for evidence of validity of implicit measures to assess the facets of neuroticism. For this purpose, two studies were conducted: (1) elaboration of the items; (2) empirical test of the measures. The first study consisted of an empirical procedure to select items and search for evidence of validity based on the content of the built measures. Participants were 408 people, all of them were university students. This procedure resulted in lists of representative words of the individual and the facets of neuroticism. On the second study, the previously chosen words were implemented on implicit association tests, so that more evidence of validity was sought and reliability indicators were obtained. Participants were 417 people with different levels of educational background, 72.2 percent of them had incomplete higher educational background and 61.4 percent were women. The results indicated that the measures of depression, insecurity and passivity presented satisfactory evidence of validity, including a structure aligned with the definition of the constructs and expected relations to other variables. For instance, we found positive correlations between the built measures and neuroticism and negative correlations with satisfaction with life. The measure of emotional instability, however, proved to be more problematic regarding the evidence of validity. In addition, the internal consistency indexes for all measures were found to be adequate and the test-retest correlation coefficients were higher than the values found in previous studies. It is argued, however, that more evidence of validity is needed so that these measures can be used in applied contexts.