Loading...
Publication
O paradoxo da atenção: a personalização algoritímica baseada em Inteligência Artificial melhora ou prejudica a atenção dos utilizadores?
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.93 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Num contexto digital cada vez mais mediado por sistemas de recomendação, esta dissertação analisa o impacto da personalização algorítmica nos processos atencionais dos utilizadores, com especial foco na atenção dirigida a conteúdos patrocinados. Partindo do contexto das plataformas de redes sociais, mais especificamente utilizando a plataforma TikTok como caso de estudo, procurou-se compreender de que forma a intensidade da personalização influencia a atenção percebida, bem como o papel mediador da perceção das filter bubbles e da fadiga algorítmica nesta relação.
A metodologia adotada consistiu num estudo experimental com 34 participantes divididos em dois grupos: um exposto ao feed “Para Ti” (n = 19), altamente personalizado, e outro ao feed “A Seguir” (n = 15), composto predominantemente por publicações de contas seguidas pelo utilizador e com menor intervenção algorítmica (ainda que sujeito a conteúdos patrocinados e alguma curadoria). Foi aplicado um questionário online após a exposição aos conteúdos, analisando-se variáveis como perceção de personalização, filter bubbles, fadiga algorítmica e atenção percebida.
Os resultados revelaram que a personalização algorítmica afeta negativamente a atenção dos utilizadores, sendo esse efeito mediado pela fadiga algorítmica, mas não significativamente pela perceção de filter bubbles. Estes dados sugerem que a sobrecarga cognitiva provocada por algoritmos excessivamente personalizados pode comprometer a eficácia da publicidade digital, quer orgânica, quer paga. As conclusões apontam para a necessidade de estratégias de
personalização mais equilibradas, que tenham em conta os limites da atenção humana e o bem-estar digital dos utilizadores. A principal limitação do estudo prende-se com o reduzido tamanho da amostra, o que condiciona a generalização dos resultados. Ainda assim, esta investigação contribui para o debate emergente sobre o futuro da atenção em ambientes digitais altamente personalizados, sendo especialmente relevante para profissionais de marketing e investigadores na área de comportamento do consumidor.
In a digital environment increasingly mediated by recommendation systems, this dissertation investigates the impact of algorithmic personalization on users’ attentional processes, with a particular focus on attention directed towards sponsored content. Drawing on the context of social media platforms - specifically using TikTok as a case study - this study aimed to understand how the intensity of personalization influences perceived attention, as well as the mediating role of filter bubbles and algorithmic fatigue in this relationship. The methodology consisted of an experimental study involving 34 participants divided into two groups: one exposed to the “For You” feed (n = 19), characterized by a high level of personalization, and another exposed to the “Following” feed (n = 15), composed predominantly of posts from followed accounts and with less algorithmic intervention (though still subject to sponsored content and some degree of curation). After the exposure, participants completed an online questionnaire assessing perceived personalization, filter bubbles, algorithmic fatigue, and perceived attention. The results showed that algorithmic personalization negatively affects users’ attention, with this effect being mediated by algorithmic fatigue, but not significantly by the perception of filter bubbles. These findings suggest that the cognitive overload caused by highly personalized algorithms may undermine the effectiveness of both organic and paid digital advertising. The conclusions highlight the need for more balanced personalization strategies that consider the limits of human attention and users’ digital well-being. The main limitation of the study lies in the small sample size, which constrains the generalizability of the results. Nevertheless, this research contributes to the ongoing debate on the future of attention in highly personalized digital environments, being especially relevant for marketing professionals and scholars in the field of consumer behaviour.
In a digital environment increasingly mediated by recommendation systems, this dissertation investigates the impact of algorithmic personalization on users’ attentional processes, with a particular focus on attention directed towards sponsored content. Drawing on the context of social media platforms - specifically using TikTok as a case study - this study aimed to understand how the intensity of personalization influences perceived attention, as well as the mediating role of filter bubbles and algorithmic fatigue in this relationship. The methodology consisted of an experimental study involving 34 participants divided into two groups: one exposed to the “For You” feed (n = 19), characterized by a high level of personalization, and another exposed to the “Following” feed (n = 15), composed predominantly of posts from followed accounts and with less algorithmic intervention (though still subject to sponsored content and some degree of curation). After the exposure, participants completed an online questionnaire assessing perceived personalization, filter bubbles, algorithmic fatigue, and perceived attention. The results showed that algorithmic personalization negatively affects users’ attention, with this effect being mediated by algorithmic fatigue, but not significantly by the perception of filter bubbles. These findings suggest that the cognitive overload caused by highly personalized algorithms may undermine the effectiveness of both organic and paid digital advertising. The conclusions highlight the need for more balanced personalization strategies that consider the limits of human attention and users’ digital well-being. The main limitation of the study lies in the small sample size, which constrains the generalizability of the results. Nevertheless, this research contributes to the ongoing debate on the future of attention in highly personalized digital environments, being especially relevant for marketing professionals and scholars in the field of consumer behaviour.
Description
Keywords
Personalização algorítmica Inteligência Artificial Filter bubbles Fadiga algorítmica Atenção do utilizador Redes Sociais Publicidade digital