<p>With the pervasiveness of learning technology, various digital media, including pedagogical agents (PAs), play a vital role in learning. Although trust is relevant for the effective use of PAs, it is not extensively researched in the context of learning technology. Although most learning technologies, including PAs, are designed with high reliability, the reliability of technology learners use may differ. Thus, trust calibration is crucial: not trusting faulty automation is as vital as trusting reliable automation. Trust calibration seems highly sensitive to contexts and tasks. It is an open question if learners show trust calibration in learning tasks. <p>

<p>Replicating de Visser et al. (2016), <em>n</em> = 34 participants repeatedly guessed the meaning of a foreign word from three alternatives. Pedagogical agents varying in anthropomorphism (human, robot, and computer) aided participants with recommendations systematically decreasing and increasing in reliability over time (in a fixed order from 100%, 0%, 67%, 83%, and 100%), thus a 3x5 within-subject design. In each of 90 trials, participants (1) guessed alone, (2) got an agent’s aid, and (3) made their final decision. At each trial, participants judged their trust in the agent. We measured trust rating (subjective trust) , and, compliance (behavioral trust). The full dataset will be analysed via mixed-effects models and ANOVAs, separately for reliability decrease and trust increase conditions</p>

<p>Preliminary results from the present data showed a strong main effect of reliability on all measures of trust, showing appropriate trust calibration with trust decreasing with decreasing reliability and trust increasing with increasing reliability. The results provide insights into trust calibration in learning settings.</p>

Indonesia is a nation renowned for its abundant array of ethnicities and cultures, which are dispersed throughout the entirety of the country. Among the most notable ethnic groups are the Papuan and Javanese tribes. While the Papuan tribe is primarily situated on the rich island of Papua, the Javanese tribe extends beyond the island of Java and can be found throughout the entire archipelago because of their adaptability. The main objective of this study is to evaluate the learning experience of the Papuan tribe, specifically concerning cross-cultural studies. To achieve this goal, the study will involve Experiential Learning Theory (ELT) approach and six Papuan young man and women to explore the Javanese Wayang philosophy as a tool for enriching learning experience. The research will be conducted through an Indonesian digital learning ecosystem, which will provide a platform for students to engage with the material and participate in interactive learning activities. The study will analyze the impact of this digital learning ecosystem on Papua student learning experience, as well as its potential to bridge cultural divides and promote greater understanding and appreciation of diverse perspectives. In the end, this study revealed that all participants demonstrated favorable outcomes in their cross-cultural learning endeavors with the aid of a digital learning ecosystem. None of the subjects reported any unfavorable feedback regarding the conducted experiments. Most of the respondents shared their positive experiences, conveying their delight, ease, and enhanced comprehension of the learning process. They also observed that their concentration and engagement in cross-cultural learning were notably improved through the utilization of a digital learning ecosystem.

The ever-growing prominence and usage of online social networking sites sparked a controversial discussion regarding their impact on polarizing communication and online interactions. These platforms are often criticized for creating environments that restrict unbiased information access, amplified by algorithmic personalization, thereby contributing to the formation of filter bubbles, in which individuals’ attitudes are further strengthened. In addition, the homophilic tendency of users to avoid interaction with outgroups holding divergent viewpoints and to consume content that reinforces their opinions contributes to the emergence of echo chambers.

Existing studies adopted a narrow focus on specific phenomena associated with polarization and often disregard a holistic view, neglecting its multidimensional nature. Further, polarization is treated, measured, and defined inconsistently, pointing towards its non-unified nature.

Polarization can manifest in (1) various types, and (2) on different dimensions, such as attitude extremity, topic diversity, social fragmentation, and linguistic usage. For example, internal polarization might be characterized with increased attitude extremity of an in-group, such as conspiracy theorists, while external polarization can appear when the emergence of opposing communities leads to social fragmentation. Multiple scenarios can exhibit polarizing dynamics, with various contexts and parties involved and with differing characteristics or polarizing properties (e.g. election campaigns, athletic competitions).

To assess the multidimensional nature of the construct, the author provides a typological model, serving as a comprehensive overview on the different manifestations of polarization in online social networks. I consider a set of multiple mini case studies, where each case study represents a real polarized conversational episode. These practical examples illustrate that polarization can manifest itself differently in different contexts. The breadth of cases helps triangulate findings and leverages the identification of patterns or similarities among different entities and the grouping of these entities based on those patterns.

I develop the typology by employing several exploratory data analytics techniques such as sentiment analysis and topic modeling. To obtain further insights into the characteristics of the conversational episodes, I conduct a qualitative content analysis with manual coding.

The research objective of this article is the provision of a comprehensive typological model for polarization phenomena in social networks. By understanding polarization as a complex multidimensional concept, I account for the dynamic nature of the phenomenon, thereby enabling comparability across different conversational episodes. Hence, this work has implications for comparing community dynamics across contexts and time, and helps mitigate detrimental consequences of social network usage for individuals and society at large.