Search results

How did I end up here? Advocating for situated media literacy

Media literacy is an important element in countering the harmful effects of dis- and misinformation. To date, the main perspective on media literacy can be described as a ‘liberalist’ view of media literacy, focused on the nature and source of individual messages (Phillips and Milner, 2021). In this approach, media literacy initiatives mainly focus on equipping ‘citizens with the necessary skills to make sense of the message they read, see and hear’ (Phillips and Milner, 2021: 151). In this paper, we report and reflect on the insights from the project “Putting Disinformation on the Map” in which we use creative, participative, visual and digital methods to understand how we can expand our understanding of media literacy. A central question to our inquiry is how to situate yourself in a media landscape which is inherently interconnected, networked and interdependent, also understanding yourself as a contributing actor.

COVID-19 infodemic and digital health literacy in vulnerable populations: A scoping review

BackgroundPeople from lower and middle socioeconomic classes and vulnerable populations are among the worst affected by the COVID-19 pandemic, thus exacerbating disparities and the digital divide.ObjectiveTo draw a portrait of e-services as a digital approach to support digital health literacy in vulnerable populations amid the COVID-19 infodemic, and identify the barriers and facilitators for their implementation.MethodsA scoping review was performed to gather published literature with a broad range of study designs and grey literature without exclusions based on country of publication. A search was created in Medline (Ovid) in March 2021 and translated to Medline, PsycINFO, Scopus and CINAHL with Full Text (EBSCOhost). The combined literature search generated 819 manuscripts. To be included, manuscripts had to be written in English, and present information on digital intervention(s) (e.g. social media) used to enable or increase digital health literacy among vulnerable populations during the COVID-19 pandemic (e.g. older adults, Indigenous people living on reserve).ResultsFive articles were included in the study. Various digital health literacy-enabling e-services have been implemented in different vulnerable populations. Identified e-services aimed to increase disease knowledge, digital health literacy and social media usage, help in coping with changes in routines and practices, decrease fear and anxiety, increase digital knowledge and skills, decrease health literacy barriers and increase technology acceptance in specific groups. Many facilitators of digital health literacy-enabling e-services implementation were identified in expectant mothers and their families, older adults and people with low-income. Barriers such as low literacy limited to no knowledge about the viruses, medium of contamination, treatment options played an important role in distracting and believing in misinformation and disinformation. Poor health literacy was the only barrier found, which may hinder the understanding of individual health needs, illness processes and treatments for people with HIV/AIDS.ConclusionsThe literature on the topic is scarce, sparse and immature. We did not find any literature on digital health literacy in Indigenous people, though we targeted this vulnerable population. Although only a few papers were included, two types of health conditions were covered by the literature on digital health literacy-enabling e-services, namely chronic conditions and conditions that are new to the patients. Digital health literacy can help improve prevention and adherence to a healthy lifestyle, improve capacity building and enable users to take the best advantage of the options available, thus strengthening the patient’s involvement in health decisions and empowerment, and finally improving health outcomes. Therefore, there is an urgent need to pursue research on digital health literacy and develop digital platforms to help solve current and future COVID-19-related health needs.

The earnest platform

Increasingly, Instagram is discussed as a site for misinformation, inau-thentic activities, and polarization, particularly in recent studies aboutelections, the COVID-19 pandemic and vaccines. In this study, we havefound a different platform. By looking at the content that receives themost interactions over two time periods (in 2020) related to three U.S.presidential candidates and the issues of COVID-19, healthcare, 5G andgun control, we characterize Instagram as a site of earnest (as opposedto ambivalent) political campaigning and moral support, with a rela-tive absence of polarizing content (particularly from influencers) andlittle to no misinformation and artificial amplification practices. Mostimportantly, while misinformation and polarization might be spreadingon the platform, they do not receive much user interaction.

Explainable misinformation detection from text

Explainable Misinformation Detection from Text: A Critical Look

With the proliferation of misinformation on the web, automatic methods for detecting misinformation are becoming an increasingly important subject of study. If automatic misinformation detection is applied in a real-world setting, it is necessary to validate the methods being used. Large language models (LLMs) have produced the best results among text-based methods. However, fine-tuning such a model requires a significant amount of training data, which has led to the automatic creation of large-scale misinformation detection datasets. In this paper, we explore the biases present in one such dataset for misinformation detection in English, NELA-GT-2019. We find that models are at least partly learning the stylistic and other features of different news sources rather than the features of unreliable news. Furthermore, we use SHAP to interpret the outputs of a fine-tuned LLM and validate the explanation method using our inherently interpretable baseline. We critically analyze the suitability of SHAP for text applications by comparing the outputs of SHAP to the most important features from our logistic regression models.

Exploring Bias in Data and Models for Misinformation Detection from Text

With the proliferation of misinformation on the web, automatic misinformation detection methods are becoming an increasingly important subject of study. Large language models have produced the best results among content-based methods, which rely on the text of the article rather than the metadata or network features. However, finetuning such a model requires significant training data, which has led to the automatic creation of large-scale misinformation detection datasets. In these datasets, articles are not labelled directly. Rather, each news site is labelled for reliability by an established fact-checking organisation and every article is subsequently assigned the corresponding label based on the reliability score of the news source in question. A recent paper has explored the biases present in one such dataset, NELA-GT-2018, and shown that the models are at least partly learning the stylistic and other features of different news sources rather than the features of unreliable news. We confirm a part of their findings. Apart from studying the characteristics and potential biases of the datasets, we also find it important to examine in what way the model architecture influences the results. We therefore explore which text features or combinations of features are learned by models based on contextual word embeddings as opposed to basic bag-of-words models. To elucidate this, we perform extensive error analysis aided by the SHAP post-hoc explanation technique on a debiased portion of the dataset. We validate the explanation technique on our inherently interpretable baseline model.