News media coverage of LGBT identities over 10 years in a 400-million-word corpus.

BACKGROUND: This study is the first to analyze LGBT portrayals in a news media dataset over a decade (2010-2020). We selected Singapore as a country of interest, emblematic of a nation grappling with state-encouraged heteronormativity and a remnant colonial law against homosexuality (377A), fraught with calls for its repeal that was only enacted in 2022. Our study is interested in this period bookended by challenge and change, particularly in newspaper portrayals of LGBT narratives. Newspapers are an important source of current information and have the power to shape societal perceptions. We lay the groundwork and provide a framework to analyze news media narratives of other Commonwealth nations with colonial pasts and inherited laws criminalizing LGBT communities. OBJECTIVES: This study analyzes LGBT portrayals in a 400-million-word news media dataset over a decade (2010-2020). First, we aimed to track the volume of LGBT media coverage over time and elucidate differences in coverage of different identity markers. Second, we aimed to track sentiments on LGBT portrayals. Third, we aimed to track salient narratives circulated about LGBT stories. METHODS: The study leveraged a 400-million-word corpus from news media in Singapore, identifying the following target keywords: LGBT, Lesbian, Gay, Bisexual, Transgender, Pink Dot (a local Pride event), 377A. First, coverage volume was tracked using annual changes in keyword mentions per million, elucidating differences in coverage of different sub-groups. Second, sentiment analysis on a valence scale was conducted on LGBT collocates. Third, we distilled salient narratives about LGBT identities using thematic labelling of top-frequency collocates. RESULTS: First, overall coverage of LGBT steadily increased over the decade, though Gay identities evidenced asymmetrical coverage-outstripping 'Bisexual' keywords by seven times, 'Lesbian' by four, 'Transgender' by two. Second, sentiment scores for Pink Dot (a local pride event) were most positive; Lesbian, Gay, LGBT, Transgender were neutral; Bisexual and 377A dipped slightly negative. Third, topics differed across the four identities: uniquely, 'Lesbian' collocates related to sensationalized cinema; 'Gay' about hate crimes; 'Bisexual' about population surveys; 'Transgender' about challenges (transitioning, alienation, suicide). CONCLUSIONS: Practically, we presented a decade-long barometer of LGBT sentiments and themes on a national level, providing a framework to analyze media for more effective communication strategies-applicable to Commonwealth countries with similar inherited colonial laws. Salient repetition through media association may unwittingly frame certain issues negatively; caution is prudent in representing each sub-group adequately, rather than portraying the LGBT identity as monolithic.

Powerful tool or too powerful? Early public discourse about ChatGPT across 4 million tweets.

BACKGROUND: This paper investigates initial exuberance and emotions surrounding ChatGPT's first three months of launch (1 December 2022-1 March 2023). The impetus for studying active discussions surrounding its implications, fears, and opinions is motivated by its nascent popularity and potential to disrupt existing professions; compounded by its significance as a crucial inflexion point in history. Capturing the public zeitgeist on new innovations-much like the advent of the printing press, radio, newspapers, or the internet-provides a retrospective overview of public sentiments, common themes, and issues. OBJECTIVES: Since launch, few big data studies delved into initial public discourse surrounding the chatbot. This report firstly identifies highest-engagement issues and themes that generated the most interaction; secondly, identifies the highest-engaged keywords on both sides of the sentiment valence scale (positive and negative) associated with ChatGPT. METHODS: We interrogate a large twitter corpus (n = 4,251,662) of all publicly available English-language tweets containing the ChatGPT keyword. Our first research aim utilizes a prominent peaks model (upper-quartile significance threshold of prominence>20,000). Our second research aim utilized sentiment analysis to identify, week-on-week, highest-frequency negative, and positive keywords and emojis. RESULTS: Six prominent peaks were identified with the following themes: 'hype and hesitance', 'utility and misuse in professional and academic settings', 'demographic bias', 'philosophical thought experiments on morality' and 'artificial intelligence as a mirror of human knowledge'. Of high-frequency valence, negativity included credibility concerns, implicit bias, environmental ethics, employment rights of data annotators and programmers, the ethicality of neural network datasets. Positivity included excitement over application, especially in coding, as a creative tool, education, and personal productivity. CONCLUSIONS: Overall, sentiments and themes were double-edged, expressing excitement over this powerful new tool and wariness toward its potential for misuse.

The Impact of Aging Policy on Societal Age Stereotypes and Ageism.

BACKGROUND AND OBJECTIVES: While studies have researched ageism in public policy, few investigated the impact of aging policy on ageism-typically, an unintended consequence. Ageism is linked to $63 billion in health care costs, so its antecedents are of interest. We test the association between Aging Policy Agenda Setting and Societal Age Stereotypes and hypothesize a mediating pathway via Medicalization of Aging, moderated by demographics. RESEARCH DESIGN AND METHODS: Scholars identified Singapore's Pioneer Generation Policy (PGP) as one of the largest policy implementations in recent years, where the agenda was set by the Prime Minister at an equivalent State of the Union address in 2013, and US$7 billion allocated to fund outpatient health care costs for aged 65 years or older. More than 400,000 older adults received a PGP card and home visits by trained volunteers who co-devised a personalized utilization plan. We leveraged a 10-billion-word data set with more than 30 million newspaper and magazine articles to dynamically track Societal Age Stereotype scores over 8 years from pre- to postpolicy implementation. RESULTS: Societal Age Stereotypes followed a quadratic trend: Prior to the Aging Policy Agenda Setting from 2010 to 2014, stereotypes were trending positive; after 2014, it trended downward to become more negative. Medicalization of Aging mediated the relationship between Aging Policy Agenda Setting and Societal Age Stereotypes. Furthermore, Old-age Support Ratio moderated the mediational model, suggesting that the impact of policy on medicalization is stronger when a society is more aged. DISCUSSION AND IMPLICATIONS: We provided a framework for policymakers to ameliorate the unintended consequences of aging policies on societal ageism-if unaddressed, it will exert an insidious toll on older adults, even if initial policies are well-intentioned.

News media narratives of Covid-19 across 20 countries: Early global convergence and later regional divergence.

BACKGROUND: Seldom in history does one get a 'front row seat'-with large-scale dynamic data-on how online news media narratives shift with a global pandemic. News media narratives matter because they shape societal perceptions and influence the core tent poles of our society, from the economy to elections. Given its importance-and with the benefit of hindsight-we provide a systematic framework to analyze news narratives of Covid-19, laying the groundwork to evaluate policy and risk communications. OBJECTIVES: We leverage a 10-billion-word-database of online news, taken from over 7,000 English newspapers and magazines across 20 countries, culminating in 28 million articles. First, we track the volume of Covid-19 conversations across 20 countries from before to during the pandemic (Oct'19 to May'20). Second, we distill the phases of global pandemic narratives, and elucidate regional differences. METHODS: To track the volume of Covid-19 narratives, we identified 10 target terms-Coronavirus, Covid-19, Covid, nCoV, SARS-CoV-2, Wuhan Virus, Virus, Disease, Epidemic, Pandemic-and tracked their combined monthly prevalence across eight months from October 2019 through May 2020. Globally, across 20 countries, we identified 18,042,855 descriptors of the target terms. Further, these descriptors were analysed with natural language processing models to generate the top five topics of Covid-19 that were labelled by two independent researchers. This process was repeated across six continents to distil regional topics. RESULTS: Our model found four phases of online news media narratives: Pre-pandemic, Early, Peak and Recovery. Pre-pandemic narratives (Oct'19-Dec'19) were divergent across regions with Africa focused on monkeypox, Asia on dengue fever, and North America on Lyme disease and AIDS. Early (Jan-Feb'20) and Peak Pandemic (Mar-May'20) evidenced a global convergence, reflecting the omnipresence of Covid-19. The brief transition from early to peak pandemic narratives underscored the pandemic's rapid spread. Emerging from the embers of the pandemic's peak were nascent recovery words that are regionally divergent-Oceania focused on hope and an uncertain future while North America centered on re-opening the economy and tackling discrimination. CONCLUSIONS: Practically, we presented a media barometer of Covid-19, and provided a framework to analyse the pandemic's impact on societal perceptions-laying the important groundwork for policy makers to evaluate policy communications, and design risk communication strategies.

Culture Linked to Increasing Ageism During COVID-19: Evidence From a 10-Billion-Word Corpus Across 20 Countries.

OBJECTIVES: Older adults experience higher risks of getting severely ill from coronavirus disease 2019 (COVID-19), resulting in widespread narratives of frailty and vulnerability. We test: (a) whether global aging narratives have become more negative from before to during the pandemic (October 2019 to May 2020) across 20 countries; (b) model pandemic (incidence and mortality), and cultural factors associated with the trajectory of aging narratives. METHODS: We leveraged a 10-billion-word online-media corpus, consisting of 28 million newspaper and magazine articles across 20 countries, to identify nine common synonyms of "older adults" and compiled their most frequently used descriptors (collocates) from October 2019 to May 2020-culminating in 11,504 collocates that were rated to create a Cumulative Aging Narrative Score per month. Widely used cultural dimension scores were taken from Hofstede, and pandemic variables, from the Oxford COVID-19 Government Response Tracker. RESULTS: Aging narratives became more negative as the pandemic worsened across 20 countries. Globally, scores were trending neutral from October 2019 to February 2020, and plummeted in March 2020, reflecting COVID-19's severity. Prepandemic (October 2019), the United Kingdom evidenced the most negative aging narratives; peak pandemic (May 2020), South Africa took on the dubious honor. Across the 8-month period, the Philippines experienced the steepest trend toward negativity in aging narratives. Ageism, during the pandemic, was, ironically, not predicted by COVID-19's incidence and mortality rates, but by cultural variables: Individualism, Masculinity, Uncertainty Avoidance, and Long-term Orientation. DISCUSSION: The strategy to reverse this trajectory lay in the same phenomenon that promoted it: a sustained global campaign-though, it should be culturally nuanced and customized to a country's context.

Aging Narratives Over 210 Years (1810-2019).

OBJECTIVES: The World Health Organization launched a recent global campaign to combat ageism, citing its ubiquity and insidious threat to health. The historical context that promoted this pernicious threat is understudied, and such studies lay the critical foundation for designing societal-level campaigns to combat it. We analyzed the trend and content of aging narratives over 210 years across multiple genres-newspaper, magazines, fiction, nonfiction books-and modeled the predictors of the observed trend. METHOD: A 600-million-word dataset was created from the Corpus of Historical American English and the Corpus of Contemporary American English to form the largest structured historical corpus with over 150,000 texts from multiple genres. Computational linguistics and statistical techniques were applied to study the trend, content, and predictors of aging narratives. RESULTS: Aging narratives have become more negative, in a linear fashion (p = .003), over 210 years. There are distinct shifts: From uplifting narratives of heroism and kinship in the 1800s to darker tones of illness, death, and burden in the 1900s across newspapers, magazines, and nonfiction books. Fiction defied this trend by portraying older adults positively through romantic courtship and war heroism. Significant predictors of ageism over 210 years are the medicalization of aging, loss of status, warmth, competence, and social ostracism. DISCUSSION: Though it is unrealistic to reverse the course of ageism, its declining trajectory can be ameliorated. Our unprecedented study lay the groundwork for a societal-level campaign to tackle ageism. The need to act is more pressing given the Covid-19 pandemic where older adults are constantly portrayed as vulnerable.