Language Bias Research Articles

Two decades of financial statement fraud detection literature review; combination of bibliometric analysis and topic modeling approach

PurposeThe emergence of machine learning has opened a new way for researchers. It allows them to supplement the traditional manual methods for conducting a literature review and turning it into smart literature. This study aims to present a framework for incorporating machine learning into financial statement fraud (FSF) literature analysis. This framework facilitates the analysis of a large amount of literature to show the trend of the field and identify the most productive authors, journals and potential areas for future research.Design/methodology/approachIn this study, a framework was introduced that merges bibliometric analysis techniques such as word frequency, co-word analysis and coauthorship analysis with the Latent Dirichlet Allocation topic modeling approach. This framework was used to uncover subtopics from 20 years of financial fraud research articles. Furthermore, the hierarchical clustering method was used on selected subtopics to demonstrate the primary contexts in the literature on FSF.FindingsThis study has contributed to the literature in two ways. First, this study has determined the top journals, articles, countries and keywords based on various bibliometric metrics. Second, using topic modeling and then hierarchy clustering, this study demonstrates the four primary contexts in FSF detection.Research limitations/implicationsIn this study, the authors tried to comprehensively view the studies related to financial fraud conducted over two decades. However, this research has limitations that can be an opportunity for future researchers. The first limitation is due to language bias. This study has focused on English language articles, so it is suggested that other researchers consider other languages as well. The second limitation is caused by citation bias. In this study, the authors tried to show the top articles based on the citation criteria. However, judging based on citation alone can be misleading. Therefore, this study suggests that the researchers consider other measures to check the citation quality and assess the studies’ precision by applying meta-analysis.Originality/valueDespite the popularity of bibliometric analysis and topic modeling, there have been limited efforts to use machine learning for literature review. This novel approach of using hierarchical clustering on topic modeling results enable us to uncover four primary contexts. Furthermore, this method allowed us to show the keywords of each context and highlight significant articles within each context.

A qualitative analysis of stigmatizing language in birth admission clinical notes.

The presence of stigmatizing language in the electronic health record (EHR) has been used to measure implicit biases that underlie health inequities. The purpose of this study was to identify the presence of stigmatizing language in the clinical notes of pregnant people during the birth admission. We conducted a qualitative analysis on N = 1117 birth admission EHR notes from two urban hospitals in 2017. We identified stigmatizing language categories, such as Disapproval (39.3%), Questioning patient credibility (37.7%), Difficult patient (21.3%), Stereotyping (1.6%), and Unilateral decisions (1.6%) in 61 notes (5.4%). We also defined a new stigmatizing language category indicating Power/privilege. This was present in 37 notes (3.3%) and signaled approval of social status, upholding a hierarchy of bias. The stigmatizing language was most frequently identified in birth admission triage notes (16%) and least frequently in social work initial assessments (13.7%). We found that clinicians from various disciplines recorded stigmatizing language in the medical records of birthing people. This language was used to question birthing people's credibility and convey disapproval of decision-making abilities for themselves or their newborns. We reported a Power/privilege language bias in the inconsistent documentation of traits considered favorable for patient outcomes (e.g., employment status). Future work on stigmatizing language may inform tailored interventions to improve perinatal outcomes for all birthing people and their families.

Discovering and Interpreting Biased Concepts in Online Communities

Language carries implicit human biases, functioning both as a reflection and a perpetuation of stereotypes that people carry with them. Recently, ML-based NLP methods such as word embeddings have been shown to learn such language biases with striking accuracy. This capability of word embeddings has been successfully exploited as a tool to quantify and study human biases. However, previous studies only consider a predefined set of biased concepts to attest (e.g., whether gender is more or less associated with particular jobs), or just discover biased words without helping to understand their meaning at the conceptual level. As such, these approaches can be either unable to find biased concepts that have not been defined in advance, or the biases they find are difficult to interpret and study. This could make existing approaches unsuitable to discover and interpret biases in online communities, as such communities may carry different biases than those in mainstream culture. This paper improves upon, extends, and evaluates our previous data-driven method to automatically discover and help interpret biased concepts encoded in word embeddings. We apply this approach to study the biased concepts present in the language used in online communities and experimentally show the validity and stability of our method

Language and Racial Capitalism: Linguistic Anthropological Approaches to Minoritized Labor, Standard Language, and Racial Bias

AbstractThis article examines racial capitalism from a semiotic perspective, arguing that economic value, like language and race, can be described in situated and indexical terms. I attempt to show how raciolinguistic bias in and around the workplace is linked to a larger labor market in which minoritized labor is reproduced in a systemic way, and to explore hegemonic formations of racialization in the workplace and beyond. The jumping‐off point for much of my argument is the work of the historian and political theorist Cedric Robinson.

Gender bias in the Chinese epicene pronoun ta

Unlike all the failed attempts proposed to mitigate gender bias in English, the Chinese language—which has an epicene pronoun, ta—took an opposite path: changes were implemented to make the third person pronoun gender specific by inserting a feminine third person pronoun and a non-human third person pronoun in its writing system (which did not affect speech). As a result, written Chinese became a mirror image of English (having the equivalents of he, she, and it). In this empirical study on gender bias in the Chinese language, we find that this institutional effort has also failed despite a century's implementation in the educational system of the country: the language exhibits a dominating male-as-norm bias as well as bias based on stereotyping, regardless of participants' gender and age groups. Our study therefore contributes to the understanding of language change as well as gender bias in language.

Multi‐lingual literature searches are needed to unveil global knowledge

Multi‐lingual literature searches are needed to unveil global knowledge

The efficacy of nutritional support and its role in the treatment of persons with sarcopenia

The aim of the current descriptive review was to identify the role of nutrition in the prevention and treatment of sarcopenia in the elderly. Most of the publications included in this review show that protein and vitamin D intake, as well as a healthy and balanced diet, plays an important protective role against the development of sarcopenia. However, current scientific evidence is insufficient to draw sound conclusions. Although the role of nutrition in sarcopenia has been assessed in scientific publications over the past decade, most of the studies conducted are highly heterogeneous and had small sample sizes. Larger meta-analyses and randomized trials with strict inclusion criteria must better describe the role of nutrition in the development and treatment of sarcopenia. In conclusion, general recommendations on nutrition are given. A literature search was conducted in three electronic databases PubMed, Cochrane Library, Scopus, for the period from 2012 to December 1, 2022. Secondary sources included articles cited in articles extracted from the above sources. Inclusion criteria were crossover or cohort studies involving individuals ≥60 years of age. There were no restrictions on the language bias of the publication. Search strategy: key words used to describe the condition of the participants were: «old age», «infirmity», and «sarcopenia».

Words matter: how ecologists discuss managed and non-managed bees and birds

Effectively promoting the stability and quality of ecosystem services involves the successful management of domesticated species and the control of introduced species. In the pollinator literature, interest and concern regarding pollinator species and pollinator health dramatically increased in recent years. Concurrently, the use of loaded terms when discussing domesticated and non-native species may have increased. As a result, pollinator ecology has inherited both the confusion associated with invasion biology’s lack of a standardized terminology to describe native, managed, or introduced species as well as loaded terms with very strong positive or negative connotations. The recent explosion of research on native bees and alternative pollinators, coupled with the use of loaded language, has led to a perceived divide between native bee and managed bee researchers. In comparison, the bird literature discusses the study of managed (poultry) and non-managed (all other birds) species without an apparent conflict with regard to the use of terms with strong connotations or sentiment. Here, we analyze word usage when discussing non-managed and managed bee and bird species in 3614 ecological and evolutionary biology papers published between 1990 and 2019. Using time series analyses, we demonstrate how the use of specific descriptor terms (such as wild, introduced, and exotic) changed over time. We then conducted co-citation network analyses to determine whether papers that share references have similar terminology and sentiment. We predicted a negative language bias towards introduced species and positive language bias towards native species. We found an association between the term invasive and bumble bees and we observed significant increases in the usage of more ambiguous terms to describe non-managed species, such as wild. We detected a negative sentiment associated with the research area of pathogen spillover in bumble bees, which corroborates the subjectivity that language carries. We recommend using terms that acknowledge the role of human activities on pathogen spillover and biological invasions. Avoiding the usage of loaded terms when discussing managed and non-managed species will advance our understanding and promote effective and productive communication across scientists, general public, policy makers and other stake holders in our society.

Validation of the cross-cultural dementia screening test in Alzheimer's disease and Parkinson's disease.

The Cross-Cultural Dementia (CCD) is a new screening tool to evaluate cognitive impairment based on a cross-cultural perspective to reduce the bias of education, and language and cultural differences. We aimed to evaluate the diagnostic properties of the CCD in Spaniards for the assessment of patients with Alzheimer's disease in mild cognitive impairment (AD-MCI) and mild dementia stages (AD-D) and patients with mild cognitive impairment associated with Parkinson's disease (PD-MCI). Sixty participants with AD (50% MCI) and thirty with PD-MCI were enrolled. Each clinical group was compared against a healthy control group (HC) with the same number of participants and no significant differences in age, education, and sex. A comprehensive neuropsychological test battery and CCD were completed. Intergroup comparisons, ROC curves, and cut-off scores were calculated for the study of diagnostic properties. Intergroup differences were found in accordance with the cognitive profile of each clinical condition. Memory measures (Objects test) were especially relevant for the classification between AD and HC. Memory and executive function scores (Sun-Moon and Dots tests) were useful in the case of PD-MCI and HC. Furthermore, CCD described differences in executive functions and speed scores comparing AD-MCI and PD-MCI. Correlations between standardized neuropsychological tests and CCD measures supported the convergent validity of the test. CCD showed good discrimination properties and cut-off scores for dementia and extended its application to a sample of prodromal stages of AD and PD with mild cognitive impairment.

Improving Patient-Centered Dementia Screening for General, Multicultural Population and Persons with Disabilities from Primary Care Professionals with a Web-Based App.

Primary care serves as the first point of contact for people with dementia and is therefore a promising setting for screening, assessment, and initiation of specific treatment and care. According to literature, online applications can be effective by addressing different needs, such as screening, health counseling, and improving overall health status. Our goal was to propose a brief, inexpensive, noninvasive strategy for screening dementia to general, multicultural population and persons with disabilities, through a web-based app with a tailored multicomponent design. We designed and developed a web-based application, which combines cognitive tests and biomarkers to assist primary care professionals screen dementia. We then conducted an implementation study to measure the usability of the app. Two groups of experts participated for the selection of the screening instruments, following the Delhi method. Then, 16 primary care professionals assessed the app to their patients (n=132), and after they measured its usability with System Usability Scale. Two cognitive tools were integrated in the app, GPCOG and RUDAS, which are adequate for primary care settings and for screening multicultural and special needs population, without educational or language bias. Also, for assessing biomarkers, the CAIDE model was preferred, which resulted in individualized proposals, concerning the modifiable risk factors. Usability scored high for the majority of users. Utilization of the Dementia app could be incorporated into the routine practices of existing healthcare services and screening of multiple population for dementia.

Image to audio, text to audio, text to speech, video to text conversion using, NLP techniques

Often, language bias between communicators can create communication problems. This article discusses a prototype that addresses this issue by enabling users to hear the content of text images. This involves extracting the text from the image and converting it into speech in the user's preferred language. Additionally, the device can be used by people with visual impairments. Overall, this device helps users to listen to the content of images being presented. The suggested system allows the user to take a picture, which is then scanned and analysed by the application to read the English text. The information obtained is then converted into voice, enabling visually impaired individuals to understand the content of the text. The output is delivered in speech format to provide access to information present on the document. To ensure better accuracy and performance, the system uses Natural Language Processing techniques. The system is designed with a Graphical User Interface (GUI) to improve accuracy and ease of use.

Analytical Review of Tax Compliance Studies in the SMEs Sector: A Bibliometric Approach

Small and medium-sized enterprises (SMEs) play a crucial role in fostering economic growth; however, they often face challenges in meeting their tax obligations due to limited resources and understanding. This bibliometric study explores the relationship between taxes and SMEs, exploring aspects such as compliance, the impact of tax burdens, and the effectiveness of incentives. The analysis of 783 Scopus documents reveals a growing interest in research, with consistent annual publications and significant citation rates. Among the fields, Business, Management, and Accounting lead in publications, closely followed by Economics, Econometrics, and Finance, also Social Sciences. Geographically, the United States, United Kingdom, Australia, South Africa, and Malaysia stand out as major contributors, underscoring the global nature of this engagement. Thematic analysis uncovers emerging and declining themes, including the cost of tax compliance, supply chain management, and country-specific SME issues. Niche themes focus on Russia’s business environment, tax avoidance, and access to finance in India. Basic themes cover SME growth, the impact of taxation, and responses to the challenges posed by the COVID-19 pandemic. Motor themes emphasize innovation in SMEs, Malaysia’s capital structure, and challenges related to tax compliance, also tax evasion. The research underscores the intricate relationship between taxes and SMEs, underscoring their global significance and the necessity for targeted policies. Despite limitations such as database representation, temporal constraints, and language bias, this study advocates for ongoing research and policy strategies. These should aim to support fair and sustainable tax obligations for SMEs, ensuring their continued growth and valuable contribution to the economy.

Multiview Language Bias Reduction for Visual Question Answering

Current Visual Question Answering models overly rely on language bias and fail to understand visual information sufficiently. Many recent works concentrate on mitigating the intra-question type bias (bias in the distribution of answers to a question type) without taking the inter-question type bias (bias in distribution between question types) into consideration, causing the model to ignore the tail question types. In addition, they neglect the overall distribution bias of the answer set, leading to the model only focusing on the head answers. In this paper, we propose the <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</b> ulti- <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">V</b> iew <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</b> anguage <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">B</b> ias <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</b> eduction (MVBR) method to solve these problems. For the inter-question type bias, we introduce the <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I</b> nter- <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Q</b> uestion <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</b> ype <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">B</b> ias (IQTB) module. IQTB exploits the question type distribution of the training set to determine the question type bias which is used to generate weighting factors to reshape the loss of each question type into a balanced form. For the overall distribution bias of the answer set, we utilize the <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</b> ecoupled <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</b> raining (DT) module. The DT module penalizes the weights of each class in the classifier and forces their distribution to be balanced. Experimental results demonstrate that MVBR is effective on VQA-CP v2 and effectively improves performance on the mainstream models.

Protocol for a scoping review of potential vaccine candidates predicted by VaxiJen for different viral pathogens between 2017–2021

BackgroundVaccination is essential for the prevention of infectious diseases and has helped to reduce disease-related mortality, such as pneumonia. However, traditional vaccine development is time-consuming and risky. Reverse vaccinology (RV) is a promising alternative to developing vaccines based on the in silico discovery of antigens, often termed ‘potential vaccine candidates’ (PVCs), using a pathogen’s proteome. RV prediction technologies, such as VaxiJen (founded in 2007), are used to take the first step toward vaccine development. VaxiJen is used by researchers to identify PVCs for various diseases. A 10-year review of these PVCs was published in 2017. There has since been no review of viral PVCs predicted by VaxiJen from 2017 to 2021. The proposed scoping review aims to address this gap.MethodsThis protocol is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) 2015 checklist. The review will employ Arksey and O’Malley’s five-stage methodological framework, which was later enhanced by Levac et al. and the Joanna Briggs Institute (JBI). The PRISMA extension for Scoping Reviews (PRISMA-ScR) reporting guideline will be utilized with this framework. PubMed, Scopus, Web of Science, EBSCOhost, and ProQuest One Academic will be searched using the term ‘vaxijen’. The inclusion criteria will be English-only full-text original articles published in peer-reviewed journals and unpublished papers from 2017 to 2021. Rayyan will be used to deduplicate, screen titles and abstracts of articles. The articles’ full texts will be examined. The data will be extracted using Microsoft Excel. Using a data charting form, data will be sifted and organized by key categories and themes.DiscussionThis protocol was submitted for publication and went through an extensive peer review process. The review has implications for novel vaccine development against various viruses. The key limitation of this study is language bias due to the selection of English-only papers because of limited resources. This study will not require ethical clearance since it will use secondary data and will not include patients. Nevertheless, this research is part of a larger project that was submitted for ethical consideration to the Biomedical Research Ethics Committee of the University of KwaZulu-Natal in South Africa. This study’s findings will be published in a peer-reviewed journal and provided to relevant stakeholders.Systematic review registrationOpen Science Framework (OSF): https://osf.io/ht8wr

Determining the safety and effectiveness of Tai Chi: a critical overview of 210 systematic reviews of controlled clinical trials

BackgroundThis overview summarizes the best available systematic review (SR) evidence on the health effects of Tai Chi.MethodsNine databases (PubMed, Cochrane Library, EMBASE, Medline, Web of Science, China National Knowledge Infrastructure (CNKI), Chinese Scientific Journal Database (VIP), Sino-Med, and Wanfang Database) were searched for SRs of controlled clinical trials of Tai Chi interventions published between Jan 2010 and Dec 2020 in any language. Effect estimates were extracted from the most recent, comprehensive, highest-quality SR for each population, condition, and outcome. SR quality was appraised with AMSTAR 2 and overall certainty of effect estimates with the GRADE method.ResultsOf the 210 included SRs, 193 only included randomized controlled trials, one only included non-randomized studies of interventions, and 16 included both. Common conditions were neurological (18.6%), falls/balance (14.7%), cardiovascular (14.7%), musculoskeletal (11.0%), cancer (7.1%), and diabetes mellitus (6.7%). Except for stroke, no evidence for disease prevention was found; however, multiple proxy-outcomes/risks factors were evaluated. One hundred and fourteen effect estimates were extracted from 37 SRs (2 high, 6 moderate, 18 low, and 11 critically low quality), representing 59,306 adults. Compared to active and/or inactive controls, 66 of the 114 effect estimates reported clinically important benefits from Tai Chi, 53 reported an equivalent or marginal benefit, and 6 an equivalent risk of adverse events. Eight of the 114 effect estimates (7.0%) were rated as high, 43 (37.7%) moderate, 36 (31.6%) low, and 27 (23.7%) very low certainty evidence due to concerns with risk of bias (92/114, 80.7%), imprecision (43/114, 37.7%), inconsistency (37/114, 32.5%), and publication bias (3/114, 2.6%). SR quality was often limited by the search strategies, language bias, inadequate consideration of clinical, methodological, and statistical heterogeneity, poor reporting standards, and/or no registered SR protocol.ConclusionsThe findings suggest Tai Chi has multidimensional effects, including physical, psychological and quality of life benefits for a wide range of conditions, as well as multimorbidity. Clinically important benefits were most consistently reported for Parkinson’s disease, falls risk, knee osteoarthritis, low back pain, cerebrovascular, and cardiovascular diseases including hypertension. For most conditions, higher-quality SRs with rigorous primary studies are required.Systematic review registrationPROSPERO CRD42021225708.

Assessing machine learning for fair prediction of ADHD in school pupils using a retrospective cohort study of linked education and healthcare data

ObjectivesAttention deficit hyperactivity disorder (ADHD) is a prevalent childhood disorder, but often goes unrecognised and untreated. To improve access to services, accurate predictions of populations at high risk of ADHD...

The Screen of Cancer Survivorship-Occupational Therapy Services (SOCS-OTS): A Delphi Study for Tool Validation

<h3>Research Objectives</h3> To formally validate The Screening of Cancer Survivorship-Occupational Therapy Services (SOCS-OTS) which researchers developed for oncology team members to identify the need for OT services. <h3>Design</h3> Researchers used a classic Delphi methodology to develop and validate the SOCS-OTS. In Round 1, a blend of deductive and inductive approaches guided item writing. The deductive approach was used to thematically analyze oncology literature using the OT Practice Framework to guide item writing. An inductive approach was then used to verify items were issues relevant to expert panelists living with and beyond cancer. In Rounds 2-4, researchers assessed 14 expert panelists' opinions to indicate which items from Round 1 met consensus for inclusion on the SOCS-OTS. <h3>Setting</h3> Content experts were survivors were living in the community and expert panelists were from various practice settings, healthcare organizations, and oncology research labs. <h3>Participants</h3> Content experts varied in age, cancer type, and activity limitations. Inclusion criteria consisted of individuals diagnosed with cancer, 18 + years, fluent in English, with access to a computer with internet. Expert panelists included OTs, OTAs, and OT researchers. Inclusion of OT practitioners required panelists to have five years of experience with a 30% caseload of survivors. OT researchers required a published oncology-related article. <h3>Interventions</h3> N/A. <h3>Main Outcome Measures</h3> Items that did not reach 80% consensus were removed from the final tool. <h3>Results</h3> The final rendition of the SOCS-OTS included 20 items that met consensus. Researchers modified and removed several items from the tool including modifying the tool's instructions for use, rewording items for clarity, readability, or appropriateness for the oncology population, eliminating biased language, and making changes to the tools rating scale. The most highly rated items on the final SOCS-OTS included bathing and dressing, feeding/eating, social participation, sexual activity and intimacy, and community mobility. <h3>Conclusions</h3> The SOCS-OTS is a content valid screening tool that can be used by cancer care teams to indicate the need for OT services. Further studies are warranted for the development of its psychometric properties. <h3>Author(s) Disclosures</h3> No conflicts of interest

Influence of cultural background on tourist flow process and outcomes: an empirical study with Chinese-speaking tourists in Japan

PurposeThis paper aims to fill a research gap on whether cultural background (Eastern or Western) influences tourist flow experiences and outcomes with Chinese-speaking tourists.Design/methodology/approachThe authors conducted flow-aware, guided cultural tours in Kyoto, Japan, from March 2019 to February 2020. Tourists visited the same spots on the same route with the same guide to avoid discrepancies stemming from experiencing different tours with different guides. Professional photoshoots were conducted to study the flow experience under high challenges. Participants were Chinese-speaking tourists worldwide. The guide spoke in Chinese to minimise language bias. Participants were asked to complete an anonymous questionnaire within one week of the tours.FindingsThe authors received 191 valid responses. Those from the Eastern regions showed significantly lower flow states and outcomes than their counterparts from the Western regions under high challenges, whereas this phenomenon was not observed under low challenges. A positive correlation between the flow state and outcomes was observed in both regional groups.Originality/valueTo the best of the authors’ knowledge, this is the first study to apply flow theory to guided cultural tours with Chinese-speaking tourists from different cultural backgrounds. It was found that those from the Eastern regions can have flow experience and good outcomes in guided tours, although they may require a lower level of challenge. A managerial implication is that cultural tours should be designed considering the balance between challenges and tourists’ cultural background to optimise their flow experiences and outcomes.

Unidirectional language bias: The Implicit Association Test with Spanish and English in Miami

Aims and objectives/purpose/research questions: Automatic preferences and implicit biases toward Spanish and English in the context of Miami are tested. We investigate whether implicit language bias is conditioned by social factors (i.e., ethnicity, gender, first language) as well as different linguistic stimuli (orthographic vs oral/aural). Design/methodology/approach: Three Implicit Association Test (IAT) experiments were designed and implemented. IAT 1 uses semantically neutral furniture terms in English and Spanish with visual attribute words. IAT 2 uses an oral/aural condition employing US Latinx city names produced with both Spanish and English phonology. In IAT 3, we use nonce words to represent the categories of English and Spanish as well as positive/negative visual imagery for the attributes. Data and analysis: d-scores were automatically calculated considering response latency and accuracy. Data were subjected to linear regression for between- and within-experiment analyses as well as a post hoc regression considering participants born outside of the United States. Findings/conclusions: The within-experiment analysis for IAT 1 showed an effect of country of birth. The between-experiment analysis showed country of birth and the experiments themselves to be significant. Length of residency for the non-US born group was also a significant factor. Across the three experiments, none of the 89 participants demonstrated a strong automatic preference for Spanish. Originality: This study is conducted in the largest situation of bilingualism in the Americas and is the first study to examine nonconscious bias toward bilingual participants’ own language varieties in the United States. Significance/implications: Despite the popular construction of Miami as “Spanish-speaking,” results of this study demonstrate strong nonconscious automated preferences for English among participants in Miami and raise questions about language, power, and the vitality of Spanish in South Florida.

China futures price forecasting based on online search and information transfer

The synchronicity effect between the financial market and online response for time-series forecasting is an important task with wide applications. This study combines data from the Baidu index (BDI), Google trends (GT), and transfer entropy (TE) to forecast a wide range of futures prices with a focus on China. A forecasting model based on a hybrid gray wolf optimizer (GWO), convolutional neural network (CNN), and long short-term memory (LSTM) is developed. First, Baidu and Google dual-platform search data were selected and constructed as Internet-based consumer price index (ICPI) using principal component analysis. Second, TE is used to quantify the information between online behavior and futures markets. Finally, the effective Internet-based consumer price index (ICPI) and TE are introduced into the GWO-CNN-LSTM model to forecast the daily prices of corn, soybean, polyvinyl chloride (PVC), egg, and rebar futures. The results show that the GWO-CNN-LSTM model has a significant improvement in predicting future prices. Internet-based CPI built on Baidu and Google platforms has a high degree of real-time performance and reduces the platform and language bias of the search data. Our proposed framework can provide predictive decision support for government leaders, market investors, and production activities.