4 datasets found
  1. f

    Multinomial logistic regression parameter estimates.

    • figshare.com
    xls
    Updated Jun 16, 2023
    + more versions
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Sydney Banton; Michael von Massow; Júlia G. Pezzali; Adronie Verbrugghe; Anna K. Shoveller (2023). Multinomial logistic regression parameter estimates. [Dataset]. http://doi.org/10.1371/journal.pone.0272299.t001
    Explore at:
    xlsAvailable download formats
    Dataset updated
    Jun 16, 2023
    Dataset provided by
    PLOS ONE
    Authors
    Sydney Banton; Michael von Massow; Júlia G. Pezzali; Adronie Verbrugghe; Anna K. Shoveller
    License

    Attribution 4.0 (CC BY 4.0)https://creativecommons.org/licenses/by/4.0/
    License information was derived automatically

    Description

    Multinomial logistic regression parameter estimates.

  2. COVID-19 Trends and Impact Survey (2020-Ongoing) - Afghanistan, Albania,...

    • catalog.ihsn.org
    Updated Nov 3, 2021
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Facebook Data for Good (2021). COVID-19 Trends and Impact Survey (2020-Ongoing) - Afghanistan, Albania, Algeria, Angola, Argentina, Armenia, Australia, Austria, Azerbaijan, Banglades [Dataset]. https://catalog.ihsn.org/catalog/9884
    Explore at:
    Dataset updated
    Nov 3, 2021
    Dataset provided by
    Facebookhttps://www.fb.com/
    Carnegie Mellon University
    University of Maryland
    Time period covered
    2020 - 2021
    Area covered
    Angola, Azerbaijan, Bangladesh, Algeria, Afghanistan, Australia
    Description

    Abstract

    Facebook partners with academic institutions to support COVID-19 research and to help inform public health decisions. The COVID-19 Trends and Impact Survey (CTIS) is designed to help researchers better monitor and forecast the spread of COVID-19. Facebook invites app users in the United States to take the survey collected by faculty at Carnegie Mellon University (CMU) Delphi Research Center and users in more than 200 countries and territories globally to take the survey collected by faculty at the University of Maryland (UMD) Joint Program in Survey Methodology (JPSM). Sampled users see the invitation at the top of their News Feed, but the surveys are collected off the Facebook app and the Facebook company does not collect or receive survey responses. UMD and CMU (“survey host universities”) each partnered with the broader public health community to design the survey. The survey includes questions about COVID-19 vaccine acceptance, barriers to getting a vaccine, symptoms, preventive behaviors, access to care, social distancing behavior, mental health issues, socio-demographic characteristics and financial constraints. This information may help health systems plan where resources are needed and potentially when, where, and how to reopen parts of society.

    CMU and UMD aggregate survey responses at a subnational level and then publish the data publicly in APIs -- one for the United States and one for the rest of the world. Microdata is also available to nonprofits and universities through Facebook’s Data for Good program.

    Geographic coverage

    The surveys are fielded daily in over 200 countries and territories.

    Analysis unit

    • Public Aggregate Data: Subnational levels
    • Microdata through Facebook Data for Good program: Individual level

    Universe

    The survey was fielded to active Facebook users ages 18 and above.

    Kind of data

    Sample survey data [ssd]

    Sampling procedure

    CMU and UMD design, collect, and analyze the survey data. Facebook provides assistance with questionnaire translation, survey sampling and recruitment, and statistical bias correction. Facebook invites a new sample of adult users on the Facebook App to take the survey each day. These users see an invitation at the top of their Facebook News Feed to an optional, off Facebook survey. The sampled users can then choose whether or not to consent to the survey. If they consent, they are redirected to a Qualtrics survey hosted by UMD or CMU. The surveys are daily repeated cross-sections. Sampled users may be invited to take the survey again in either a few weeks or months, depending on the density of their area.

    We stratify the sample using administrative boundaries within countries and territories to provide geographic coverage. We are constantly working with the survey host universities to optimize the sampling design, including incorporating adaptive sampling, which could improve statistical power for local area estimates in priority areas as the pandemic progresses.

    The responses of sampled users who participate more than once will not be linked longitudinally. In order to enable an agile public health response, we aim to provide data that can detect either outbreaks or successful containment over time rather than cumulative or overall prevalence alone.

    Mode of data collection

    Internet [int]

    Research instrument

    CTIS includes questions about COVID-19 vaccine acceptance, barriers to getting a vaccine, symptoms, preventive behaviors, access to care, social distancing behavior, mental health issues, socio-demographic characteristics and financial constraints. The survey instruments are owned by the survey host universities and are available, along with their translations, with the data at the following links:

    https://gisumd.github.io/COVID-19-API-Documentation/docs/survey_instruments.html https://cmu-delphi.github.io/delphi-epidata/symptom-survey/coding.html https://umdsurvey.umd.edu/jfe/preview/SV_2mWYHEMq5ZoUBNj?Q_CHL=preview&Q_JFE=qdg https://cmu.ca1.qualtrics.com/jfe/preview/SV_cT2ri3tFp2dhJGZ?Q_SurveyVersionID=current&Q_CHL=preview

    Response rate

    Response rates to online surveys vary widely depending on a number of factors including survey length, region, strength of the relationship with invitees, incentive mechanisms, invite copy, interest of respondents in the topic and survey design.

    Sampling error estimates

    Any survey data is prone to several forms of error and biases that need to be considered to understand how closely the results reflect the intended population. Sampling error is a natural characteristic of every survey based on samples and reflects the uncertainty in any survey result that is attributable to the fact that not the whole population is surveyed.

    Facebook provides analytic weights that adjust for non-response and coverage biases. By non-response bias, Facebook means that some sampled users are more likely to respond to the survey than others. To adjust for this, Facebook calculates the inverse probability that sampled users complete the survey using their self-reported age and gender as well as other characteristics we know correlate with non-response. They then use these inverse probabilities to create weights for responses, after which the survey sample reflects the active adult user population on the Facebook app.

    By coverage bias, Facebook means that not everyone in every country has a Facebook app account or uses their account regularly. To adjust for this, Facebook adjusts the weights created in the first step even further so that the distribution of age, gender, and administrative region of residence in the survey sample reflects that of the general population. Making adjustments using the weights ensures that the sample more accurately reflects the characteristics of the target population represented.

    The weights will be available for the United States as well as 114 other countries and territories globally where we are able to generate high-quality weights. The current set of weighted countries and territories are listed on the next page. The set of countries and territories for which weights are available will be revised over the course of data collection as Facebook and the survey host universities evaluate sample coverage within each country. For more details about the weighting methodology and the general population benchmarks used, please see this weighting documentation: https://arxiv.org/abs/2009.14675

  3. o

    Data for: Thriving in a pandemic: determinants of excellent wellbeing among...

    • explore.openaire.eu
    • data.niaid.nih.gov
    • +1more
    Updated Feb 3, 2022
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Ben Beaglehole; Jonathan Williman; Caroline Bell; James Stanley; Matthew Jenkins; Philip Gendall; Janet Hoek; Charlene Rapsey; Susanna Every-Palmer (2022). Data for: Thriving in a pandemic: determinants of excellent wellbeing among New Zealanders during the 2020 COVID-19 lockdown; a cross-sectional survey [Dataset]. http://doi.org/10.5061/dryad.66t1g1k36
    Explore at:
    Dataset updated
    Feb 3, 2022
    Authors
    Ben Beaglehole; Jonathan Williman; Caroline Bell; James Stanley; Matthew Jenkins; Philip Gendall; Janet Hoek; Charlene Rapsey; Susanna Every-Palmer
    Area covered
    New Zealand
    Description

    For further details or analysis code for paper (R format) please contact james.stanley@otago.ac.nz. Data and data dictionaries are included in the uploaded files. The data were collected via two cross-sectional national surveys. For the majority of the analyses undertaken, this datasets from the two surveys was combined and survey weights were not used. However, for comparisons with the New Zealand General Social Survey the two datasets were analysed independently and post-stratification survey weights were used. Post-stratification survey weights are included with the datasets (used to re-scale collected data to the NZ population structure), but should only used if datasets are analysed independently. A total of n=3487 participants are included (meeting eligibility criteria and substantively completed the surveys). Data were originally collected in Qualtrics; cleaned and coded in R; and analysed using R. The "raw + clean" dataset provides both the original variables and the variables as analysed for the paper. Columns with names starting with "Q" are the original responses in the survey The "clean only" dataset provides just the variables analysed for the paper. Additional raw data and code available on request Some additional variables are available on request (with corresponding code) for data sensitivity or to reduce the complexity of the files (e.g. individual item responses on psychological scales). These are detailed below. WHO-5 (scores and categorised variables included; original per-item responses available on request with code for scoring) Age in years (excluded for data confidentiality) Age group (some people were only asked a "screening" question on age group for response quotas, others also provided age at end of survey) Gender (some people were only asked a "screening" question on gender for response quotas, others also confirmed gender at end of survey) Ethnicity (prioritised ethnicity included in file; original recorded ethnicity columns and free text fields available on request along with code for scoring) Alcohol intake before/during lockdown (summarised data included: original free-text columns and recoding step available on request with code for scoring) Cigarettes smoked each day/week (original free-text columns and recoding step available on request) Number of people living within bubble (original free-text column and recoding step available on request) Other main sources of stress (original free-text column available on request) Other prior trauma (original free-text column available on request) Objective: The COVID-19 pandemic and associated restrictions are associated with adverse psychological impacts but an assessment of positive wellbeing is required to understand the overall impacts of the pandemic. Methods: The NZ Lockdown Psychological Distress Survey measured excellent wellbeing categorised by a WHO-Five Well-being Index (WHO-5) score ≥22. The survey also contained demographic and pre-lockdown questions, subjective and objective lockdown experiences, and questions on alcohol use. The proportion of participants with excellent wellbeing is reported with multivariate analysis examining the relative importance of individual factors associated with excellent wellbeing. Results: Approximately 9% of the overall sample reported excellent wellbeing during the New Zealand lockdown. Excellent wellbeing status was associated with older age, male gender, Māori and Asian ethnicity, and lower levels of education. Excellent wellbeing was negatively associated with smoking, poor physical and mental health, and previous trauma. Conclusion: A substantial minority of New Zealanders reported excellent wellbeing during severe COVID-19 pandemic restrictions. Demographic and broader health factors predicted excellent wellbeing status. An understanding of these factors may help to enhance wellbeing during any future lockdowns. Methods of data collection are in the published paper and its parent (both open-access, see related works below). Briefly: Respondents were survey participants from an internet Panel survey firm. Data have been cleaned and processed: this was mostly simplifying/collapsing response options to fewer options for reporting.

  4. f

    Data supporting AHRC Project "Screen Encounters with Britain. What do young...

    • kcl.figshare.com
    Updated Feb 10, 2025
    Share
    FacebookFacebook
    TwitterTwitter
    Email
    Click to copy link
    Link copied
    Close
    Cite
    Jeanette Steemers; Andrea Esser; Matthew Hilborn (2025). Data supporting AHRC Project "Screen Encounters with Britain. What do young Europeans make of Britain and its digital screen culture? [Dataset]. http://doi.org/10.18742/22153928.v2
    Explore at:
    Dataset updated
    Feb 10, 2025
    Dataset provided by
    King's College London
    Authors
    Jeanette Steemers; Andrea Esser; Matthew Hilborn
    License

    Attribution 4.0 (CC BY 4.0)https://creativecommons.org/licenses/by/4.0/
    License information was derived automatically

    Area covered
    United Kingdom, Europe
    Description

    The data set includes:WP1 the catalogue & landing page research database - 1. the main catalogue database, which combines the Lumiere database , plus columns to indicate which titles were found during the Landing page research; 2. a list with titles found during the landing page research; 3. a list with findings from the JustWatch research; and 4 UK Titles mentioned by survey respondents, interviewees and in digital diariesWP2 Excerpts from buyer interviews, fully pseudonymised and consolidated within 1 document -WP2 Excerpts from cultural intermediary interviews, fully pseudonymised and consolidated within 1 documentWP3 fully pseudonymised and shortened individidual audience interviewsWP3 fully pseudonymised and shortened focus group interviews with 16-19 year oldsWP3 the survey for 4 countries with three sheets: 1. raw data as downloaded from Qualtrics; 2. the cleaned/translated and weighted database; 3. the weighting table -

  5. Not seeing a result you expected?
    Learn how you can add new datasets to our index.

Share
FacebookFacebook
TwitterTwitter
Email
Click to copy link
Link copied
Close
Cite
Sydney Banton; Michael von Massow; Júlia G. Pezzali; Adronie Verbrugghe; Anna K. Shoveller (2023). Multinomial logistic regression parameter estimates. [Dataset]. http://doi.org/10.1371/journal.pone.0272299.t001

Multinomial logistic regression parameter estimates.

Related Article
Explore at:
96 scholarly articles cite this dataset (View in Google Scholar)
xlsAvailable download formats
Dataset updated
Jun 16, 2023
Dataset provided by
PLOS ONE
Authors
Sydney Banton; Michael von Massow; Júlia G. Pezzali; Adronie Verbrugghe; Anna K. Shoveller
License

Attribution 4.0 (CC BY 4.0)https://creativecommons.org/licenses/by/4.0/
License information was derived automatically

Description

Multinomial logistic regression parameter estimates.

Search
Clear search
Close search
Google apps
Main menu