# Data Records This page describes the data used in this project, specifically: where we collected the data from, some information about the data collection methods, any data processing which occurred to transform the raw data in deriving the variables which are viewable in the map. @[toc]( ) ---------- ## Local Support Variables ---------- ### WCVA Registered Volunteers **Definitions**: - *WCVA Registered Volunteers (per 100 pop)*: Number of volunteers who have signed up with the Welsh Council for Voluntary Action to provide voluntary support. **Source**: [Wales Council for Voluntary Action](https://wcva.cymru/) (WCVA). The original file was obtained on request and so only exists as a local copy. **Methods**: The counts are divided by the population size and multiplied by 100 to give the % volunteers in each geographical area. **Threats to validity:** - Counts of WCVA volunteers may not include volunteers registered with other third sector organisations. It is possible that more volunteers are registered in areas with a greater density of third sector organisations, which may not be reflected in this data. ---------- ### WCVA Increase in Volunteers **Definitions**: - *WCVA Increase in Volunteers (since March 2020, %)*: The number of volunteers who have signed up since the start of the COVID-19 pandemic is also given as a separate data layer designed to measure the number of volunteers likely involved in COVID-19 relief efforts. **Source**: [Wales Council for Voluntary Action](https://wcva.cymru/) (WCVA). The original file was obtained on request and so only exists as a local copy. **Methods**: The counts are divided by the population size and multiplied by 100 to give the % increase since March 2020 in each geographical area. **Threats to validity:** - Counts of WCVA volunteers may not include volunteers registered with other third sector organisations. It is possible that more volunteers are registered in areas with a greater density of third sector organisations, which may not be reflected in this data. ---------- ### Community Support Groups **Definitions**: - *Point Locations*: Location of identified community support groups. - *Community Support Groups (per 100 pop)*: A count of the number of groups is also given as a separate data layer as a percentage of the local population. **Source**: Data about known community support groups is collected from the [open source list provided by Police Rewired](https://www.policecoders.org/home/covid-19/communities) (COVID UK volunteer support v2, Police Rewired). It is a list of support group locations maintained by volunteers. They regularly scrape data from the following sources: - Mutual aid communities, crowdsourced by [COVID-19 Mutual Aid UK][3]. - Halo coronavirus communities, kindly provided by [Halo][4]. - Council community hubs (crowdsourced by [@boffbowsh][5]). For more information about their methodology, please see their [website][6]. **Methods**: Please see the outline of our methodology for obtaining and processing this data on the [Methods page](https://osf.io/c48hw/wiki/Methods/). **Threats to validity**: - Provides exact locations of community support groups, as registered by the groups themselves. - Not all community support groups will be registered through the sources used through this application. ---------- ### Twitter Community Support **Definition**: - *Twitter Community Support (estimated per 100 users)*: Number of users who were identified as having posted at least one tweet about community support, as a percentage of the total number of users who are estimated to have tweeted in that area. **Source**: Tweets from Wales, Twitter API **Method**: Twitter data was collected using the Twitter API. For more detailed information please see the [Methods](https://osf.io/c48hw/wiki/Methods/) page. **Threats to validity**: - Uses a probabilistic determination of where the tweet is coming from to aggregate and locate tweets to areas in Wales. - [We estimate that 15-29% of Welsh Tweets are in Welsh][17] (and this varies by area) and our keywords will not pick up these tweets - If the underlying determinants of increased twitter use decrease support (or vice versa), the model will produce a highly misleading description of the distribution of support. ---------- ### Sense of Community Belonging **Definition:** - *Sense of Community Belonging (per 100 pop)*: The percentage of people who agreed, or strongly agreed with the statement “I belong to my local area” within each geographical area. **Source**: National Survey for Wales, 2018-2019 (UK Data Service. SN: 8591). This data was obtained by request and so exists as a local copy. Methods can be downloaded from the [link][1]. Welsh Government, Office for National Statistics. (2020). National Survey for Wales, 2018-2019. [data collection]. UK Data Service. SN: 8591, http://doi.org/10.5255/UKDA-SN-8591-1 **Methods:** The question is a measure of self-reported sense of belonging in the community. Respondents were asked if they agreed or disagreed with the statement: "I belong to my local area." More information on the survey data collection is given [here][2]. **Threats to validity**: - The sample frame (random sample of households with someone over the age of 16) may not be representative of all at need groups (e.g. elderly not living at home). - Response rate was low, but weighted to be nationally representative. - Potential risk of information bias due to self-report. Standardised computer aided interviews should reduce interviewer bias. - Survey is over a year old, and may no longer be representative. ---------- ### Symptom Tracker: Can Count on Someone Close **Definition:** - *Symptom Tracker: Can Count on Someone Close (per 100 pop)*: The percentage of individuals who agreed that they could count on someone close to them if they need help. **Source**: This work uses anonymised data provided by participants of the [COVID-19 Symptoms Study](https://covid.joinzoe.com/) stored in the [Secure Anonymised Information Linkage (SAIL)](https://saildatabank.com/) Databank (project number 1095). **Methods:** This question is collected once from anybody who submits their data to the [COVID-19 Symptom Tracker](https://covid.joinzoe.com/) for the purposes of research. The data is collected using a smartphone app. The data we use is from up to 31st May 2020. **Threats to validity**: - The sample is likely to be limited to those who have access to a smartphone and the internet (though it is possible to submit data on behalf of someone else). - The response rate is not guaranteed to be representative of the population the data is from. ---------- ## Local Need Variables ---------- ### COVID-19 Cases **Definition**: - *COVID-19 Cases (per 100 pop)*: Confirmed cases of COVID-19 as a percentage of the local population. **Source**: Rapid COVID-19 surveillance run by Public Health Wales. Data collection methods are described by following the [link][7]. **Methods**: Individuals who have tested positive for COVID-19 are recorded as confirmed cases. We obtain this using a python script which remotely accesses and records data from the most recent testing episode. **Threats to validity**: - Will underestimate total number of cases as not everyone is tested. - Potential ascertainment/selection bias due to unequal probabilities of being tested. ---------- ### COVID-19 High Risk **Definition**: - *High Risk of COVID-19 (per 100 pop)*: Percentage of the local population who are at high risk (clinically extremely vulnerable) from coronavirus. **Source**: Public Health Wales, with permission of use granted by Welsh Government. The original datafile is not available for download, and only exists as a local copy. **Methods**: Data extract (updated on 03/04/2020) was provided directly by Public Health Wales with permission of Welsh Government, as a count of the number of people in each area who were identified as being at high risk (clinically extremely vulnerable), and were sent a letter advising them to self-isolate. Information on identifying vulnerable patient lists is given [here][8] and information on who is at higher risk from COVID-19 is given [here][9]. **Threats to validity**: - Assumes Public Health Wales has accurate information on all individuals to identify who is at high risk. This may create ascertainment bias. ---------- ### COVID-19 Moderate Risk **Definition**: - *Moderate Risk of COVID-19 (estimated per 100 pop)*: Percentage of the local population who are at moderate risk (clinically vulnerable) from coronavirus. **Source**: National Survey for Wales, 2018-2019 (UK Data Service. SN: 8591). This data was obtained by request and so exists as a local copy. Methods can be downloaded from the [link][10]. Welsh Government, Office for National Statistics. (2020). National Survey for Wales, 2018-2019. [data collection]. UK Data Service. SN: 8591, http://doi.org/10.5255/UKDA-SN-8591-1 **Methods**: We used responses to the National Survey for Wales as a proxy to approximate the population in each Local Authority that matches the NHS definition of who is at moderate risk (clinically vulnerable) from coronavirus. Note that this is not the same as the population who have been advised to shield themselves, who are at high risk (clinically extremely vulnerable). Definitions are given [here][11]. *Criteria*: The estimated vulnerable population is based on those who reported that they meet one of the following conditions: - 70+ - Asthmatic - Heart or circulatory illness - Respiratory system illness - Kidney complaints - Other digestive complaints including stomach, liver, pancreas etc - Learning disability - Diabetes (including hyperglycaemia) **Threats to validity**: - The sample frame (random sample of households with someone over the age of 16) may not be representative of all at need groups (e.g. elderly not living at home). - Response rate was low, but weighted to be nationally representative. - Potential risk of information bias due to self-report. Standardised computer aided interviews should reduce interviewer bias. - Survey is over a year old, and may no longer be representative. ---------- ### Over Age 65 **Definition**: - *Over Age 65 (per 100 pop)*: Percentage of the population who are 65+ years of age. **Source**: Mid-year population estimates. These are calculated by the Office for National Statistics. The survey data collection methods are described [here][18]. **Methods**: Data was collected from the following two sources: 1. [LA level data](https://statswales.gov.wales/Catalogue/Population-and-Migration/Population/Estimates/Local-Authority/populationestimates-by-localauthority-age) 2. [LSOA level data](https://statswales.gov.wales/Catalogue/Population-and-Migration/Population/Estimates/Small-Area/populationestimates-by-lowersuperoutputarea-agegroup) **Threats to validity**: - Numbers are based off modelled projections, which may not be perfect. Generally agree with census in years which the census is taken. ---------- ### Population Density **Definition**: - *Population Density (per sq. km)*: People per square kilometre. **Source**: Mid-year population estimates. These are calculated by the Office for National Statistics. The survey data collection methods are described [here][16]. **Methods**: Data was collected from the following two sources: 1. [LA level data](https://statswales.gov.wales/Catalogue/Population-and-Migration/Population/Density/populationdensity-by-localauthority-year) 2. [LSOA level data](https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/lowersuperoutputareapopulationdensity) **Threats to validity**: - Numbers are based on modelled projections, which may not be perfect. Generally agree with census in years which the census is taken. ---------- ### Most Deprived / Index of Multiple Deprivation **Definition**: - *Most Deprived (% areas in lowest quintile of deprivation)*: The % of LSOAs in the 20% most deprived in Wales in each LA. - *Index of Multiple Deprivation (rank)*: LSOA level rank of the Welsh Index of Multiple Deprivation. When visualised, a higher rank indicates higher relative deprivation (this is the reverse of the original order). **Source**: Welsh Index of Multiple Deprivation. This is the Welsh Government’s official measure of relative deprivation. Data was collected from the following two sources: 1. [LA level data](https://statswales.gov.wales/Catalogue/Community-Safety-and-Social-Inclusion/Welsh-Index-of-Multiple-Deprivation/WIMD-2019/localauthorityanalysis) 2. [LSOA level data](http://geoportal.statistics.gov.uk/datasets/b4fddfae9a0a4259ade22cab349208f7_0/data) **Methods**: Methods for generating the Welsh IMD are described [here][14]. **Threats to validity**: - Objectively defined and measured at LSOA level, using standardised methods. ---------- ### Digital Exclusion: No Internet Access **Definition**: - *Digital exclusion: No Internet Access (per 100 pop)*: Percentage of the local population without access to the internet. **Source**: National Survey for Wales, 2018-2019 (UK Data Service. SN: 8591). This data was obtained by request and so exists as a local copy. Methods can be downloaded from this [link](https://gov.wales/sites/default/files/statistics-and-research/2019-07/national-survey-for-wales-april-2018-to-march-2019-technical-report_0.pdf). Welsh Government, Office for National Statistics. (2020). National Survey for Wales, 2018-2019. [data collection]. UK Data Service. SN: 8591, http://doi.org/10.5255/UKDA-SN-8591-1 **Methods**: Information on the Internet and Media national survey can be found [here][13]. **Threats to validity**: - The sample frame (random sample of households with someone over the age of 16) may not be representative of all at need groups (e.g. elderly not living at home). - Response rate was low, but weighted to be nationally representative. - Potential risk of information bias due to self report. Standardised computer aided interviews should reduce interviewer bias. - Survey is over a year old, and may no longer be representative. ---------- ### Digital Exclusion: Not Registered with Online GP Services **Definition**: - *Digital Exclusion: Not Registered with Online GP Services (per 100 patients)*: Percentage of total patients who are not registered with GP online patient services. This is a measure of how likely a local population is to engage with digital and online services. **Source**: NHS Wales Informatics Service. This data was obtained by request and so exists as a local copy. **Methods**: Registered patients at GP surgeries who were signed up to "My Health Online" were counted. **Threats to validity**: - Possible bias from self-selection, e.g. requires access to digital/online services, if used as a proxy for something other than propensity to engage with NHS digital and online services. ---------- ## Other Data ---------- ### Population **Definition**: - Count of population (all ages) in each area. **Source**: Mid-year population estimates. These are calculated by the Office for National Statistics. The survey data collection methods are described [here][15]. **Methods**: Data was collected from the following two sources: 1. [LA level data](https://statswales.gov.wales/Catalogue/Population-and-Migration/Population/Estimates/Local-Authority/populationestimates-by-localauthority-age) 2. [LSOA level data](https://statswales.gov.wales/Catalogue/Population-and-Migration/Population/Estimates/Small-Area/populationestimates-by-lowersuperoutputarea-agegroup) **Threats to validity**: - Numbers are based off modelled projections, which may not be perfect. Generally agree with census in years which the census is taken. ---------- ### Geographical Boundaries<br> This data was collected from open source ONS Geography Open Data and NHS Digital data. <br> <br> **Map resolution**: <br> Government boundaries were used to accurately map areas of Wales at two resolutions: - [Local Authority ][21]('LA') - [Lower Layers Super Output Area][22] ('LSOA': [definition][19]) In producing this map some other resolutions and boundaries were used and translated to LA and LSOA for mapping: - [Local Health Board ][20]('LHB') - [GP Practice locations][24] - Point coordinates ---------- ## Acknowledgements and further information We have combined NHS and census data with online sources such as Twitter and Police Rewired. Most of the data used for the map is publicly available. Some data is not publicly available, but we always report our sources and the results from this data can still be viewed in the map. We kindly thank the [Wales Council for Voluntary Action][25], [Mutual Aid][26], [Welsh Government Statistics and Research][27] for collecting and making this data available. We also thank Open Street Map for their map of Wales, you can suggest improvements [here][28]. We are using the [Police Rewired community support dataset][29] to map community support groups. If you don't see your local group then you can add it by registering on the [COVID-19 Mutual Aid website][30]. Further information on our data pipeline, along with the sampling and data collection procedure of surveys is arrayed in our [data sources, methods and treatment spreadsheet][31]. ---------- [1]:https://gov.wales/sites/default/files/statistics-and-research/2019-07/national-survey-for-wales-april-2018-to-march-2019-technical-report_0.pdf [2]:https://gov.wales/sites/default/files/statistics-and-research/2019-07/national-survey-for-wales-april-2018-to-march-2019-technical-report_0.pdf [3]:https://covidmutualaid.org/ [4]:https://www.localhalo.com/coronavirus [5]: https://twitter.com/boffbowsh?lang=en [6]:https://www.policecoders.org/home/covid-19/communities [7]: http://www2.nphs.wales.nhs.uk:8080/CommunitySurveillanceDocs.nsf/61c1e930f9121fd080256f2a004937ed/77fdb9a33544aee88025855100300cab/$FILE/Rapid%20COVID-19%20surveillance%20data.xlsx [8]: https://nwis.nhs.wales/news/latest-news/identifying-vulnerable-patient-lists/ [9]: https://www.nhs.uk/conditions/coronavirus-covid-19/people-at-higher-risk-from-coronavirus/whos-at-higher-risk-from-coronavirus/ [10]:https://gov.wales/sites/default/files/statistics-and-research/2019-07/national-survey-for-wales-april-2018-to-march-2019-technical-report_0.pdf [11]:https://www.nhs.uk/conditions/coronavirus-covid-19/people-at-higher-risk-from-coronavirus/whos-at-higher-risk-from-coronavirus/ [12]:https://gov.wales/sites/default/files/statistics-and-research/2019-07/national-survey-for-wales-april-2018-to-march-2019-technical-report_0.pdf [13]: https://gov.wales/national-survey-wales-internet-and-media [14]:https://gov.wales/welsh-index-multiple-deprivation-index-guidance [15]:https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/methodologies/annualmidyearpopulationestimatesqmi [16]:https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/methodologies/annualmidyearpopulationestimatesqmi [17]: https://mfr.de-1.osf.io/render?url=https://osf.io/9cgs8/?direct%26mode=render%26action=download%26mode=render [18]:https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/methodologies/annualmidyearpopulationestimatesqmi [19]: https://www.datadictionary.nhs.uk/data_dictionary/nhs_business_definitions/l/lower_layer_super_output_area_de.asp?shownav=1 [20]:https://opendata.arcgis.com/datasets/473fed063ce84d98a2f7c4fb0b0ff320_0.geojson [21]:https://opendata.arcgis.com/datasets/3a4fa2ce68f642e399b4de07643eeed3_0.geojson [22]:https://geoportal.statistics.gov.uk/datasets/lower-layer-super-output-areas-december-2011-boundaries-ew-bsc [23]:http://geoportal.statistics.gov.uk/datasets/postcode-to-output-area-to-lower-layer-super-output-area-to-middle-layer-super-output-area-to-local-authority-district-february-2019-lookup-in-the-uk [24]:https://digital.nhs.uk/services/organisation-data-service/data-downloads/gp-and-gp-practice-related-data [25]: https://wcva.cymru/ [26]: https://covidmutualaid.org/ [27]: https://gov.wales/statistics-and-research [28]: https://www.openstreetmap.org/fixthemap [29]: https://www.policecoders.org/home/covid-19 [30]: https://covidmutualaid.org/local-groups/ [31]: https://osf.io/ue29y/