Issues from Humdata Checklist

[collinschwantes]

HDX quality assurance (QA) officers should check each new or updated dataset against this checklist to ensure datasets shared on HDX meet the minimum HDX quality standards for datasets. The checklist is in three parts: a) a checklist for dataset resources, b) a data responsibility checklist and c) a metadata quality checklist. Each of these checklists are presented below.

A. Data Resource

• [x] Download link works
• [x] Link is not password protected
• [x] File is in a machine readable format
• [x] No PII in the data file
• [x] No DII ( need to confirm what DII means)
• [x] No individual survey data
• [x] stated file type and actual file type are consistent
• [x] if Zipped shaped, file is properly typed

B. Data responsibility

• [x] Data is safe to share and not potentially sensitive

C. Metadata Quality

• [x] Dataset has a title
• [x] title is consistent with contents
• [x] title is not in camel case
• [x] title does not start with a place name
• [x] title does not include a relative time reference (current, latest, new, old, etc)
• [x] dataset has a description
• [x] description reflects contents
• [x] No pii in the description - generic email addresses are okay (data@ecohealthalliance.org)
• [ ] Dataset has sources
• [ ] sources are valid
• [x] Dataset has a date of dataset attribute
• [x] date is not not in the future
• [ ] Dataset date is consistent with the freshness of the data - data are as up to date as suggested by the dataset date
• [x] the dataset has an expected update frequency
• [x] data are fresh in accordance with update frequency
• [x] The dataset has at least one location
• [x] Specified location matches location for the data
• [x] The dataset has a license
• [ ] The dataset has a methodology
• [ ] The caveats and comments text is clear and meaningful
• [x] If the dataset contains subnational data, has it been flagged?

[collinschwantes]

@sebaum Need to make sure that all datasets have appropriate source information. Many of the manually updated datasets do not contain a source URL or doi for the dataset. Ideally source points directly to the source information not an organization's homepage.

• [x] JEE Scores
• [x] Rabies Management CDC
• [x] Political Stability and Absence of Violence/ Terrorism - FAO?
• [x] Population growth, 2010-2019 (%) - world bank?
• [x] FAO Protein
• [ ] Combined data sheet_African Countries?
• [x] OIE simulation
• [x] eid_risk - from allen et al 2015
• [x] mammals_count - IUCN?

[sebaum]

JEE Scores - Came from each country's individual JEE report (https://extranet.who.int/sph/jee)

Rabies Management CDC - Came from country look up on CDC's National Rabies Assessment (https://www.cdc.gov/rabies/resources/countries-risk.html)

Political Stability and Absence of Violence/ Terrorism -- World Bank I believe (https://databank.worldbank.org/source/worldwide-governance-indicators/Series/PV.EST)

Population growth, 2010-2019 (%) - Calculated from original population data from world bank

FAO Protein - This captures "Average protein supply (g/cap/day) (3-year average)" and "Average supply of protein of animal origin (g/cap/day) (3-year average)" from FAO's Food Security Data. Average protein supply (g/cap/day) (3-year average): http://data.un.org/Data.aspx?q=Economy&d=FAO&f=itemCode%3A21013 Average supply of protein of animal origin (g/cap/day) (3-year average): http://data.un.org/Data.aspx?q=Average+protein+supply+(g%2Fcap%2Fday)+(3-year+average)&d=FAO&f=itemCode%3A21014

Combined data sheet_African Countries: This is an excel file that Catherine had put together for her dissertation, from which Sarah Hayes pulled in data. I think all the data here is account for except the "malaria incidence per 1000 people". I checked the World Bank data and it doesn't match what is in this dataset, so I will likely need to ask Catherine where she got it from.

OIE simulation -- (https://www.woah.org/en/what-we-do/animal-health-and-welfare/disease-data-collection/simulation-exercises/)

eid_risk - from allen et al 2015 This is from the Allen et al. (2015) repo. Emma had aggregated the data to the country level for another project and this data is included.

[collinschwantes]

@sebaum we can could also cite Catherine's dissertation if thats the source of the data

[sebaum]

@collinschwantes But it's not her own data though. Let me ask and see if she can get back to me quickly.

[collinschwantes]

@sebaum can't find a paper where Toph Allen was lead auther in 2015. Is this a different EHA Allen?

[sebaum]

@collinschwantes sorry collin, it's 2017. https://www.nature.com/articles/s41467-017-00923-8. It's the hotspots paper.

[collinschwantes]

Found the source with some help from Emma - https://github.com/ecohealthalliance/hotspots2/blob/master/inst/out/ncd-eid-link/hs2_country.csv

[collinschwantes]

@sebaum I requested permission to use IUCN redlist data and WOAH data from outside WAHIS.

[collinschwantes]

@sebaum see https://github.com/ecohealthalliance/gheri-africom-data/blob/feature/humdata_compliant/R/update_humdata_metadata.R - there are a couple of places where I need some help to fill out the metadata for HDX. (where ever there is an "@Sarah" placeholder ).

[collinschwantes]

• [ ] Check that lauren has access and see how to move ahead