Difference between revisions of "Talk:Country codes"
(→Missing values for related entities, such as "population over 500K": new section) |
(→Country names (for Wesis upload): new section) |
||
Line 47: | Line 47: | ||
Entities like Hong Kong only appear on the "related entity" column and therefore, the following values are not given for it: population over 500K, start-end, wiki URL | Entities like Hong Kong only appear on the "related entity" column and therefore, the following values are not given for it: population over 500K, start-end, wiki URL | ||
+ | |||
+ | == Country names (for Wesis upload) == | ||
+ | |||
+ | We (A04) discovered when trying to upload data to Wesis that also the country names (not only cow codes) seem to be standardized according to how the name is written in cow. For other spelling of country names, an error is returned in the upload if we interpreted it correctly. Is that correct? Do we need to standardize the spelling of the country names as well? Another possibility would be to just add the respective country name to the cow code during the upload, but this maybe is too prone to mistakes? |
Revision as of 14:40, 12 December 2019
Shall we discuss COW and V-Dem definitions at length? Or just refer to them?
Shall we list countries in alphabetical order by continents? Or further refer to regions like V-Dem (e.g. Central Africa, East/Horn of Africa etc.)
-> I think it would be good if the countries are listed alphabetically but the table has a column for regions and sub-regions to filter them. I would suggest to use the UN defintions for regions and sub-regions, see https://unstats.un.org/unsd/methodology/m49/
ND: For the current testing, there is a "legacy" list in WeSIS that - unfortunately - does not recognize all "related entities"; we @A01 will fix it soon!
Contents
Spelling Mistake in Bosnia and Herzegovina
I think there is a spelling mistake in one of the entries in the country table of Bosnia and Herzegovina (o instead of e). I cannot edit the page, would be great if someone who has access can correct it. Thanks!
Code for Palestine
We suggest a code for Palestine: 667 Israel has 666, thats why we suggest this code
ND: The CoW code for Palestine is 665!
Code for Puerto Rico
We suggest a code for Puerto Rico: 43 Dominican Republic (regionally close) has 42
ND: The CoW code for Puerto Rico is 6!
Why ignore some countries
for example, we have data for North Vietnam during the split, why would we not incorporate this into Wesis? You have it listed as a related entry cow_code 817 there are other examples like this (eg., Serbia, Serbia and Montenegro)
ND: If there is a CoW code, you can upload it; in other words, 817 is valid for WeSIS!
Gran Colombia
It was a small empire including Ecuador, Panama and Venzuela <- these countries should be listed as 'related entities' no?
ND: But from what I can read of the CoW codebook this was prior to 1880, right? If so, it is missing as it is beyond the CRC's scope. If you, nevertheless want to upload data for it we should talk about adding it manually to the list of valid country-year observations. In this case just get in touch with us @A01.
East and West , North and South
Most of the data we have is radically different for East and West Germany. You have country codes for both (260 and 265), why not use them? Please let me upload data for both. We are throwing away both data and history by forcing Germany to be one code. Or please tell me how I am supposed to choose which data to upload. Should I upload data for East Germany for 1965 and have it coded as Germany (code number 355, i.e., West Germany)? This seems like a bad idea. Also if we ever go public with our data, this would really look not so good in my opinion.
ND: Of course you could use both - and should, as it was agreed upon to collect data as "disaggregated as possible". As 260 and 265 are valid CoW codes, you can upload data for both entities separately. In fact, because we rely on CoW, you can upload data for every entity that has a CoW code.
Entities like Hong Kong only appear on the "related entity" column and therefore, the following values are not given for it: population over 500K, start-end, wiki URL
Country names (for Wesis upload)
We (A04) discovered when trying to upload data to Wesis that also the country names (not only cow codes) seem to be standardized according to how the name is written in cow. For other spelling of country names, an error is returned in the upload if we interpreted it correctly. Is that correct? Do we need to standardize the spelling of the country names as well? Another possibility would be to just add the respective country name to the cow code during the upload, but this maybe is too prone to mistakes?