ORCID.DataExplained

原帖：

学术圈都在往哪润？来自 1847 万名学者 1145 万次的用脚投票

终极赛博斗兽棋之终结一切院校排名的院校排名

盘点老中老印最好留的硕士项目（附国家与地区层级数据）

This is to analyze the annual released ORCID Public Data File

Sorry I chose thousandths place as scale. I would like to regenerate the graph later.

Here is the explain:

• When the value comes >1,000, it means the country has a net inflow of scholars, meaning the country intakes more scholars than it loses
• When the value comes <1,000, it means the net outflow of scholars >0%, meaning the country intakes less scholars than it loses
• When the value comes =1,000, it means the country stands at a balance of the scholars intake and lose

For the scholars:

• A scholar is a person who has registered on ORCID (whose database I have used in these graphs), which is a website for academic authors and contributers. More info on Wikipedia
• The scholars' data I used is based on ORCID Public Data File, which has 11 million flows from 18 million scholars

For every country:

• Ratio counted without self flow = Total inflow count / Total outflow count
• Ratio counted with self flow = (Total inflow count + Total self flow count) / (Total outflow count + Total self flow count)

For every flow:

• An inflow means a scholar chose the inflow country as the destination country
• An outflow means a scholar made the flow from the outflow country as the origin country
• An self flow means the inflow country is the same as the outflow country, internal flow in other words

For every flow count:

• A Count without HDI adjusted = A scholar move once from a position to a position
• A HDI adjusted count = A Count without HDI adjusted * (Outflow country HDI / Inflow country HDI)

HDI adjusted count can make a flow having a higher or lower count with high HDI difference, which can eliminate attraction from the difference of living condition to some extent.

GitHub Page: ORCID.DataExplained

Where you can download all the data the graph based on.

Source:

• ORCID Public Data File 2023
• UNDP HDI 2022
• HDI for Taiwan and Macau

Tool:

• Program by Python, released on GitHub
• IMAGE
• Watermark tool

HDI 加权算法：流动数 = 流出国HDI / 流入国HDI

如此一来当学者从 HDI 高的国家流动到 HDI 低的国家时能有更高的加成，反之亦反，以此可以更清晰地划分层次

例如：同样是流入，但是挪威的流入多为发达国家，即 HDI 高的国家的流入，而法德则多为发展中国家，即 HDI 低的国家的流入，所以在加权后挪威的指标可以明显和法德拉开差距