Peers in science and technology consulting all know this: when it comes to a talent policy or science and technology planning report, how persuasive it is often depends onthe density and quality of data points.
“The total number of high-level talent in China continues to grow.” With a sentence like this, the client’s follow-up questions are almost inevitable: How many? Based on what definition? When were the statistics compiled? If you cannot answer, the report’s credibility immediately drops to zero.
So let’s start with the conclusion:every data point in a report should satisfy the requirement of being “traceable.” This article uses the MuchSci talent profile as an example to explain two usable data dimensions for talent data in policy reports, along with three citation rules.
I. Two Data Dimensions
Data Dimension 1: Repository-scale statistics. The actual number and categorized counts of experts included in the repository across fields directly support statements such as “high-level talent reserves in a certain field.” Technically, it is worth noting that the statistics are generated by direct counting,unaffected by the 100-result limit per search—so even large-scale inventories can be calculated accurately.
Data Dimension 2: Categorized queries. View distributions across academic divisions, fields, and levels. Regional differences and structural differences can be directly incorporated into policy analysis frameworks to support judgments about “where the talent base is strong and where it is thin.”
II. Three Citation Rules
When data goes into a report, discipline matters more than the data itself:
First, use only public data. For anything not included in the profile, explicitly write “not currently provided by the profile”—consultants do not make estimates to fill gaps; this is the bottom line.
Second, specify definitions and timing. Clearly state the repository’s scope and the statistical reference time. Only reproducible data is qualified to go into a report.
Third, do not extrapolate conclusions. A distribution within the repository cannot support a nationwide judgment. “Experts in a certain field account for X% of the repository” is a fact; “talent in that field accounts for about X% nationally” crosses the line. Leave room for uncertainty when making statements.