China and the RS

[Loet Leydesdorff]

Dear Quentin, 

 

I take the freedom to forward to the list because I followed your advice
below and the results are interesting. 

 

Ø  Loet, Andrew

 

Ø  I am enjoying this exchange but could I please add a small remark.

 

Ø  When applying a simple regression model - linear, exponential or whatever
- estimates of unobserved values within the range of observation are not too
debatable. When used for extrapolation beyond the original range, it is
almost obligatory - or at least it was for my students! - to provide a
confidence  band for the expected value or a prediction band for the
observed value. Point predictions presume rather more accuracy than they
actually have!

 

Ø  BW

 

Ø  Quentin

 

Figure 1 for Scopus data (articles, reviews, and proceedings papers): 



 

Figure 2 for WoS data (articles, reviews, proceedings papers, and letters –
however, the letters do not change the curves:

 



 

Some notes: 

1. The percentages for 2010 in the Scopus set are uncertain because this is
based on publication years. In the WoS case, data collection was based on
calendar years.

2. The USA is weaker represented in the Scopus data, and China is weaker in
the ISI set;

3. The linear regression line does not fit the data in the case of the USA
(r2 = 0.015); I added the quadratic regression line which has a fit of
0.876, but of course does not have a further interpretation for the
extrapolation;

3. The differences between the two measurements are large (e.g., confidence
intervals do not overlap);

4. Scopus data can indeed be used to legitimate the conclusions of the Royal
Society report.

 

Best wishes, 

Loet

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.asis.org/pipermail/sigmetrics/attachments/20110404/4a2bd965/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: application/octet-stream
Size: 32856 bytes
Desc: image001.png
URL: <http://mail.asis.org/pipermail/sigmetrics/attachments/20110404/4a2bd965/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.png
Type: application/octet-stream
Size: 26840 bytes
Desc: image002.png
URL: <http://mail.asis.org/pipermail/sigmetrics/attachments/20110404/4a2bd965/attachment-0001.obj>