Here is a usefull links :http://www.dit.ie/library/ditlibrarycentralservices/researchsmarts/understandbibliometrics/https://www.library.auckland.ac.nz/subject-guides/med/pdfs/H-index%20and%20impact%20factors.pdf
h index is a simple and standard method described by Jorge Hirsch: http://www.pnas.org/content/102/46/16569
So, Scopus determines h the same way any other source does: take the collection of papers of interest (it does NOT have to be an individual author's papers - any collection of papers has an h index). Rank them in descending order by the count of citations. Find the highest numerical rank where the citation count exceeds the rank. That's h .
The intricacy of h-index, however, is that its precise value depends on several non-quality factors, including:
1. the dataset you are using (WoS, Scopus and Google Scholar all feature h-index) which will affect both the source items you can include and the citing item population that will be considered If you are comparing h-indexes, it is best to compare within a single resource.
2. the subject matter of the work - as with any non-normalized citation metric, the expected high/low value will be dependent on subject area. Medicine, biomedicine, physics, chemistry - all pretty high; engineering, agriculture, linguistics...lower; history, art, literary criticism - comparatively low. It is best to compare within similar subjects. There are some adaptations/enhancements to h that attempt to account for some aspects of this by including normalizations based on citation levels in the h-core.
3. Age of the corpus - Older things more likely to be cited- and more likely to have achieved near-final citation records. h-index can only increase over time, even if the body of work being considered does not add new items. Again- see some of the h index adaptations.
4. the effect of highly cited papers - There are two edge-cases that would give you an h-index of 10: 10 papers with exactly 10 citations each - and nothing else. (Total citation count for the body of work: 100); or 10 papers with 1000 citations each - and nothing else (Total citation count for the body of work: 10,000). Are these really equivalent academic records? And, you guessed it, there are a pack of h index adjustments that address this kind of thing too.
http://www.pnas.org/content/102/46/16569.
The index is based on a list of publications ranked in descending order by the number of citations these publications received. The value of h is equal to the number of papers (N) in the list that have N or more citations.
Here is a usefull links :http://www.dit.ie/library/ditlibrarycentralservices/researchsmarts/understandbibliometrics/https://www.library.auckland.ac.nz/subject-guides/med/pdfs/H-index%20and%20impact%20factors.pdf
h index is a simple and standard method described by Jorge Hirsch: http://www.pnas.org/content/102/46/16569
So, Scopus determines h the same way any other source does: take the collection of papers of interest (it does NOT have to be an individual author's papers - any collection of papers has an h index). Rank them in descending order by the count of citations. Find the highest numerical rank where the citation count exceeds the rank. That's h .
The intricacy of h-index, however, is that its precise value depends on several non-quality factors, including:
1. the dataset you are using (WoS, Scopus and Google Scholar all feature h-index) which will affect both the source items you can include and the citing item population that will be considered If you are comparing h-indexes, it is best to compare within a single resource.
2. the subject matter of the work - as with any non-normalized citation metric, the expected high/low value will be dependent on subject area. Medicine, biomedicine, physics, chemistry - all pretty high; engineering, agriculture, linguistics...lower; history, art, literary criticism - comparatively low. It is best to compare within similar subjects. There are some adaptations/enhancements to h that attempt to account for some aspects of this by including normalizations based on citation levels in the h-core.
3. Age of the corpus - Older things more likely to be cited- and more likely to have achieved near-final citation records. h-index can only increase over time, even if the body of work being considered does not add new items. Again- see some of the h index adaptations.
4. the effect of highly cited papers - There are two edge-cases that would give you an h-index of 10: 10 papers with exactly 10 citations each - and nothing else. (Total citation count for the body of work: 100); or 10 papers with 1000 citations each - and nothing else (Total citation count for the body of work: 10,000). Are these really equivalent academic records? And, you guessed it, there are a pack of h index adjustments that address this kind of thing too.
http://www.pnas.org/content/102/46/16569.
The index is based on a list of publications ranked in descending order by the number of citations these publications received. The value of h is equal to the number of papers (N) in the list that have N or more citations.