Dear Ones:
Since last January we have been doing a month-by-month simulation of the JCR Communication journals to see how close we could get to the final JCR data. As you well know, JCR 2018 was published yesterday, and we have important discrepancies regarding what we have been doing with data obtained from Web of Science. About the denominator to calculate the IF, the numbers coincide, there is no difference here. However, in the numerator for this calculation there are important discrepancies in the number of citations of many of the journals in this category.
However, in the numerator for this calculation there are significant discrepancies in the citation count of many of the journals in this category. JCR identifies for a large number of journals more citations than the Web of Science is able to detect.
Perhaps it is our mistake: The methodology we have been using to calculate the impact factor is the following: At WoS we do an advanced search for the Communication category in the 2016-2017 range of Social Sciences Citation Index (SSCI). Then, we refine by document type by selecting article and review. When analyzing results by source title, it shows us the number of documents per journal. In this way, we get the denominator to calculate the IF. This number in all cases matches our estimations.
To obtain the numerator we take into account all types of documents published in both 2016 and 2017 for the counting of citations. Then, by selecting each source title one by one, we access the citation report for each of them. In the report, we obtain the citations issued in 2018 that appear in the graph provided by WoS. This is the number that does not match in most cases (it only matches in 23 of the 88 journals) with the number given by JCR used as numerator. These discrepancies reach more than 80 citations in some cases (e.g. New Media Studies).
Below, you can see in these links the evolution that we have been making and the difference with the JCR published as FI and, therefore, to the final positions of the journals. The text is in Spanish, but you can see the evolution of the FI in the ranking.
January
http://aliciamorenodelgado.com/index.php/2019/01/17/factor-de-impacto-provisional-de-las-revistas-de-comunicacion/
February
http://aliciamorenodelgado.com/index.php/2019/02/19/las-revistas-espanolas-de-comunicacion-en-jcr-mejoran-sus-posiciones-con-los-datos-de-febrero/
March:
http://aliciamorenodelgado.com/index.php/2019/03/21/778/
April
http://aliciamorenodelgado.com/index.php/2019/04/23/803/
May
http://aliciamorenodelgado.com/index.php/2019/05/24/853/
June
http://aliciamorenodelgado.com/index.php/2019/06/17/los-ultimos-datos-de-junio-alteran-las-posiciones-de-las-revistas-del-area-de-comunicacion-en-jcr/
JCR2018:
http://aliciamorenodelgado.com/index.php/2019/06/20/clarivate-analytics-publica-el-journal-citation-reports-2018/
Could someone tell me if we have done the counting of citations in a wrong way or why the WoS ones do not match the ones offered by JCR 2018?
We are thinking in two options:
1. We have made a mistake.
2. JCR is done with a database that is more updated than the one currently shown in WoS.
Thank you very much.
Best regards,
Alicia.
########################################################################
To unsubscribe from the LIS-BIBLIOMETRICS list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=LIS-BIBLIOMETRICS&A=1
|