Hello again Jenny,

 

I am glad it could be of help. I have been using SciVal/Scopus a lot for about a year now and have become very familiar with things. Charles Martinez from Elsevier has been a really good support whilst learning to use the product (and still is amazing at helping me with things/questions). I am also happy to help you with interpreting the metrics if you want some guidance on that.

 

For small departments I would investigate a bit further, look at citations per publication, see if there are any publications skewing the results. Try to understand the subject area a bit before reporting. See if the subject areas that make up the FWCI accurately represents the department. I think it is still okay to report the FWCI for small departments as long as you explain what it means and its limitations.

 

We currently do not have a standard template but would advise to work closely with departments whilst looking at metrics to make sure they understand the data you provide them with. I think we might be looking to develop a template in the future but we are still working on it J

 

Please feel free to email my personal email ([log in to unmask]) if you would like some help with interpretation of the data.

 

Kind regards,

Kat

 

Katrine Sundsbo

Scholarly Communications and Research Support Manager

Library Services

University of Essex

 

T 01206 873153

E [log in to unmask]

► libwww.essex.ac.uk

katrinesundsbo

 

WE ARE ESSEX

TOP 20 FOR RESEARCH EXCELLENCE

TEF GOLD 2017

 

 

 

 

 

 

Hi Kat

 

This is really timely for me thank you as I have just been asked to provide something similar for one of our Schools here, to try and give a flavour of where they are at, to identify strengths and collaborations etc.  We are very new to using SciVal and we don’t have a research team to support it so I’m trying to do my best in between my other work.  One of the things that concerns me is that I can use the system and produce the reports but it’s being able to properly understand the metrics and explain them and, as you point out, be able to identify any instabilities in the data that could give a false impression.  So for example, for small groups, is there a better metric to use than FWCI?

 

Also, have you, or anyone else out there created a standard template that you use for reporting to departments, and if so would anyone be prepared to share?

 

Jenny

 

 

 

 

 

 

 

Hello Maria,

 

Here at the University of Essex we have been using Scopus/SciVal.

 

Institution

For this we use the institution groups created by SciVal. These are created based on the affiliation details researchers put in, so there are some output that is not included as there are misspellings or cases where people don’t add an affiliation. However, I’ve added an alert to try to ‘catch’ and add the misspelled versions (there are a few strange ones. Recently had a “University of Essex, Canada” which wasn’t recognised as belonging to us, so this one I had to add manually via the web feedback form.

 

For the whole institution we use scholarly output, Number of citations, FWCI (Field-weighted citation impact, where a score of 1 means that the document/research group/institutions is receiving as many citations as expected, compared to other similar document in the same subject area) and have proposed to use citations per publication as well though we are not doing that currently.  We also use this for benchmarking against other institutions.

 

Departments

We have created groups on SciVal using information from HR so that we can get an overview of the departments. Again this is not accurate as it brings all publications from researchers, regardless of their affiliations. So there is some ‘dirty data’ in there but we explain that to the departments. We explore the same data here as institutions.

 

We’ve tried doing the same for benchmarking other institutions, but it has been very time consuming… and departments usually differ a lot between institutions in terms of research areas (e.g. our Mathematics Department does not include any physics, but several others have physics within this department so our scores are a lot lower as there are more citations within physics.  Because of this these scores will probably always be lower and benchmarking becomes a bit pointless)

 

We also check for stability in our data, e.g. FWCI can be very unstable for small groups so we check the FWCI by each year to see if there are any peaks. If there are we explore these before reporting. I wrote up a report a year ago and we had a sudden peak in our most recent data because of 5 highly cited documents on the same topic in a small group of publications. In this case I highlighted the data and explained it to make sure stake holders did not interpret any information incorrectly.

 

I think no matter what you use, explaining anything strange or unusual in your data is key. Always explain what’s going on and what it means.

 

Hope this helps.

 

Kind regards,

Kat

 

Katrine Sundsbo

Scholarly Communications and Research Support Manager

Library Services

University of Essex

 

T 01206 873153

E [log in to unmask]

► libwww.essex.ac.uk

katrinesundsbo

 

WE ARE ESSEX

TOP 20 FOR RESEARCH EXCELLENCE

TEF GOLD 2017

 

 

 

 

 

From: A bibliometrics discussion list for the Library and Research Community [mailto:[log in to unmask]] On Behalf Of maria c
Sent: 09 March 2018 08:44
To: [log in to unmask]
Subject: Report on bibliometric research outputs