I have a similar issue with the search box in Elsevier’s ClinicalKey. It’s almost impossible to search for a specific author and, again, you have to put all your search terms in the single box.

 

I have been asking for a more advanced search engine ever since we first subscribed to ClinicalKey (which has full text books as well as journals) to no avail.

 

 

 

Hello,

 

I promised to collate the replies I got to my question about the new search functionality in Science Direct. A similar question was also posed on the LIS-MEDICAL list by a colleague, so this this a summary of replies from both lists.

 

Firstly, thank you to everyone who replied to these questions and provided feedback and examples.

 

One question asked was whether researchers use databases or indexes to do systematic searching , rather than publisher platforms. From the responses, I would say that the ability to search the big publisher platforms like Science Direct in a systematic way is still valued, perhaps more so in some subject areas than others. This ability provides a ‘belt & braces’ approach - researchers and students would like to be able to employ the same search strategy that they’ve used in Medline, for example, in Science Direct to check that they have found everything of relevance. In many subject areas, students are required to demonstrate their systematic searching techniques, so it’s important that such a search can be done and the process understood.

 

Onto the feedback. Overall the feedback was negative. Some quotes:

 

-I’m really disappointed by the way they’ve reduced the search functionality and removed previously available personalisation features, particularly the facility to combine searches – I think the expert search was a replacement for this but that’s also gone now. As far as I can see it makes any kind of systematic searching impossible.

-It is so atrocious that rather than demonstrating it to students I put up a screenshot of the search page annotated with something along the lines of “Don’t bother using this - it isn’t powerful enough to meet your needs”.

-Imagine my disappointment when they released the new interface. It really is not very intuitive.

-ScienceDirect used to be the go-to resource that I would recommend for both my Psychology and Speech therapy students. Since Elsevier have made these changes, I have found it really difficult to use for Advanced searches, as the page doesn’t match the ‘standard’ that we have come to expect from other databases.

 

There was more detailed feedback on some of the specific functionality and features:

 

-the Advanced search doesn’t support wildcards or truncation searches. (One respondent had been told by Elsevier that by the end of 2018, the new advanced search will start supporting wildcards and proximity operators, so it might be worth checking this again)

-Phrase searching also seems entirely optional and random (eg. The search was asking for either “child abuse” OR “child neglect” but found the word ‘child’ on its own – and this was given a high ranking (it was the second result)

-the Advanced search feature also limits students and staff to 8 Boolean connectors which for many is not enough.

-we had to include search brackets within the single search box – not doing this meant that you got two completely different sets of results.

 

Also, some confusion about the way it works and how this is explained:

 

-the instructions that are given for Boolean searching don’t make sense and are limited in number.

-it’s really difficult to explain to a student how it does work, as the help files/videos aren’t helpful

-This was not at all initiative for the student. I then had to explain the disparity in the results

-The help pages contain quite confusing information – particularly when discussing parentheses as the two examples given in the help pages seem to contradict one another. … is not clear how and why they impact the search in such a way.

 

There were some comments on the earlier version of Science Direct Advanced and Expert searches:

 

-the previous advanced search was also flawed, better but still flawed, because it only had two search boxes, and if you wanted to use a more complex search string you had to use the expect search option, which was a non-starter for search novices.

-It would be better if the search operated like Ebsco or Proquest’s advanced search features, whereby you can add multiple search boxes, to accommodate longer, more complex search strings.

-The previous version of the Advanced Search just needed to have the option to add extra lines to the search, but instead it has been completely re-worked so that you have to put all of your search terms in a single line, and then it still doesn’t work in the same way!

 

I accept that the biggest proportion of use of Science Direct comes from users linking directly to journals or articles from databases, indexes or discovery tools, but there is still a significant demand for systematic searching tools. Even if this is only a small proportion of the site usage, the actual numbers of people requiring this is still high. It would be interesting to see more detail from Elsevier on how the decision was reached to make these changes, and to what consultation was undertaken.

 

I should say as well that some other publisher platforms came in for a bashing too in some of the replies (SpringerLink, PQ Ebook Central and Wiley Online Library were named) so this issue is not limited to Elsevier, but part of an overall perceived move to simple Google-style-single-search-box approaches by many publishers.

 

Kind regards,

James

 

 

 

 

Over recent months, Elsevier have made changes to the search functionality on Science Direct.

 

My colleagues  and I are interested to know what the overall opinion is of this.  Does the search functionality on Science Direct meet the requirements of you and your students?

 

I’m happy to take replies off-list, and I’ll summarise in the new year.

 

To start things off, here’s the opinion of one of my colleagues, an Academic Support Librarian from our School of Health & Society :

 

The content has always been great for health subjects in Science Direct, but the recent changes have made it more difficult to use it for systematic health searches.

 

The screen shot below shows the advanced search options available, with the intervention, comparison and outcome missing off the screen.

 

It’s not made clear how ‘Find articles with these terms’ differs from ‘Title, abstract or keywords’, though both get vastly different results.

 

 

A search using the PICO framework has to be compressed in to a single search line. This makes it really hard to add in a search that may have been written and produced results in other databases that support complex multiple line searches.  Not  all of the above search can be seen on the screen, and so a check for mistakes is hard to do. The workaround previously available (expert search) seems to have disappeared. This has implications for the reliability and robustness of any searches conducted in Science Direct if users start to simplify searches to get round this.

 

If a complex search run in several other databases cannot be reliably replicated in Science Direct, then a selection bias may be introduced into the search that skews the sensitivity and the specificity of the results found. This may have implications for systematic reviewing, and clinical searching generally, where users have to apply the principles of evidence-based practice to systematically search for their evidence base.

 

Kind regards

James

 

lis-e-resources is a UKSG list - http://www.uksg.org UKSG groups also available on Facebook and LinkedIn Follow us on Twitter: https://twitter.com/UKSG