iiinova — where your digital self keeps evolving

SRLibProblems @srlibproblems.bsky.social · 29/06/2026

When was the initial search conducted? I've done 3 (or 4) updates occasionally, but even that is not normal per se. Is it a case of teams not completing screening in time, and then requesting an update as they are picking it up again (and then falling into the same trap)?

210

SRLibProblems @srlibproblems.bsky.social · 15/05/2026

🤔 Its a trend. EBSCO uses the fish 🐟. I wonder what we will see next, from a different publisher.

220

SRLibProblems @srlibproblems.bsky.social · 15/05/2026

I also had this happen a few times last week when I was browsing through different reviews.

110

SRLibProblems @srlibproblems.bsky.social · 07/03/2026

Odd! When I search (SO ("nurse.com nursing spectrum") in CINAHL, I get 4929 results (image 1). When I add TI "patient violence", it finds 4 articles with the same title (see image of 1 record). If I click one of the regional titles, I can see all issues from 2011-2015 (image 3).

000

SRLibProblems @srlibproblems.bsky.social · 06/03/2026

Other regional titles with the same name (Nursing spectrum) are found in the NLM catalog and CINAHL. The title may have changed names from Nursing Spectrum to "Nurse.com" CINAHL kept both names in the journal title field, but the NLM catalog has "Nurse.com", Nursing Spectrum, and the combined option

catalog.nlm.nih.gov

120

Reposted by SRLibProblems

JBI @jbiebhc.bsky.social · 04/03/2026

🔥 Hot tip! Before starting a systematic review, make sure an information scientist/research librarian is part of the review team. They have specialised skills to develop & implement a comprehensive search strategy 🔎 Learn more: bit.ly/4spgp1z #JBIMethodology #medlibs

0138

SRLibProblems @srlibproblems.bsky.social · 26/02/2026

This is a move in the right direction, for use in #SystematicReview: 1) Stem/no stem ability in web app 2) Truncation & wildcard operators - and in combination with the stem/no stem toggle 3) Proximity operator (Limited like PubMed's - no truncation allowed with proximity) Still a big improvement!

130

SRLibProblems @srlibproblems.bsky.social · 26/02/2026

Longer search strings might still require use of AND instead of proximity because the character limit won't allow the writing out of each combination (compared to operators that allow nested OR within the proximity string). If #OpenAlex is listening: larger character limits would be really great.

120

SRLibProblems @srlibproblems.bsky.social · 26/02/2026

New studies will follow. The longevity of the studies' results depend on the decisions that were made. E.g. the Stansfield study used AND in lieu of proximity which broadened the search (better sensitivity, worse precision) - the new functions should give similar or better results (for what matters)

110

SRLibProblems @srlibproblems.bsky.social · 26/02/2026

I use their documentation page so many times, that I usually land on the page and immediately use the menu on the left to scroll down to the search section, without looking at the top of the page. Scrolling too fast! 😅. Having seen the same info I was used to (no proximity/wildcard), I asked the Q

010

SRLibProblems @srlibproblems.bsky.social · 26/02/2026

I was looking at this page which is still linked to at the bottom of the OpenAlex page docs.openalex.org/how-to-use-t.... I skipped past the note pointing to the new developer page you linked to. The wildcard and proximity search is a very exciting development. Thanks for sharing. Will check it out!

docs.openalex.org

Search entities | OpenAlex technical documentation

210

SRLibProblems @srlibproblems.bsky.social · 25/02/2026

Do you mean proximity in the semantic search? They don't mention any proximity option in their Boolean search, so my ears perked up when I saw your mention of proximity.

220

SRLibProblems @srlibproblems.bsky.social · 22/02/2026

Since I used #OpenAlex to design and export a #systematicreview search last night, I had to double check just now that my search string still gives the same number of results (it should!).... and it does 🎉. This is exactly what I would expect from a stable system (which OpenAlex is).

020

SRLibProblems @srlibproblems.bsky.social · 22/02/2026

Or, 5) My search query once executed (except via the API query or URL, which isn't the same) Today it looks like what it was on Friday 🤔. What a trip! I guess I saw a peek of the changes that are coming??? Their blog hints at some of the features I saw (semantic search, full-text search).

120

SRLibProblems @srlibproblems.bsky.social · 22/02/2026

3) A combined title, abstract, & fulltext search option (in addition to the existing title only, or title & abstract search). This will be useful for scoping/special cases But, I could no longer see: 4) A way to do one concept per line, with the various lines automatically adding together with AND

110

SRLibProblems @srlibproblems.bsky.social · 22/02/2026

2) Improved filter experience. I was delighted to see how quick & easy it was to see how many/what results are categorized according to individual "type" so I know what "types" to include/exclude (very handy)

110

SRLibProblems @srlibproblems.bsky.social · 22/02/2026

As I began designing my search, I had to familiarize myself with the new look, where certain functions were and the new things I could do, and to work around what I could no longer do. New things I saw: 1) the ability to do a semantic search (using a toggle between Boolean and Semantic)

210

SRLibProblems @srlibproblems.bsky.social · 22/02/2026

I've used #OpenAlex (web) a lot this week, for exploratory searching & scoping, to show a colleague how it works, & even creating & exporting a search for a #SystematicReview. Interestingly, I was on it yesterday and was surprised to see the interface had changed from what I had used the day before

120

SRLibProblems @srlibproblems.bsky.social · 17/02/2026

I have used many of the tips on this blog. Today, I used the instructions (and tool mentioned) in this post (uxcaucustips.blogspot.com/2025/04/tip-...) to generate two sets of records from a draft Ovid Medline search, so the research team can do some sampling. #MLAUXCaucus #MedLibs #CanMedLibs

uxcaucustips.blogspot.com

Tip #58: Exporting a Random Set of Citations from Ovid Databases to Avoid Bias

091

Reposted by SRLibProblems

Matt Grainger @drmattg.bsky.social · 15/01/2026

New guidelines for the use of #AI in evidence synthesis in the environmental fields

024

SRLibProblems @srlibproblems.bsky.social · 29/09/2025

I'm going to have to look into the options (and tools) for citation chaining again, in the near future, but from this perspective. #EvidenceSynthesis

120

SRLibProblems @srlibproblems.bsky.social · 29/09/2025

I tried Zotero earlier, which sources metadata from Crossref & was surprised to see the same (missing abstracts) when using the "add item by identifier" function It was great (timely) to see @aarontay.bsky.social discuss abstracts (+ the implications) this week aarontay.substack.com/p/the-petrol...

120

SRLibProblems @srlibproblems.bsky.social · 29/09/2025

In my case, it was a small number of records with missing abstracts & I had the DOIs, so I decided to manually locate the missing data (from Primo at my institution, or via the publisher page) to add/edit the RIS file. Doing so will allow the records to be screened without disruption, in Covidence.

110

SRLibProblems @srlibproblems.bsky.social · 29/09/2025

To be clear, the problem is not the tool (Citation Chaser or Lens.org, where Citation Chaser gets its metadata from). The problem is the publishers that are withholding and preventing the abstracts from being openly available. And the impact in this case, is on screening at the title-abstract level

131

SRLibProblems @srlibproblems.bsky.social · 29/09/2025

Chasing down abstracts while pulling records for citation chaining (in this case, backwards citation chaining) is no fun and reduces efficiency. Tools like Citation Chaser are awesome, but (as I noticed earlier this week) the process becomes less efficient when some records are missing abstracts.

lens.org

140

SRLibProblems @srlibproblems.bsky.social · 18/09/2025

I think you might also enjoy "PRISMA compliant"[tiab:~2]

030

SRLibProblems @srlibproblems.bsky.social · 03/09/2025

students who use it for a one off, who want to download 20-50 search results so they can explore them further in a citation manager, will be inconvenienced. I wish database vendors would allow small batch record downloading without accounts, and have the RIS format as a standard available format.

082

SRLibProblems @srlibproblems.bsky.social · 03/09/2025

Not liking this spreading practice where database providers now require use of an account to download any records or in specific formats (RIS); not even a small number. I know advanced users (including evidence synthesis librarians/researchers) will create their accounts and continue along but

1112

SRLibProblems @srlibproblems.bsky.social · 06/08/2025

You would not get those records if you searched human rights in TI OR AB because it will parse each line separately and use AND between the words but stay within the field. This is how I understood what the page said. Please do share if I am missing something. From what I see, I can safely use XB

100

SRLibProblems @srlibproblems.bsky.social · 06/08/2025

In this case, " ", or the proximity operator would override the default mode (find all terms, etc) so no impact. But if you searched human rights (with no " ") in XB versus TI or AB, you would get different results. In XB, you will find instances where human is in title and rights is in the abstract

100

SRLibProblems @srlibproblems.bsky.social · 06/08/2025

It sounds like the issues would only happen in situations where the default mode is activated (find all terms, or proximity). But, this won't apply to most systematic searches. For example, most librarians will use "human rights" with " " around the phrase, or a proximity (human N3 rights).

100

SRLibProblems @srlibproblems.bsky.social · 04/08/2025

Never been so happy to have followed good #DataManagement practices. There were times that I almost didn't take the time to use the naming convention, but I resisted. I must remember to thank my awesome colleagues for inspiring this in my practice (they know who they are ❤️). #CanMedLibs

030

SRLibProblems @srlibproblems.bsky.social · 04/08/2025

Saw a post about moving searches out of custom folders in EBSCO, which reminded me to get on it. I moved 105 saved searches from 20 custom folders into one saved searches folder 😒. Thankfully, no renaming needed as I used this naming convention from the start: libguides.ucalgary.ca/DMPforSR/dat...

libguides.ucalgary.ca

Library: Guide to the DMP Assistant Template for Systematic Review Projects: Data Collection

Companion guide to the Portage DMP Assistant Template for Systematic Review Projects

143

SRLibProblems @srlibproblems.bsky.social · 27/05/2025

Why is it an OR? Go with AND - do both!

130

SRLibProblems @srlibproblems.bsky.social · 14/05/2025

This is also true when switching from Ovid to other platforms. Adj3 (Ovid) = N2 (EBSCO) = NEAR/2 (WoS). I don't have Embase on the other platform (only Ovid) so I'm of no help here.

110

SRLibProblems @srlibproblems.bsky.social · 14/04/2025

I often say that PRESS is for requests between search experts, and give an example of peer reviewing another librarian's search. But that I would be happy to review and advise on their search as part of our consultation options. TL;DR - we don't offer a PRESS option for student searches.

140

SRLibProblems @srlibproblems.bsky.social · 14/04/2025

Review & provide feedback - yes (in consult or via email), but I don't think of that as "peer review" & certainly not with PRESS. It is rare to receive such a request AND where the search is good enough to qualify as a peer review. Not impossible though, as there have been 3 or 4 in 5 years.

140

SRLibProblems @srlibproblems.bsky.social · 11/04/2025

The intent is (Controlled fields) NOT (uncontrolled fields). In MEDLINE, we wouldn't use the same terms in both since we'd browse and find the MeSH terms. But, in Scopus, in the absence of deliberate/separate Index term gathering, you could use the same set of terms as shown in my example above.

000

SRLibProblems @srlibproblems.bsky.social · 11/04/2025

(Index terms) NOT (title, abstract, authkey) would match the intent. So, in Scopus syntax it would be (for e.g.): INDEXTERMS("corporate social responsibility" OR {CSR}) AND NOT ( ( TITLE-ABS("corporate social responsibility" OR {CSR}) ) OR ( AUTHKEY("corporate social responsibility" OR {CSR}) ) )

100

SRLibProblems @srlibproblems.bsky.social · 10/04/2025

But if new synonyms keep finding new relevant articles (and recall matters at all - such as in lit review assignments), then you keep going with the iterative refinement. I do tend to emphasize the importance of starting with a good set of terms/synonyms (grouped in concepts).

010

SRLibProblems @srlibproblems.bsky.social · 10/04/2025

Comprehensiveness may be too strong a term. I teach synonym generation in the initial search string development & then it gets added to as part of iterative search refinement (i.e. evaluate the first set of abstracts and add new synonyms to the search). The stopping point depends on the search need.

110

SRLibProblems @srlibproblems.bsky.social · 09/04/2025

Yes, strategies to broaden/narrow is commonly taught, with the objective of either finding more articles or reducing noise or volume. Comprehensiveness of synonyms used and its impact (on number of results) is also covered. Fundamentally, this gets at the principles of it, without defining the terms

100

SRLibProblems @srlibproblems.bsky.social · 09/04/2025

This is part of info lit though (matching the search approach to the objective). Not doing so is definitely a mistake. High recall searches are normal in SR work, but an unexpected perk for me was an improvement in my precision searching, as I know many more advanced strategies.

100

SRLibProblems @srlibproblems.bsky.social · 04/04/2025

The 'Search History' screen on the new EBSCO UI now shows result numbers for each line. We can finally take screenshots to demonstrate the remaining problems 😅 - Happy Friday #IYKYK #Medlibs #CanMedlibs

1170

SRLibProblems @srlibproblems.bsky.social · 27/03/2025

Yes, I agree. Also, there is no label next to the button (like in Scopus), so some folks may not know what it means.

100

SRLibProblems @srlibproblems.bsky.social · 26/03/2025

The fear of the #EBSCO switch is real! Even with the planned July 2025 features, we are going to have to spend more time (clicks) & change the way we do things (for the worse) in order to make up for what is being lost I just want to keep the functionality I have now - is that too much to ask for 😅

030

SRLibProblems @srlibproblems.bsky.social · 25/03/2025

I have been playing around in the new #EBSCO interface and have thoughts 😅 Has anyone noticed that the symbol used to combine queries in the new EBSCO interface is very similar to that used by #Scopus? And now I'm starting to see more similarities across the two systems 🤔 #Medlibs #ExpertSearching

A screenshot of the "combine queries" symbol in the new EBSCOhost user interface. The symbol looks like 2 lines merging into 1 line with a right facing arrow at the end

Screenshot of the symbol that follows the Combine queries text in Scopus. The symbol looks like 2 lines merging into a single line with a right facing arrow at the end.

260

SRLibProblems @srlibproblems.bsky.social · 18/03/2025

They state "For publication or regulatory-grade systematic reviews, you will often have to default to traditional, keyword-based search methods" I guess, all researchers who thought they could do a "systematic review" in minutes and submit for journal publication are going to be disappointed!

193

Reposted by SRLibProblems

JBI @jbiebhc.bsky.social · 04/03/2025

🔥 Hot tip! Before starting a systematic review, make sure an information scientist/research assistant is part of the review team. They have specialised skills to develop and implement a comprehensive search strategy 🔎 synthesismanual.jbi.global #MedLibs #JBImethodology #EvidenceSynthesis

12215