Finding reviews on any topic using Lens.org and 2d search — a new efficient method
TL;DR — This article shows you a reliable search strategy that you can use in Lens.org to find meta-analysis, systematic reviews, review articles of topics you are interested in regardless of your discipline to help kick start your research. I covered the rationale for doing this with Lens.org earlier, in this article, I will show an easy way to adept a well tuned search strategy to your topic of interest in 5 minute flat using 2D Search (a free tool that now supports Lens.org).
Starting off your research in a new area like leadership, creativity and overwhelmed by the amount of papers out there and have no idea where to start? Perhaps you have heard that one good way is to look for things like “review articles”, “systematic reviews”, “meta-analysis” or “bibliography” that summarise the research in an area. But how do you find them in an efficient and comprehensive way?
If you are a user in the medical/life science field, you may know of Pubmed filters that find such items but what if you are working in other areas?
In such situations, I highly recommend Lens.org which includes Pubmed, Pubmed Central, Microsoft Academic (which is comparable to Google Scholar in scope) and more as the tool to use.
In an earlier blog post, I worked out a efficient search strategy that you can use in Lens.org to do so which you can adapt for your own use. But to be honest, it can still be daunting if you are not very skilled with Boolean searches as the search strategy need to be cut and pasted and looks really complicated to alter to change to your topic.
The solution is 2D search. With this, you can get the search off and running in 5 min flat.
You will quickly generate high quality results of reviews of any topic you might think of.
Some examples I generated with no effort by just changing the topic keyword (in fact you can actually improve recall by adding more synonyms for the topic but I did not borther).
- Mergers and acquisition — generated 125 Lens.org results
- Audit committees — generated 31 Lens.org results
- Executive compensation — generated 41 Lens.org results
Interested? Here are the steps
Step 1 : Click on the following link to my saved search 2D search template (no login or registration needed).
Click on the 2D search link (newer higher recall template here), you should see something like
You might be panicking now by the seemingly complicated search. But don’t worry all you need to do is change one small part.
Step 2 : Scroll to the bottom and locate the block in the search strategy that is on creativity and change the terms to your topic
The search strategy as complex as it looks is designed to find content that are reviews, systematic reviews, meta-analysis on Creativity. For you to adapt it for your use , all you need to do is to change the part of it that matches the word Creativity to your topic.
For example if your topic is “leadership” or the narrow “Servant leadership”. change it to look like the below.
Advanced tip #1 — taking advantage of controlled vocabulory — Field of study and MeSH
While you can enter any query term you want for the fields of title, abstract, Keyword, the ones for “field_of_study” and “mesh_term.mesh_heading” are controlled terms.
They are also the few fields that are case sensitive (e.g. Leadership not leadership).
By controlled terms, this means that they accept only specific terms on a list.
(Note if you enter a term that is not included in the controlled term list, the overall search strategy will still work, even though that part of the search will get no results, because we are using an OR operator with title/abstract matches)
What how do you find the right terms? For field_of_study which comes from Microsoft Academic , you can use the topic browser here to search and see what is suggested.
You can do the same with the MeSH browser.
Alternatively if you can’t find a suitable term, you can work backwards by searching for a paper that covers the topic you are interesting in Microsoft academic or Lens.org and look at what field of study is tagged.
For example, I was trying to find a suitable field of study to describe anti-trust law and this paper in Microsoft academic gives an idea to use “unfair competition”
Do note that both Microsoft academic Field of study and MeSH provide a hierachy of terms, so for example “Leadership” is the parent concept of “Servant leadership”. In Lens.org, if you search for the parent concept “Leadership”, you will not automatically get the child concept — “Servant leadership”, so you probably want to include the most specific level of concept you can find.
Advanced tip #2 — adding more matches for synonyms
Say you are looking in the area of “Executive compensation”, you check the Microsoft academic topic browser and find the term “Executive compensation “ does exists as a controlled term.
But that might still miss content if items use other synonyms in title/abstract/keywords or are not tagged in field of study. If so, you can easily add more matches to the search. In the example below, I added matches for items with the words “executive renumeration”, “CEO compensation”, “CEO renumeration” in the title.
You can copy and paste existing parts of the search, then click to overwrite the text.
Because they are all in a block with the OR function, they will be added to existing matches.
That’s basically it, now to generate the result
Step 3: Generate the search in Lens.org
To generate the result you can click on “Show results” on the right corner of the screen to generate an inline view of Lens.org results or you can click on the icon on the right corner to open the results for Lens.org in a new page.
As I write this, this gives you 79 items on systematic reviews, meta-analysis and other reviews of various kinds in servant leadership!
Step 4 -Do visualization of results in Lens.org (optional)
With the above search you get only 79 results which you can still eyeball. On a broader subject like “leadership” you will get far more reviews (>2,000 results)
In such a case, since you are already usings Lens.org, you can make use of Lens.org’s amazing visualization capabilities to have a broad overview.
Click on the analytics tab and you can see increasing visualizations on journals published in, the institutions that produce such reviews and more.
https://link.lens.org/L522p7Gqbrj
Conclusion
I have been trying this out with various topics like
- Mergers and acquisition — generated 125 Lens.org results
- Audit committees — generated 31 Lens.org results
- Executive compensation — generated 41 Lens.org results
- and more corporate governance, school class size, affiliation action
and they work very well often producing very high quality results for reviews on the topic with pretty much no effort on my part.
The default search template, is designed to go for precision over recall, and advanced users might consider enabling different parts of the search if they prefer to boost recall to get more results at the cost of false hits.
For example you can enable (right click — enable), parts of the search that are greyed out that find bibliogaphies or the parts that match by publication titles.
Advanced users might also want to try out truncation and wildcards (note Lens.org stems by default and drops stop words, truncation and wildcards also works for single words not phrases but will drop stemming if done). Proximity is supported in Lens.org with the tilde symbol but it is not functioning in 2D search currently.
It’s unclear to me how good the results produced by this search strategy without a gold standard to validate against, but it looks promising to me. By just using Field of Study as a filter for systematic reviews, meta-analysis and comparing against Pubmed’s filters gives very close results, implying that Field of Study works pretty well in the domains covered by Pubmed and may also work in other disciplines.
I will be interested to see what advanced users will make of this, now one can easily do long complicated boolean in Lens.org. Librarians for example can use this in classes if you teach Boolean.
23th June 2020 — Minor update to templates to handle the NOT publication types
