Dear Dr Ali

 Based on my humble experience of conducting two systematic reviews, I would suggest , that you include all possible keywords that appear in your articles when you must have done while conducting preliminary research on the topic or research question that you formulated. I am specifically not from your field , so I cannot specifically comment on the perfect search strategy for you, but you are advised to separate all possible keywords in full form and abbreviations that have anywhere appeared in the literature to validate your search strategy. Make your search strategy as elaborate as possible separated by Boolean terminology so the maximum articles appear in your  databases. If we try to be very specific, we are liable to omit articles of relevance. Systematic review is a time taking process and you may manually have to read all the abstracts to finally decide which ones to attain full text for. I hope I was of some help.

Best

 I agree with Priyanka Kapoor that you should be using all possible combinations so as to you do not miss out any relevant articles regarding your review.

I agree with Priyanka and Saif. A systematic review needs to be broad in reach. The protocol you use will then help you to winnow the papers systematically. I'd avoid using the NOT operator as you can easily discount important material. 

Do consider using "'World Wide Web" as as search term too.

PRISMA has some really clear guidelines for winnowing http://www.prisma-statement.org/

Best

Thanks, Priyanka and Saif . Indeed a protocol is the backbone of a SLR.

Sarah, I really appreciate your contribution both for the search term and the link.

BR,

Ali

i agree with Sarah. Sometimes i avoid using NOT as it definately makes me loose hits that are relevant; but this decision depends also on the number of final hits i retrieve: if i get a very high number of hits that are not reasonably manageable then i will keep the NOT, and vice versa.

If you opt to use the NOT, then increase sensitivity by including other keyword to the search option 2 specially for the last two concepts.

hope this helps...

Thank you Aida.

You are right, there is trad off between ease and loosing relevant studies. I have already administrated search using Option 2. However, will do with option 1. 

It will further enhance understanding regarding use of NOT as an operator in SLR.

Thanks once again.

NOT though sometimes becomes absolutely essential to eliminate related but unnecessary literature. For eg. if my research question is on cytokines in tooth movement, I would separate all cytokines by OR and eliminate all other mediators (enzymes, hormones etc) by  NOT. It is to specifically eliminate noise in literature.

Yes of course that's the purpose of NOT. Perhaps it is not possible to generalize boolean operators use in SLRs as every SLR has different needs.  I perhaps will go with string without NOT as there's chance that personality has been studied side by side for both normal and problematic usage or usage has been meaured using problematic scale.

Dear Dr. Ali,

I agree with Priyanak and seif.

Dear Mr Farooq,

I am sorry but I am in disagreement. If we are talking about exploiting massive referential databases, as MEDLINE and EMBASE are, both of them have got two classes of laguage in order to search in: natural and structured. If you want to conduct search strategies in this kind of database, you must construct strategies with terms fromn both languages.

Your search strategies seem to have been written only with terms from natural languaje. For this, is useful to apply terms related with concepts dealed in the known papers (as author keywords are) but the strategy will be incomplete.

For any research, and a SLR is a research, you shoud analyze your research question in four aspects: Poblation, Intervention, Compartor and Outcome. This is the called "PICO format" and it is widely used in the health research field, particulary when we carrying on SRL. Each of the four aspects must be explored with terms from natural languaje and from structured.

From nature of your question I assume you will explore databases as PsycINFO and MEDLINE. All terms of structured languaje of MEDLINE database compose a thesaurus (called MeSH: Medical Subject Headings) and is the same for PsycINFO. A some knowledge of thesaurus is needed in order to perform right search strategies My advice is that you shold seek a help from an expert librarian with knowledge in this kind of searches or from a MD or psychologist with skills in the field of literature research (better if it is for SLRs)

Please, feel free to contact me if you need further information.

With best wishes.

Dear Mr. Antonio,

I must say I heard about structured class for the first rather. Rather I couldn't find a single study in my field of study where structured class has been made use of.

As long as PICO is concerned, I have already taking into consideration. Rather I using PICOCE (Population, Intervention, Comparison, Outcomes, Context and Experimental Designs) as a method to formulate the research questions Kitchenham & Charters (2007). However, I have adapted PICOCE to PIOCE as I do not aim at making any comparison, at least not a the moment.

Having said that all, I find your view point very interesting. As a student, beyond this SLR, I would like to know more regarding the structured class. I would appreciate if you could refer publications where both natural and structured class has been address in conducting SLR or a mapping/scoping study.

Thanks,

Ali

Dear Dr Antonio

Thank you for the useful insights, I would just like to add that even if you apply the same search strategy to two different databases , SCOPUS and Embase, to get relevant results, you might need some alteration regarding use of Boolean terminology as well as some words have different terminologies to be used for the same words. But if you spend some time exploring the two databases , you may be able to on your own (I am telling by my experience) 

 Also, I need to ask if someone has access to Embase in their universities, do let me know I would be really grateful.

 My mail ID : [email protected]

Dear Mr. Farooq,

This link points to a RLS and metaanalysis titled "Efectividad y seguridad de la artroplastia total discal en lumbalgia crónica. Revisión sistemática y metaanálisis". Is a SLR in Spanish but whit an Executive summary in English (Lumbar total disc replacement effectiveness and safety in chronic low back pain) I am a coauthor of it. Pages 76 to 79 contain strategies for EMBASE and MEDLINE both of them are performed using natural and structured language (EMTREE and MeSH, respectively). For instance, p 76, lines 1 to 16 are a filter to select RCT in humans. Lines 17 and 18  are a strategy written in natural language in order to identify our Poblation. Line 19 is to identify our Poblation but with structured lenguage MeSH. Intervention is analyzed in a same way.

Researchers should become familiar at least with MeSH terminology. It is easy searching with MeSHs through an inteface as Ovid, but advanced search in PubMed allows it as well (a bit harder). Is very useful performing an initial analysis of MeSH in non-systematically screened (but relevant) and referenced papers in PubMed, when these papers are truly relevant for our study, if these articles are available.

With best wishes.

Book Efectividad y seguridad de la artroplastia total discal en l...

HI Ali

find out how many databases are there in your field of interest and which ones are you going to use and why! then get in touch with your librarian and get him/her to suggest the search strategy and the key words for your search ,

Hope this helps,

Regards

Zahid

We have recently published guidelines for conducting systematic literature reviews. We used the example of SCM, but it can be adapted to other disciplines. You can find our article on my ResearchGate page. Christian F. Durach, Joakim Kembro & Andreas Wieland (2017), A new paradigm for systematic literature reviews in supply chain management, Journal of Supply Chain Management, Vol. 53, Issue 4.

I am agree with Priyanka Kapoor and base on her note i propose this search strings :

(World Wide Web OR Cyberspace OR internet OR web* OR Cyberspace) AND ( Personalit* OR Personality Type OR Personality Traits ) AND (use* OR using OR usage OR search* OR navigati*) .

Take a look at this paper

Conference Paper Smart School Guidance and Vocational Guidance System Through...

It is always a good idea to make the search strategy in conjunction with a librarian. Use a wide search strategy encompassing all major database and appropriate key words, MESH terms. I have attached an article for your reference. Hope this helps!

Article Does Mini Screw Assisted Rapid Palatal Expansion (MARPE) hav...