NLPN

A network for new and aspiring library professionals

Constructing a literature search

Carrying out a literature search can be daunting- knowing where to start, understanding the foreign language of “search syntax”, how to choose your search terms, and not finding exactly what you’re after, or being sure you’ve found everything despite what feels like an eternity scrolling through pages and pages of results.

There is plenty of support available, including excellent training such as the LIHNN MOOC. Also take a look at NLPN top tips on literature searching for useful resources which will help you to develop your techniques. The important thing to remember is, it is a skill that will get better the more you practice and gain confidence. And, although there are some important techniques, I often think of searching as the “art” of applying the science of information retrieval – people may take slightly different approaches to the same search topic, there is not always a single “right” way, and sharing ideas can help – we find in our team in the NHS, chatting to a colleague about a tricky search helps us refocus and come up with a slightly different range of search terms.

Literature searching requires a methodical approach – but is an iterative process, there’s nothing wrong with refining your search terms as you go along- a degree of flexibility and adaptability is needed.

We won’t attempt to reproduce the excellent training resources such as the LIHNN MOOC here, but here are a few tips to help your literature searching…(in no particular order)

  • The boring bit, but have a think about the topic or question before you start – make sure you fully understand the user’s request (if it’s a search on their behalf) and any strict inclusion or exclusion criteria, dates, etc. Making sure you fully understand the topic and type of literature you’re after will help you pick the right sources, and make a judgement call on results, such as whether slightly broader results would be acceptable (e.g. from a different country if there is nothing available from the UK).
  • Frameworks like PICO can help, but not all searches fit perfectly… mapping out your search question to break it down into grouped concepts is the important part to help identify key search terms and synonyms. The main point of the frameworks is to turn a “natural language” question or phrase into the key words or phrases you want to find in your results, and you want the computer to look for.
  • Sources are really important, but you may need to be pragmatic depending on what you have access to – there are some good quality free databases and sources (PubMed, etc…) and it’s not wrong to use these.
  • …and for ‘grey’ literature, Google (or other web search engine) is your friend. Using a structured approach, and appropriate Booleans and a domain search (or the advanced search function), can help locate key documents on government or research institution websites.
  • Using the subject headings or thesaurus terms in the database is of course seen as the ‘proper’ way to search, and I would always recommend starting by looking up appropriate subject headings. But, if you find that there are no subject headings to match your search, you’re probably not doing anything wrong! Newer or very niche topics may not have made it through into the subject headings. A well thought out keyword search (using appropriate variations) can be just as effective [LIHNN MOOC explains SHs in detail].
  • Beware that different databases have ever-so-slightly different quirks, such as different symbols for truncation (some use $, others *) – if you’re getting peculiar results, that could be why. The “help” or “search tips” page should explain which symbol to use,
  • However perfect your search terms, you will have to do some filtering of the results – not everything will be relevant and time spent refining the results, particularly to present to a user, is not wasted. Inevitably there will be results that are a “match” in terms of having the key words, but the context is wrong – scanning abstracts is an important part of the search process.
  • Practice makes perfect! The more you try out different techniques, the more you play around with different search tools, and the more you get to know a topic, you’ll develop a “sixth sense” for what works.

 

 

Does CBT improve self-esteem in adults with eating disorders, compared with psychotherapy?

Using the PICO table below, see if you can add any relevant words or phrases, we’ve provided a few to get you started:

pico

Now think about how you will use Boolean logic to structure your search, grouping relevant terms together and combining different sets in order to achieve the maximum number of relevant results from your strategy:

boolean_and

boolean_or

Once you have developed your PICO strategy and identified which terms you would combine using Boolean logic, you can start to sketch out your search strategy.

An example of the search strategy using OVID Medline can be found below; you should be able to identify the three key concept groups and how they have been defined. As we mention searches are subjective so you may have included/excluded different terms to the ones found below:

Database: Ovid MEDLINE(R) <1946 to September Week 2 2016>

Search Strategy:

——————————————————————————–

1     “Feeding and Eating Disorders”/ (12576)

2     Anorexia Nervosa/ or Anorexia/ (16046)

3     Bulimia Nervosa/ or Bulimia/ (6871)

4     (binge* adj4 eat*).ti,ab. (3831)

5     (eat* adj4 (disorder* or condition* or illness*)).ti,ab. (14668)

6     (anorex* or bulimi* or orthorex* or nervosa*).ti,ab. (29833)

7     (extrem* adj4 diet*).ti,ab. (497)

8     or/1-7 (47490)

9     Cognitive Therapy/ (19588)

10     (cognitiv* adj4 (behavior* or behaviour*) adj4 therap*).ti,ab. (10145)

11     (CBT or CBT-E or CBT-Ef).ti,ab. (5946)

12     or/9-11 (23414)

13     self concept/ (49961)

14     (self* adj4 (care* or caring* or confiden* or esteem* or concept* or regard* or assuranc* or respect*)).ti,ab. (43134)

15     or/13-14 (79992)

16     8 and 12 and 15 (138)

17     animals/ not humans/ (4285612)

18     16 not 17 (138)

Once you have developed a first draft, take a look through the final results and see if you can identify any words or phrases that you might have missed. For instance by adding ‘care* or caring*’ to line 14, 14 relevant results were added to the final number.

 

Next, try developing a strategy for one of the following search questions:

  • Does aspirin or paracetamol work better in relieving headache in adults?
  • Does Calpol or Nurofen for Children work better in relieving earache symptoms in a 5 year old?
Advertisements

One comment on “Constructing a literature search

  1. Pingback: Top Tips: Literature Searching | NLPN

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Information

This entry was posted on December 9, 2016 by in Skills and tagged , , , .

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 2,042 other followers

Twitter Updates

%d bloggers like this: