Tag Archives: Google Scholar

Strategies for systematic literature searching

Searching

Image by Philm on Flickr (CC BY-NC-SA)

Literature reviews are a crucial part of the research process. Being able to search for the literature effectively is key to ensuring that you identify information for your research topic. Ultimately, a literature search should help you to identify gaps in the literature, as well as identify appropriate techniques or methodologies (Booth, Sutton & Papaioannou, 2016).

Of course, when searching the literature you shouldn’t restrict yourself to one database and one database alone. Rather you should seek to identify what databases will be relevant to your research topic and then seek to search them systematically, using the same keyword combinations and approaches across all resources as much as possible. So, for example, you shouldn’t settle on PsycInfo as the one database to search, you should look at others such as Academic Search Complete, ScienceDirect, CINAHL, Scopus etc etc (again, dependent on your topic).

Not only should you seek to search across a number of databases, you should also seek to employ a variety of strategies in conducting your searches. By using a variety of strategies across a range of databases, you ensure that you conduct a comprehensive search and, consequently, limit the chances of overlooking crucial papers that could have a serious impact upon your research.

The following are a number of key strategies to employ when searching the literature. They should never be used in isolation, they should be used in conjunction with each other to ensure effective lierature searching.

Thesaurus Searching

Thesaurus searching is a good and effective way to find all of the articles on a particular topic. Some databases (eg PsycInfo) use a subject term index to categorise articles. These terms are used consistently across the database, minimising the kind of variation you see with author keywords that are applied to articles (different authors may well use different keywords to describe the content of their articles). This is what is known as a controlled vocabulary (because there is over-arching control of the categorisation of all the articles).

Instead of searching the database, a thesaurus search requires the user to interrogate the thesaurus or subject index and use that to construct their searches (more on how this works on this video). As well as searching for an individual term, these subject terms can also be used in conjunction with each other (eg “Subject A” AND “Subject B”). When searching for a particular term, you can be pretty confident that all of the articles on that topic will be returned in your results.

However, it is not without its problems. New terminology or niche areas, for example, may not be indexed within the subject index (in which case you will need to do a keyword search). You may also find that it is quite a narrow search that results in a small number of results (depending on the topic), so you may need to broaden it out a little (which is again where a keyword search can help).

Advantages: Locates all the articles on a particular topic due to its use of a controlled vocabulary consistently applied throughout the database.
Disadvantages: May not be useful to find articles on emerging topics or niche areas of interest.

Databases: PsycInfo, Academic Seach Complete, CINAHL, MEDLINE (PubMed).

Keyword searching (free text searching)

Keyword searching is a much broader strategy and simply requires the user to enter their keywords in the search box to find results (much like how you would search on an internet search engine like DuckDuckGo). Compared to a thesaurus search, you are likely to find many more results, particularly if you only use one or two keywords. However, it is a more effective strategy for emerging or niche topics, because you will very quickly and easily find all the articles on that particular topic.

When keyword searching you can also use truncation (*) and wildcards (?) to help make your search more effective. So, for example, searching for child* would find child, children and childhood. Likewise, organi?e looks for organise and organize.

When using keyword searching, it’s best to restrict the search in the database to title, keywords and abstract, otherwise you may find you retrieve a large number of irrelevant results.

Advantages: Better for searching for niche or emerging topics if you use specific terminology.
Disadvantages: Can be too broad, returning large numbers of results which require careful filtering.

Databases: PsycInfo, Academic Search Complete, CINAHL, ScienceDirect, Scopus (and others).

Citation searching (bidirectional searching)

Citation searching can be a very effective way of finding studies. When finding useful papers, one of the first things most people do after reading it is look into the references to see details of the papers referred to in the article to then investigate. Bidirectional searching comes into effect when you look at not only the articles referenced in a particular paper (Paper A), but who has referenced Paper A.

Rather than searching for articles on a particular topic, you identify a key paper, then investigate who has cited that particular paper. You can thing dig into the papers that have cited that paper and repeat the process (this is known as snowballing). By digging into who has cited a particular paper or papers, you can quickly and easily build up a relatively decent amount of papers. Indeed, it is sometimes considered to be more effective than other forms of searching (Hinde and Spackman, 2015).

However, it’s worth noting that this kind of searching may not necessarily find all sources that cite a particular paper. Scopus, for example, indexes over 50 millions sources, but it’s primarily concerned with journal articles and books. Blog posts and other resources will not be picked up. Google Scholar, on the other hand, will include these kinds of web resources. However, the flip side to that is that there will consequently be more resources to plough through.

Advantages: Good way of building up literature by finding articles that have cited an article.
Disadvantages: Not all sources are necessarily indexed.

Databases: Scopus, Google Scholar.

Author searching

If you are aware of an author that has written a couple of articles on your research area, it may be worth digging into the rest of their research outputs. Scopus has an author search option (including an option to search by the author ORCiD) which can help you to find all their outputs (and, of course, all the times these outputs have been cited). You can also click on any author name in Scopus to find all the papers by that particular author.

Databases: Scopus.

Hand searching

Another way to search is to identify key journals covering your topic and search within them. Regardless of whether we have access to the journal in question or not, all journals enable you to at least search within them and have a look at the abstracts of the articles (you can then submit an inter-library loan if we don’t have access to one that you want).

Hand-searching can be effective, however it is a very time consuming process (Armstrong et al, 2005). One study (Greenhalgh and Peacock, 2005) found that a month of hand searching led to only 24 papers being identified – an average of one paper for every nine hours of hand searching (Booth, Sutton & Papaioannou, 2016). Of course, this could not be considered as a comprehensive method on its own, but it’s a useful additional strategy to employ.

Key Points

* Don’t rely on only one strategy. Use multiple techniques to ensure that you conduct a comprehensive literature search.

* Identify keywords and terms before you start searching. If you identify these first it will make the search process much easier and you will be less likely to miss key papers.

* Consider the databases you need to be using (as well as other resources) in the broadest sense. Will you need to look at clinical/health databases? Education databases? Searching in PsycInfo alone will not be enough. If your topic touches on other subject areas, then you will need to look at the relevant databases in other subject areas.

* Make sure, once you have identified your keywords, that you use the same keywords and keyword combinations across all the databases you search. This will help to ensure your search is comprehensive, as well as being systematic.

* Look for patterns when you search. If particular authors or journals keep cropping up, it may be worth considering an author search to find papers by a particular researcher, or hand searching to search more deeply into a particular journal.

* Google Scholar is useful for searching for grey literature, and can find more than citation databases or grey literature identification methods (Haddaway et al, 2015). However, as Haddaway et al note, it does miss relevant evidence and should not be used as a standalone resource. It should, as with other approaches, be used in combination with other search strategies.

References

Armstrong, R., Jackson, N., Doyle, J., Waters, E., & Howes, F. (2005). It’s in your hands: the value of handsearching in conducting systematic reviews of public health interventions. Journal of Public Health, 27(4), 388-391. http://dx.doi.org/10.1093/pubmed/fdi056.

Booth, A., Sutton, A. & Papaioannou, D. (2016). Systematic approaches to a successful literature review (2nd ed.). Croydon, UK: Sage.

Greenhalgh, T. & Peacock, R. (2005). Effectiveness and efficiency of search methods in systematic reviews of complex evidence: audit of primary sources. BMJ, 331, 1064-1065. http://dx.doi.org/10.1136/bmj.38636.593461.68.

Haddaway, N. R., Collins, A. M., Coughlin, D., & Kirk, S. (2015). The Role of Google Scholar in Evidence Reviews and Its Applicability to Grey Literature Searching. Plos ONE, 10(9), 1-17. http://dx.doi.org/10.1371/journal.pone.0138237.

Hinde, S. & Spackman, E. (2015). Bidirectional citation searching to completion: an exploration of literature searching methods. PharmacoEconomics, 33(1), 5-11. http://dx.doi.org/10.1007/s40273-014-0205-3.

Searching PsycInfo

Advertisements

The problem with searching for journal articles using internet search engines

(Image c/o JD Hancock on Flickr)

Internet search engines are great. They help you find information quickly and easily. They are, however, not always necessarily the best tool to use in searching for academic papers. They are not particularly sophisticated and don’t necessarily point you in the right direction to access the papers you need.

Take Google Scholar (other internet search engines are available!). A lot of people tell me that they’ve searched on Google Scholar but either cannot find many papers that are relevant to their needs or they find the papers but cannot access them. This is not necessarily because we do not have access to the papers required, but is down to the way Google and the databases we subscribe to work.

For my school (Psychology) the key databases are PsycInfo and PsycArticles, both provided by EBSCO (you can access both of them via this link). These two databases include a lot of papers published by a number of different publishers (and by the way, everything in PsycArticles is also in PsycInfo!). But publishers also have their own sites that host content for each of their publications. So you can find journal titles hosted on the Taylor and Francis or Wiley websites, for example. However, often our access to a particular journal is not via the publisher website, but via EBSCO. And this is where the problems can arise.

Internet search engines (include Google Scholar!) are unable to search within EBSCO. Consequently, if you conduct a search and find a journal article published by, for example, Taylor and Francis, the search engine will take you to the Taylor and Francis page for that particular article. At this point, you will think you do not have access to the article because access is provided by EBSCO rather than Taylor and Francis.

There are a couple of ways to get around this. First of all, restrict your use of Google Scholar and internet search engines. It can be useful, but you are much better off using any of our dedicated databases. The databases will direct you to the content you can access, and also enables you to conduct more sophisticated searching, particularly in terms of filtering your search results. EBSCO in particular enables you to filter by methodology, age, gender, subject heading and more. The fact that it makes our subscribed content easier to find and easier to filter is a huge advantage over internet search engines, particularly when conducting research.

Second, if you do use an internet search engine and come up against an article that you cannot access, head to Library Search, paste the title of the journal in the search window (under the “Books and more” heading) and it will tell you whether we have access to the journal in question, which database includes the journal and what coverage the database provides access to.

I would never say avoid using it altogether as it can be a useful starting point. However, it is only ever a starting point. Your research should never begin and end on Google Scholar (or any internet search engine).  In short, when conducting your literature searching, focus the majority of your searching on our databases and Library Search. Use internet search engines if you must, but keep in mind that they are limited and should never be your sole source when literature searching. And if you need help searching any of our specialist databases, please do get in touch to arrange a one-to-one, either with myself (for Psychology students) or your subject librarian.

Happy searching!

Ian Clark

email: i.clark@uel.ac.uk | twitter: @PsycLib_UEL