This is a guide about writing a systematic literature study. If you need more help or if you have questions or comments about the content, contact firstname.lastname@example.org or turn to the Biology library. NB, you do not need to be this thorough in a small literature project within a course, or if you are writing a thesis that is not mainly a literature study.
A systematic literature search (i.e. a search for a literature study) is somewhat different from a 'regular' literature search. These are its general attributes:
A systematic literature study should capture all relevant material. Ideally your search will find all relevant sources and at the same time capture as few non-relevant sources as possible. In practice, it is a tricky balance act. Start with a broad search to capture all relevant sources. Even if the search will be gradually narrowed down to be more specific it is unavoidable that some non-relevant sources will remain. These can be sorted out manually in the final stages.
You need to thoroughly present how you ended up with the sources that you choose to analyze and summarize. The reader must understand exactly how you did your search and be able to repeat it. In the method section you therefore need to include:
The bibliographic databases have a transparent search system and contains almost exclusively publications that are quality controlled.
LUBsearch has in many aspects similar attributes to the bibliographic databases. An advantage with LUBsearch is that beyond academic publications you can find many other document types such as reports, book chapters and news articles. A disadvantage is that the result list often contains duplicates. LUBsearch searches simultaneously in several different databases that Lund University has access to, which you need to present (see Databases A-Z).
In Google and Google Scholar you will get more search results as they will generally search in the full text, not just the bibliographic information (title/abstract/keywords etc). In some cases this is a huge advantage. The downside is that you have little control over your search (these search tools have a sort of "hidden help"), which makes it non-transparent. Boolean operators (AND/OR/NOT) and truncation doesn't work well. Furthermore, the search you do in Google is not repeatable or non-partial as the Google search algorithm will generate different search results depending on for example your search history, the language setting on your computer and the popularity of documents. Google Scholar appears to be somewhat better in this aspect, but it is unclear to what extent. However, as some sources can only be found through Google (aka. "grey literature"), it can be a good complementary search tool. Keep in mind that in both Google and Google Scholar, you may find sources of doubtful quality.
One option is to present your search strategy in a table:
Additional searches that may be difficult to report in a table you can describe in the text. Here are a few examples:
In the method section, you have to describe your search strategy, but also your criteria for selection and assessment of relevance. In other words, how you after the search choose the sources that you will analyze in more depth. The assessment of relevance may be based on for example your hypothesis or the hypothesis in the sources, reading of title/abstract, presence of keywords in the sources, type of study (e.g. qualitative or quantitative) or the method used in the sources. Some people use a point system to classify the quality of the sources based on certain criteria. In this way you can get a relatively independent (or at least transparent) measurement of the quality of the sources. The selection can be presented in a figure:
Don't forget to describe your reflections during your work. Examples of questions you may ask yourself is:
The analysis is described in the method section after presenting the search strategy and selection criteria. Here is an example of a way to arrange the analysis:
The headings and subheadings in this section may look different. Look at previous student papers to get some ideas on how to do it. Here is an example of a possible layout: