How to do a Systematic Literature Review

Lecture slides will become available here

Abstract

A well-developed literature review is critical to any PhD project. During the early stages of your research it guides the selection of your research questions, whilst in the thesis and viva it demonstrates your understanding of the field and the place of your work within it, to your examiners.

Many questions arise during the construction of a literature survey:

• What new avenues or directions should I consider?
• What don't I know about this topic?
• Have I read enough of the literature?
• Why should I believe the conclusions in this paper?
• How should I present the findings of this paper?

There are several different philosophical approaches to conducting a review of the literature, reflecting different potential audiences and purposes. We will provide you a step-wise guideline to conduct a literature review, including how to decide whether particular techniques and/or methods are appropriate.

This will include:

• How to search for material relevant to your topic of study using both search tools and snowballing techniques.
• How to decide on relevance criteria to filter results.
• How to make your literature selection construction transparent and traceable.
• How to construct a *meta-analysis* of the literature and decide when it is appropriate to do so.

Finally, we will discuss how to assess, present, and critically discuss the contribution made by each article, and make a rigorous decision about excluding work where appropriate.

Learning Outcomes

• LO1: A good understanding of different stages of a literature review process in computing science.
• LO2: Participants should be able to confidently support their research with complete overview of the relevant literature.