- Research can be understood as systematic investigation to develop theories, establish evidence and solve problems. We can either undertake new research or we can learn from what others have already studied. How, then, do we go about finding out what has already been studied, how it has been studied, and what this research has found out?
- Reviewing evidence, and synthesising findings, is something that we do all the
time when going about our daily lives. For example, consider the range of activities
involved in buying a new car. We approach the problem with an overarching question:
‘which car shall I buy?’ that can be broken down into a series of questions
including: ‘what cars are available?’; ‘what type of car do I need?’; ‘which cars can
I afford?’; and, if manufacturers' marketing departments are doing their job,
‘which car will make me happy?’. We then gather data together to help us make
our decision. We buy car magazines, read online reviews, talk to people we know
and, when we’ve narrowed our options down a little, visit car showrooms and take
some cars out for a test drive. We critically review the evidence we have gathered
(including our personal experience) and identify possible reasons for doubting the
veracity of individual claims.
- Reviewing research systematically involves three key activities: identifying and
describing the relevant research (‘mapping’ the research), critically appraising
research reports in a systematic manner, and bringing together the findings into a
coherent statement, known as synthesis.
- The idea that different research questions may be answered best by different
methods and by different types of data also applies to reviews. For instance, systematic
reviews addressing questions about the effects of health interventions have
widely agreed systematic methods for setting the scope, judging the quality of studies,
and presenting the synthesised findings, often using statistical meta-analysis of
the results of randomised controlled trials (see, for example, Higgins and Green
2011). However, a systematic question-driven approach to reviews can apply equally
to research questions of process or of meaning that are addressed by more qualitative
primary research and by review methods that reflect those qualitative research
approaches.
- Anyone working on a review will always bring their own subjective ways of looking
at the world that they are researching, and their own ways of seeing the practice
of research itself.
- Criteria to choose, e.g. The diagram summarises the many reasons for why, out of the 14,439
references found initially by an extensive search strategy, 25 were eventually identified as being relevant for the in-depth review. It enables readers to see at a glance that, for example, 1,394 were duplicates, 10,302 were not about PDP, 1,164 were not empirical, etc.
- Utilising at least two of these sources is recommended since individually these sources may not cite all referencing papers or reports. For example, forward reference checking using Google Scholar will identify research reports and dissertations as well as some journal papers; however, forward reference checking with Web of Science might cite additional journal papers.
- The process of reading each title and abstract to ascertain its relevance is known as screening.
- Explicit criteria, based on the review’s scope and question(s), are applied to each record in order to determine if it should be included in or excluded from the review. Criteria often cover the range of dates that the review covers, characteristics of the study population (and intervention, if applicable) and methods used in the research.
- pg. 87 > Table 5.1 Data increase through a review
pg. 92 > Figure 5.3 Flow of references through a review
pg. 122 > Box 6.3
pg. 128 > Table 6.1 Example of record-keeping log