An introduction to Systematic reviews
On the 5th of November has been held scientific seminar from series "Culture Matters". Sebastian Bamberg, PhD, professor at University of Applied Sciences, Bielefeld (Germany), made a presentation on following topic: “An introduction to Systematic reviews”.
Most empirical sciences are of a cumulative nature, research results of others are used for decision which questions are sufficiently examined and which need further research. Good research synthesis should be as comprehensive and unbiased as possible. However, everyone who has tried to conduct a research synthesis has experienced the difficulties associated with such an enterprise. There are two methodological standards of dealing with problems of research synthesis: systematic reviews and meta-analysis. One should remember that these two terms are not interchangeable, meta-analysis is only the statistical part of a systematic review used to combine the results of several studies addressing the same question into a single summary measure.
Systematic reviews are derived mostly from information explosion. For example, these are numbers from medicine: annually there are 3 million articles are published in more than 30,000 journals. Not all this information has high quality. That’s why researchers make review articles to give a good overview of everything was published. A “review” is the generic term for any attempt to synthesis the results and conclusions of two or more publications on a given topic. In former times reviews were usually conducted on a narrative basis, produced by a “content expert”. But it is an expert opinion rather than a real review.
Main problem for reviews is bias problem. Bias means something that will cause a consistent deviation from the truth. There are three possible sources of bias in reviews: bias arising from the studies included in the review (there might be poor studies); bias arising from the studies not included in the review (with insignificant or “good” results but interesting); and bias arising from the way the review is done (it should be objective and correct). So in systematic reviews researchers try to minimize the effects of anything that will cause results to deviate from the truth.
Stages of conducting systematic review are following: first of all it is asking precise question (it leads to clearer answer). Then researcher should select studies, which is rather time-consuming. Next steps are to assess quality of chosen studies and to extract data from them. And last step is data synthesis, which is actually meta-analysis.
Systematic reviews help to find research gaps and make one’s research better. There are units that make such reviews, for psychology it is Campbell Collaboration.