In a systematic review you must research and integrate the findings of empirical studies in order to answer a specific question. Reviews such as this are often published in general psychology journals and are also useful to non-psychology professionals, so you should write in a way that any intelligent, professional reader could reasonably be expected to understand – no specialist jargon or unexplained abbreviations. Remember here that your marker is unlikely to be a specialist in the area of your review and will penalise you if you don’t explain yourself clearly.
A systematic review is different from a typical undergraduate essay or literature review at least five ways.
The format for a systematic literature review is more like a laboratory report than an essay, and includes an abstract. In broad terms, you should think of research articles as data and the literature search as collecting data.
The method section is a detailed account of the literature search, including databases and keywords used. So, if you want to get an early start on this before the relevant seminar, keep careful records and save everything you do.
The review gives a definitive answer to the research question because you have searched systematically rather than looking for articles that build towards a particular argument. An answer is expected and required – it is not sufficient to say that “findings were mixed and more research is needed”.
The review is aimed at being useful in a real-world setting; if you were asked to complete such a review in a real job or within a Masters placement in clinical, organisational, or another speciality in psychology, it would be because your boss wants an answer, not a prevarication.
Many systematic literature reviews are also quantitative. Such meta-analyses involve statistically combining the findings from the studies reviewed. This is NOT required for the current assignment.
Your task is to conduct a systematic literature review and to write a report on it. Your report should contain STRICTLY no more than 2500 words of text plus one table. Words of text include the review itself, citations, and headings, but exclude the title, the abstract, the reference list, and the table. In your report you must ask and answer a question. The key to this task is to select a question to which there is a definite answer, and which is narrowly defined so that you can answer it within the word limit. You must provide a full list of references. Use the sixth edition of the APA Manual to format for citations and for references the review.
Your final review, around a dozen papers will suffice and is probably about right.
1. Your research question is “What predicts psychological ill-health and/or absenteeism in the workplace?”
– Much of this research is undertaken on health-care workers or teachers – you should probably focus on one group of workers only, and on a particular type of ill-health (for example stress or burnout or depression).
Identify the published review paper (Michie and Williams, 2003, which covered papers till 1999), and then write your review as an update, covering papers from 2000 onward.
Electronic filtering using the database
– English language,
– Some aspects of methodology were then used on 2 databases searched: PsychINFO (limits were used to articles published from 2000 onwards), using the search string((SU.exact(“STRESS”) OR ORG.exact(“STRESS”)) AND SU.exact(“OCCUPATIONAL HEALTH”)) which yielded 40 results and PMC (2000 to 2016) using the search string “What predicts psychological ill-health and absenteeism in the workplace?”. The abstracts were then vetted for further exclusion. Duplicate publishing – check for multiple reports of the same data (usually by the same authors so easy to detect)
Each sample/finding should be represented only once
DISCUSS THIS EXPLICITLY
TABLE (SHOULD BE APA FORMATTED)
The table should include the following:
– Study identifier (name, date in short form, e.g. Michie et al. 2003)
– Rows are in alphabetical order by first author
– Participants (n, gender, age, response rate, occupation)
– Design (e.g. wait-list control vs placebo control or cross-sectional vs longitudinal)
Discuss your findings from the evidence gathered. Limitations can encompass: Many studies measured multiple outcomes
Although all reported positive changes, there may be type I error.
Student self-reports were more likely to change than other outcome measures.
Was this particularly true for studies without active control?
Variable delivery length and expertise
Group level effects and
Publication bias. Make a conclusion based on your findings.
not structured (i.e. with subheadings) although some examples do have such an abstract
word limit 150 words – not included in the overall word limit.
The final review and table must include at least six papers – more than 20 is not recommended, 10-12 is ideal. Although I have stipulated 20 references these can be from other sources that may or may not be included in the actual study.