The literature review contains the researcher’s appraisal of relevant literature associated with the research problem. It should include up-to-date references and provide a critical perspective of the existing literature. A combination of theoretical and research-based knowledge on the topic and sample population, where possible, should be presented. Biased opinions should not be evident and it is important that the researchers have provided evidence which supports as well as contradicts personal opinions (Marshall 2005). A good literature review draws from a wide range of sources, contains relevant information and provides a summary of current knowledge on the research topic (Carnwell & Daly 2001). It should begin broadly, narrow down to the specific problem and conclude with a clear justification for the research (Russell 2005). Secondary sources should only be used when absolutely necessary as the reader is subsequently analysing another’s interpretation of the study. The review should highlight flaws in previous research methods that the current study will overcome (Summers 1991). The literature review, through identification of what is already known on a given topic, provides a rationale for the proposed study. The literature should not be presented as a descriptive narrative. Instead, conclusions should be drawn from individual research studies and analysed and compared to others. Research variables should be clearly defined to ensure consensus on what is actually being investigated. Overall the review should flow logically, include all major research previously conducted in the area and be consistently referenced. Word limitations in many journals mean that literature reviews are often very short, which can result in difficulties making balanced judgements about the review. Sometimes reviews are published separately and a quick look at the reference listing should confirm if this is the case.

When evaluating the research methodology, ask the question ‘could I replicate this study?’ The provision of clear accounts on how the research was conducted is essential. In the methodology section there are a number of subsections, which mainly include the research design, sampling, data collection procedure, data analysis, results and ethical issues. A well-defined research problem is paramount to assessing whether the chosen design was able to address the problem appropriately (Rowan 1997). The aims, objectives, questions and hypotheses should be clearly stated and consistent with the research problem. Evaluation should be concerned with judging how capable the aims and objectives are of answering the research problem.

The research design or approach should be explained sufficiently and include the identification of theoretical or conceptual frameworks used to underpin the study (Hek 1996). Evaluation of the design includes assessing the appropriateness of the approach to the research questions posed, the aims and objectives. For example, if the researcher is investigating ‘lived experiences’, then a phenomenological design is the most suitable; however, if exploring the relationship between variables, then a quantitative correlation design is appropriate (Polit & Beck 2004). In some instances a number of approaches may be suited to a particular study and evaluation should be concerned with how the researchers justify using the chosen design. The research design should be reflected in the research problem, sampling procedure and methods used in data collection and analysis (Benton & Cormack 1996).

The sampling procedure should be described in detail, incorporating sample size and inclusion/exclusion criteria, as well as the rationale for the sampling procedure implemented. It is important to evaluate whether or not the sampling procedure was used appropriately by the researchers, i.e. was the sample type in keeping with the design. For example, in a quantitative experimental study, randomisation is the recommended sampling method (Burns & Grove 2001). Enough information should be provided to consider whether the sample selection method was valid. Questions surrounding the issue of sampling bias should also be raised. Sampling bias occurs when there is over- or under-representation of any characteristics of relevance to the research problem (Polit & Beck 2004). Consideration of whether the sample is large enough and representative of the population under investigation is important. While small samples may be appropriate in qualitative studies, a larger sample size is required for statistical analysis in quantitative research. Research reports should be specific about size and how researchers arrived at the number of subjects. The sample should be sufficiently explicit to allow informed judgements about whether the findings can be generalised to others (McCaughan 1999).

There are ethical issues in all research and it is necessary to evaluate whether appropriate ethical approval was obtained for the research. Normally, this is achieved after submission of a detailed proposal to a local ethics committee. Specific description of subject recruitment and retention should be provided. Subjects should be given adequate information, and be assured of their rights to confidentiality, anonymity and freedom to withdraw from the study at any time. It is also important to establish whether ethical principles such as informed consent were upheld throughout the research (Seals & Tanaka 2000) and whether data storage and disposal were addressed.

Evaluating the methods used to collect data is important as any flaws in data collection may undermine findings. Procedures taken to collect data should be described in detail. How key concepts were defined is important. The relationship of this definition to method of data collection should be explicit. It is important to decide whether the way data were collected was appropriate to the specific data required. Irrespective of how data were collected, sample questions or interview topics should be provided so that the reader can make informed judgements concerning the relevance of the questions asked (Beeman 2002). Ascertaining whether a pilot study was conducted is a useful way of evaluating if potential problems with data collection methods were identified and altered (Polit & Beck 2004).

The conclusions or results reached are largely dependent on the adequacy of data analysis. It is important to evaluate whether the researchers analysed the data in the most appropriate way (Seals & Tanaka 2000). Adequate descriptions of how data were analysed should be provided and in keeping with the research design. Owing to the different techniques quantitative and qualitative researchers use to analyse data, these issues will be dealt with separately below.