Validity and reliability--what's it all about? Part 3: Issues relating to qualitative studies.

Validity and reliability in quantitative research were discussed in the November and December 2004 issues of Paediatric Nursing. Judging the quality of qualitative research requires consideration of different factors which are discussed this month. Rather than validity and reliability, qualitative researchers talk about rigour. Rigour relates to the overall planning and implementation of the research design and is concerned with whether study has been carried out in a logical, systematic way. Rigour is described as: ‘The striving for excellence in research through the use of discipline, scrupulous adherence to detail and strict accuracy’ (Burns and Grove 1997, p793) When considering the rigour of a study you need to consider credibility, transferability and auditability (Lincoln and Guba 1985). A measure of the credibility of the data can be obtained by the degree of confidence the researcher inspires in us (Lincoln and Guba 1985, Carter and Porter 2000). The researcher can assure the reader of the credibility of the study through: n use of prolonged data collection – for example, when observing nurses’ pain management practices Alison collected data over a four-month period n verification or member-checking – the process of returning interview transcripts to participants who are asked to check them for accuracy n theoretical verification – where the findings of the study are compared to the results of previous studies (Morse and Field 1995). Transferability relates to the extent to which we can see similarities in the findings that may relate to other settings. This requires the researcher to provide a lot of detail about the setting and the events taking place, and is described as thick description (Lincoln and Guba 1985, Morse and Field 1995). The detailed description allows the reader to assess whether the people and events could be applicable elsewhere, and this is sometimes referred to as applicability or fittingness. For auditability the researcher needs to provide a sufficiently clear and full account of the research process so that the reader can judge the dependability of the qualitative study (Lincoln and Guba 1985). This has been described as a decision trail or audit trail (Morse and Field 1995). An example of an audit trail can be seen in Figure 1 where the stages taken by one of the authors (AT) when analysing some qualitative data are outlined. It is not necessary to provide this information in a flowchart but doing so makes it easier for the reader to follow the steps taken. The steps taken throughout the study, and not just during data analysis, should be clearly identified (Field and Morse 1995). Unlike in quantitative studies the information you need to assess whether a research study is of a good quality is not always placed within the methodology section of a paper. You may need to seek this information in other sections. This is one of a series of short papers on aspects of research by Alison Twycross and Linda Shields