Table 2 Risk of bias and quality assessment criteria
Research design | Criteria a | Satisfied if |
|---|---|---|
Qualitative | Credibility | 1.Collection of data over a prolonged period and from a range of participants 2.Use of a variety of methods to gather data 3.Use of a reflective approach through keeping a journal of reflections, biases, or preconceptions and ideas 4.Triangulation used to enhance trustworthiness through multiple sources and perspectives to reduce systematic bias. Main types of triangulation are by: sources (people, resources); methods (interviews, observation, focus groups); researchers (team of researchers versus single researchers); or theories (team bring different perspectives to research question) 5.Member checking |
Transferability | 1.Can the findings be transferred to other situations? 2.Are the participants and settings described in enough detail to allow for comparisons with your populations of interest? 3.Are there concepts developed that might apply to your clients and their contexts? 4.Were there adequate (thick) descriptions of sample and setting? | |
Dependability | 1.Is there consistency between the data and the findings? 2.Is there a clear explanation of the process of research including methods of data collection, analysis and interpretation often indicated by evidence of an audit trail or peer review? 3.An audit trail described the decision points made throughout the research process | |
Confirmability | 1.What strategies were used to limit bias in the research, specifically the neutrality of the data not the researcher? For example, was the researcher reflective and did they keep a reflective journal, peer review such as asking a colleague to audit the decision points throughout the process (peer audit) and checking with expert colleagues about ideas and interpretation of data, checking with participants (participant audit) about ideas and interpretation of data and having a team of researchers | |
Quantitative | Sample | 1.Sample is representative 2.Selection bias reduced: population based/representative/convenient 3.Size of study in relation to design and question (power) 4.Clearly described participant characteristics |
Measure | 1.Measure is valid for purpose and reliable 2.Measurement bias is reduced: validity of tools for purpose/reliability of tool/recall/memory | |
Analysis | 1.Analyses are appropriate to the research question and outcome measure 2.Statistical significance reported 3.Point estimates and variability provided and clinical importance discussed |