Quality assessment of research articles
If you are doing research project or a master thesis, you may need to assess the quality of the research articles you include. Depending on the study design your project has, different approaches and level of rigour may be required.
General assessment tips for research articles
-
Study design- is it suitable for the research question?
-
Study population size (sample size) and composition- is the intention of the study general conclusions or are individual variations in focus?
-
How is the study participants selected?
-
Dropouts- how are they explained, and is the statistical analysis used appropriate?
-
Confounders- how may these affect correlation or causality presented in the study?
-
Is there a risk of bias (methodological flaws) that may affect the credibility of the result?
-
Are conflicts of interest reported? E. g. patent ownership, participation in corporate advisory boards.
-
Is financial support to the study reported?
Evidence hiearchies
The certainty and strength of the evidence in a research article may depend on the study design. Different study designs have pros and cons. Depending on your research area, different study designs may be included in a relevant hiearchy of evidence.
Video Evidence hierarchies (6:46 min)
Risk of bias assessment
Risk of bias is a method for assessing the methodological quality of research articless. Assessing risk of bias is an essential part of e. g. a systematic review and in evaluating research. There are checklists for reporting of studies, and for risk of bias assessment, both types are depending on study design. Reporting checklists may be a starting point, to check how well a study is reported. A full risk of bias assessment is usually done with a risk of bias checklist, which lists signalling questions for each included study as part of a structured assessment. A risk of bias checklist also adressess how well a study is conducted, not just reported.
Reporting guidelines- EQUATOR networkLinks to an external site.
Reporting guidelines for different study designs may help you reporting flaws in studies.
RoB2 and ROBINS checklistsLinks to an external site.
Risk of bias checklists for randomized and non-randomized studies of interventions.
Newcastle-Ottawa ScaleLinks to an external site.
Risk of bias checklist for observational studies
JBI Manual of Evidence SynthesisLinks to an external site.
Method tips for qualitative systematic reviews
Tools for evidence grading
If you do a master thesis you may need tools for evidence grading. The GRADE method is international standard for evidence grading. The GRADE handbook and website can give you further guidance.