Data analysis

  • How did the authors analyse their data?
  • Is it appropriate to the research approach?
  • Validity? For example, if a study investigates braking distance in cars aged over ten years and above, how might the age of the cars affect the results?
  • Reliability. Can any of the test results obtained in the article be reproduced?
  • Blindness? This refers to the amount of potential bias in the study. 
  • Completeness:
    • Compliance. This can be a problem in studies in health. If there is compliance in a study, this may have the effect of reducing any apparent improvement in a new system.
    • Missing data? Check that data has not been missed out in the final results section. If no explanation is given for this, then there is the potential that the result is biased and the validity of the findings compromised.
Qualitative approach - Usually content analysis or thematic analysis.
  • Do the authors show in detail the process that occurred to get to the final categories (auditability)?
  • Do the authors show the categories of themes by using participant dialogue (quotes) or field notes to represent the themes (credibility)
  • Do more than one researcher analyse the data (peer review)?
  • How does all this affect the findings?
Quantitative approach - Usually either descriptive or inferential statistics
  • What statistical tests did the researcher use?
  • Did the researcher explain why they used descriptive and/or inferential statistics?
  • Did they explain which tests they used and the relevance to the data?
  • Are the tables and graphs easy to understand and well explained?
  • This all affects validity of the findings
  • What is the overall effect of this on the results?

Creative Commons Licence

SMILE - How to assess a research article by GCU School of Health and Life Sciences modified by Marion Kelt, Glasgow Caledonian University is licensed under a Creative Commons Attribution 4.0 International License.