What is good quality research?
A guide from the UKCIS Evidence Group
Research comes in many shapes and forms. It employs qualitative, quantitative and/or mixed methods depending on the research question being asked. Data also comes in many shapes and forms and not all of it qualifies as ‘research’ or ‘evidence’. Without trying to summarise the vast published literature on the nature, quality, conduct and uses of research, we note key points that research users should have in mind when faced with a new report or article. Our focus is on research relevant to the UK Council for Internet Safety. We have in mind research with children, although other populations may also be studied.
Good quality research provides evidence that is robust, ethical, stands up to scrutiny and can be used to inform policy making. It should adhere to principles of professionalism, transparency, accountability and auditability.
Design and data collection
- Ensure that the research design and methods are appropriate for the research topic or question.
- Ensure that sampling is fit for purpose – e.g. if a survey, use a sufficiently large and representative sample to enable meaningful statistical analysis; not using qualitative data for quantitative conclusions; ensure that qualitative interviewees are purposively selected to reflect the range and diversity of the population of interest.
- Employ researchers trained to work with children (or vulnerable groups), as appropriate, and ensure that sensitive issues are ethically addressed.
- Be aware of the strengths and limitations of the methodology, data collected or findings reported.
- Report the actual questions asked of children and the exact dates and circumstances of data collection.
- Publish the findings of the research along with an account of its methodology (sufficient for another researcher to replicate the study).
- If key percentages are to be released to the press, this should be linked to publication of the report (and not precede it).
- Report findings (e.g. percentages) with a clear statement of the sample or subsample (e.g. children aged 9-16; girls who use the internet aged 9-16).
- Qualitative reports should present the range and diversity of views expressed.
- Report comparisons and trends conscientiously so as not to exaggerate minor differences (e.g. use statistical analysis) or confuse correlation with causation (e.g. higher internet use plus lower grades may not mean internet use reduces grades) or other interpretative errors.
- Avoid describing findings in headline-grabbing, exaggerated or panicky ways.
- Disclose any funding sources and any potential conflicts of interest.
- Ensure the full research report is accessible and/or provide contact details for interested parties to follow up with the researchers and ask about the research conduct or data analysis.