Primary research is like growing your own crops collecting original data, while secondary research is like shopping at a farmers' market—analysing what's already been gathered by others then creating a new dish.

These can be broken down further..........think of these methods along a continuum.  At one end highly controlled experiments, the other less controlled qualitative research. NB Both are equally valuable in output!!

There are various types of Secondary Reviews.  Secondary reviews is a research approach that synthesises and analyses data from previously published primary studies, reports, and other secondary sources, rather than collecting new empirical data directly. The key purposes are to provide a comprehensive overview of the current state of knowledge on a topic and to inform future research, policy, and practice.   Here are some review methods:

Least Potential Bias:

• Cochrane Reviews (most steps): These highly rigorous reviews follow strict methodological guidelines and often involve multiple reviewers at every stage.

Moderate Potential Bias:

• Standard Systematic Reviews: These follow established methods but might have some flexibility depending on the research question. Reviewers are still crucial to minimise bias throughout the process.  These systematic reviews (least bias to most) can be split into:

• • Meta-Analysis (least bias): This approach synthesizes quantitative data from multiple similar studies, combining effect sizes to produce a single, comprehensive estimate. By pooling data, meta-analyses offer increased statistical power and a more robust evaluation of the research question. However, this method is most effective when studies are sufficiently homogeneous in terms of population and outcome measures. When significant heterogeneity exists among studies, a quantitative review (below) may be more appropriate.

  • Quantitative Review: focuses on numerical data from research studies that are not similar. It analyses the findings of multiple studies using statistical methods but doesn't necessarily combine them into a single summary measure. Quantitative reviews are broader, encompassing any systematic review that analyses numerical data, populations and outcomes may differ.

  • Narrative Synthesis: This review would summarise findings from various studies (qualitative and potentially quantitative) on online learning. It would highlight common themes across studies and describe the overall picture of student experiences.

  • Qualitative Review (most bias): This review would delve into interview transcripts, analysing themes around student perceptions, challenges, and benefits of online learning.

Moderate Potential Bias (Structured / Systematic route):

• Rapid Reviews: Conducted with a tighter timeframe, these reviews might involve fewer reviewers or streamlined processes in some stages (like data extraction). This can increase the risk of bias, but it's balanced by the need for faster results.

Other less structured reviews (Unstructured route)

Moderately high Bias:

• Critically Appraised Topics (CATs): These are concise summaries of the best available evidence on a specific topic. Compared to systematic reviews, they are less rigorous and often involve fewer reviewers. This can introduce bias, particularly selection bias during the search and data extraction stages.

Higher Potential Bias:

• Narrative Reviews: While still valuable for summarising existing research, these reviews often rely on a single author's interpretation, leading to a higher potential for bias in selection and analysis.

NB For more detailed descriptions click this link

For more depth go to: 

Experimental

• Quantitative Research (Primary Research)

Quasi Experimental

Non Experimental

• Qualitative Research (Primary Research)

• Reviews (Secondary Research)

References

Last J (2001) A dictionary of epidemiology, 4th edn. Oxford University Press, New York

The pyramid was produced by HLWIKI Canada: http://hlwiki.slais.ubc.ca/index.php/File:EBMpyramid.gif