Confounding

Confounding

1. Controlling

   • •randomisation
   • •restriction (use of inclusion/exclusion criteria)
   • •matching
   • •the analysis stage of research (i.e. stratification and modelling).
   • design of study methods before the data is collected
   • through analysis after the data has been collected
   • Study Designs

2. Systematic reviews

   • sum up the best available research on a specific question. This is done by synthesising the results of several studies.
   • uses transparent procedures to find, evaluate and synthesise the results of relevant research.
   • Procedures are explicitly defined in advance
     • in order to ensure that the exercise is transparent and can be replicated
   • also designed to minimise bias
     • relation to the articles selected for inclusion in the systematic review
       • Studies included in a review are screened for quality
         • the findings of a large number of studies can be combined
         • Peer reviewed
   • Criteria
     • •clear inclusion/ exclusion criteria
     • •an explicit search strategy
     • •systematic coding and analysis of included studies
     • •meta-analysis (where possible).
       • involves statistically combining the results of several separate studies to arrive at an overall result, which is often in the form of an 'effect size'.