Selection criteria are important to determine in advance of screening records and are especially important for reviews that aim to minimize the chance of bias.
Inclusion criteria: variables that are required for an article to be eligible for inclusion in a literature review. Examples: experimental design, human subjects, English language, match of population/intervention/outcome
Exclusion criteria: variables that render an article inappropriate for inclusion. Exclusion criteria may not be as easy to identify in advance, but may become apparent with screening a test batch of records. Examples:healthy subjects (rather than patients), mixed age groups, funding from pharmaceutical industry.
Chapter 7 of the Cochrane Handbook provides an overview of the process and tools available.
Assessing risk of bias in a randomized trial
Chapter 8 provides specific guidance for assessing risk of bias and quality of randomized studies.
Resources from GRADE working group, an international collaborative that has produced a standard for grading quality of evidence and strength of recommendations.
Tools developed for assessing quality of various study types for guidelines and systematic reviews from the National Heart, Lung, and Blood Insitute and developed in collaboration with groups recognized internationally as EBP leaders.
Created by librarians at Duke University, this resource provides an extensive list of risk of bias or quality assessment tools for evidence synthesis projects.