Borman et al. included 29 CSR models that had been replicated in at least 10 schools and had been evaluated in at least one study. They did an extensive search for all types of studies. Studies were included if they met the following criteria:
- Schools using each program had to be compared to equivalent control groups
- Achievement data had to be sufficient to allow effect sizes to be computed
- Schools had to be in the U.S.
- Students had to be in the regular education program
CSR programs were rated in terms of the strength of their evidence of effectiveness in four categories:
Strongest Evidence of Effectiveness
Highly Promising Evidence of Effectiveness
Promising Evidence of Effectiveness
Greatest Need for Additional Research
Programs were assigned to categories based on a combination of three criteria (Borman et al., 2003, p. 154):
- Quality of evidence: Does the CSR model have research evidence from control-group studies and third-party control group studies?
- Quantity of evidence: Does the CSR model have a relatively large number of studies? (10 studies overall and 5 third-party studies were required to be in the top category)
- Statistically significant and positive results: Does the evidence from control-group studies show effects that are significantly positive?