Table 1 Critical appraisal tool for study validity assessment
Category | Bias and generic data quality features | Specific data quality features | Validity | Design of assessed study |
|---|---|---|---|---|
1 | Selection bias: study design | Design (i.e., well-controlled) | High | BACI |
Medium | BA or CI | |||
Replication | High | Replication at level of intervention (n > 5) | ||
Medium | Replication at level of intervention (true n = 1 with pseudoreplication) or n = 2–5 | |||
Low | Unreplicated (i.e. no variance, or variance across years only) | |||
Control matching | High | Control and treatment samples well-matched or likely well-matched at baseline | ||
Medium | Control and treatment samples moderately matched | |||
Low | Control and treatment samples poorly matched | |||
2 | Assessment bias: measurement of outcome | Measured outcome | High | Quantitative |
Medium | Quantitative approximations or semi-quantitative | |||
Low | Qualitative | |||
3 | Performance bias: baseline comparison | Other confounding environmental factors | High | Intervention and comparator sites homogenous |
Medium | Intervention and comparator sites moderately comparable with respect to confounding factors | |||
Low | Intervention and comparator sites hardly comparable with respect to confounding factors or lacking sufficient information to judge | |||
N/A if BA design and before measurement taken immediately prior to restoration |