Meta-Analysis

Where insufficient data exist for meta-analysis, findings will be presented as a narrative synthesis with descriptive tables and visual aids. Random-effects meta-analysis will be conducted when at least three homogeneous studies report the same statistic. Results will be organized by tool and measurement property using tables and narrative synthesis. Diagnostic accuracy outcomes will be pooled using bivariate or hierarchical summary receiver operating characteristic models. Reliability outcomes will be pooled using Fisher Z transformations. Funnel plots are used to explore small-study effects and potential publication bias. A fixed effects model is used when five or fewer studies are included; a random effects model applies when more than five studies are available. Heterogeneity is quantified using the I² statistic, with values exceeding 50% considered substantial. Heterogeneity will be quantified with I², τ², and H² statistics. Pooled prevalence estimates will use random-effects models such as logit-transformed models or binomial GLMMs. The review will analyze subjective, behavioural, and physiological markers separately because they reflect different neurocognitive processes. T…