Table 2.
The 24 graph-functionality features used to describe the 208 retrieved graphical displays for meta-analysis
| Functionality features of meta-analytic plots | |
|---|---|
| 01 - Displays summary outcome point estimate | |
| 02 - Displays summary outcome interval estimate | |
| 03 - Displays heterogeneity summary estimates (e.g., I2, Q; also, inconsistency in network meta-analysis) | |
| 04 - Displays individual study effect-size point estimates | |
| 05 - Displays individual study effect interval estimate | |
| 06 - Displays individual study meta-analytic weight/precision/N (including contribution of comparisons in network meta-analysis) | |
| 07 - Displays individual study names or identifiers | |
| 08 - Displays more than one outcome per study | |
| 09 - Displays individual study significance dichotomously (i.e., significant vs. not) | |
| 10 - Displays individual study significance continuously (i.e., allows to assess how close a study p value was to statistical significance thresholds) | |
| 11 - Informs about the likelihood, or posterior distribution, of meta-analytic parameter values | |
| 12 - Suitable to display association of effect sizes with categorical study features | |
| 13 - Suitable to display association of effect sizes with continuous study features | |
| 14 - Suitable to display individual study or study-group features (additionally or exclusively) | |
| 15 - Suitable and informative for small-sized meta-analyses (10 studies or less) | |
| 16 - Suitable and informative for medium-sized meta-analyses (say, about 50 studies) | |
| 17 - Suitable and informative for large-sized meta-analyses (say, hundreds of studies) | |
| 18 - Suitable to assess small-study effects/publication bias and other forms of biases | |
| 19 - Suitable to assess the temporal development of meta-analytic estimates | |
| 20 - Suitable to assess an excess of between-study (or study-group) effect heterogeneity (also, inconsistencies in network meta-analysis) | |
| 21 - Suitable to assess assumptions about the distribution of estimates (e.g., normality of effects) | |
| 22 - Suitable to assess the robustness of summary effect(s) | |
| 23 - Suitable to assess the robustness of heterogeneity statistics (e.g., I2, τ2, Q) | |
| 24 - Suitable to identify influential studies (i.e., outliers, leverage points) |