- Cost above baseline for 28% to 86% of projects, depending on which source you read
Protocol 1.0; shares use a Wilson 95% CI. The sources disagree beyond sampling noise (chi² 114, df 5), so no pooled figure is published.
- Cost more than 10% above baseline for 19% to 78% of projects, depending on which source you read
Protocol 1.0; shares use a Wilson 95% CI. The sources disagree beyond sampling noise (chi² 47, df 5), so no pooled figure is published.
- Cost growth of 12% to 109% among the projects that moved above baseline, depending on which source you read
Protocol 1.0; medians use a bootstrap 95% CI (2000 resamples, seed 0). The sources disagree beyond sampling noise (chi² 110, df 5), so no pooled figure is published.
- Finishing later than planned for 39% to 89% of projects, depending on which source you read
Protocol 1.0; shares use a Wilson 95% CI. The sources disagree beyond sampling noise (chi² 276, df 5), so no pooled figure is published.
- 5 to 40 months of slip among the projects that slipped, depending on which source you read
Protocol 1.0; medians use a bootstrap 95% CI (2000 resamples, seed 0). The sources disagree beyond sampling noise (chi² 49, df 4), so no pooled figure is published.
- 11% of 555 completed projects, from one source (Victorian IT Project Dashboard) and not a sector figure
Protocol 1.0; shares use a Wilson 95% CI. Only one source reports this measure, so there is nothing to pool it against.