Wald Compatibility Curve

Reconstruct which candidate effect sizes are more or less compatible with an observed estimate and reported 95% confidence interval under a one-parameter Wald approximation.

This educational research tool does not calculate an exact profile likelihood, validate a clinical threshold, or provide clinical decision support.

Observed estimate and interval

Supply a published two-sided 95% confidence interval. The point estimate is optional because the interval midpoint determines the reconstructed estimate on the working scale.

Effect measure

Point estimate (optional)

Lower 95% CI

Upper 95% CI

Null value (optional)

Defaults to 1 for ratios and 0 for additive effects.

Reference thresholds or MCIDs (optional)

Separate values with commas or spaces. They are displayed as user-supplied reference markers, not validated claim rules.

Display lower (optional)

Display upper (optional)

A plausible display range changes only the plotted and exported grid, never the reconstructed summary.

Advanced display controls

Ratio-axis spacing

Grid points

Show 90%, 95%, and 99% compatibility guides

Compatibility result

Reconstructed summary

Reference-threshold lookup

No reference thresholds supplied. Add one or more values to compare their compatibility with the observed estimate.

Compatibility at user-supplied candidate values
Natural/display value	Working-scale value	Compatibility / two-sided Wald p-value	Relative to estimate	Relative to null	Inside reported 95% CI

Figure caption

Exports

When the local runtime is ready, update the curve to populate the plot, reconstructed summary, threshold lookup, caption, and exports.

Higher curve values indicate greater compatibility with the observed estimate under the reconstructed Wald model. They are not posterior probabilities that a candidate effect is true.