Confidence Evaluation Formats

For verbal confidence evaluation, you will be invited to select the appropriate verbal term from a pre-specified list (that used by the IPCC).

For numerical confidence evaluation, three options are available:

  • Numerical
  • Numerical (aided)
  • Intervals

Note that you can change the evaluation option during reporting (using the Events tab), and, as far as possible, existing evaluations will be translated into the language of the new format.


Specify your confidence evaluation of the probability judgement at issue directly using the slider, on the 0-100 scale described here (note that only even numbers are possible, because the numbers correspond to numbers of draws, half of which are red).

The display recalls the interpretation of the confidence scale in terms of confidence in a certain probability judgement after a (varying) number of observations. You should place the pointer at the number of observations for which you are equally confident in the probability judgement to be evaluated and the 40%-to-100% assessment about the urn.

Numerical (aided)


It may not be easy to directly provide a numerical confidence evaluation for a given probability judgement. This entry format provides support, by asking a series of confidence comparisons: are you more confident in the probability judgement on the table, or in the 40%-to-100% probability of red in the urn case, after a certain number of observations (50 in the Figure)? Answering several such questions, the tool will recommend (using a standard elicitation technique) a confidence evaluation, on a slider like that in previous Numerical case. You may refine the suggestion before confirming. Don’t forget to confirm when you are ready by clicking on Save.


“Precise” confidence numbers may not be needed. The interval format allows you to asses your confidence by selected one of a list of pre-specified intervals on the numerical scale which best captures it. An urn figure recalls the interpretation of the scale.

… Continue to User Guide