cardioception.scales.ConfidenceScale

class cardioception.scales.ConfidenceScale(kind='vas', low=0.0, high=100.0, granularity=1.0, labels=('Guess', 'Certain'))

How confidence is collected and what its numbers mean.

Parameters

kind

"vas" for a continuous visual analogue scale, "discrete" for a fixed number of steps.

low, high

Ends of the scale, inclusive.

granularity

Step size. 1 on a 0-100 VAS gives 101 positions; 1 on a 1-7 discrete scale gives 7.

labels

Text for the two ends.

__init__(kind='vas', low=0.0, high=100.0, granularity=1.0, labels=('Guess', 'Certain'))

Methods

__init__([kind, low, high, granularity, labels])
believes_correct(value)	Whether the participant thought they were right.
clamp(value)	Snap a raw position onto the scale.
describe()	Columns identifying this scale, written on every trial.
from_unit(unit)	Inverse of to_unit(), snapped to the scale.
magnitude(value)	Confidence, on 0-1, ignoring which outcome was believed.
to_unit(value)	Rescale to 0-1, which is what a bounded model wants as input.
with_labels(labels)	The same scale with translated end labels.

Attributes

bounds
granularity
high
kind
labels
low
midpoint
n_levels	Number of positions a participant can select.
signed	Whether the scale spans zero, so its sign carries meaning.
ticks	Tick positions for a slider, one per label.