hearz.uk

Estimation and reporting

Results and analysis

Statistical analysis methods and sanitised result summaries.

Every completed run is saved on the server. This public repository contains only a sanitised result summary: no participant id, no raw response, and no device fingerprint.

Analysis methods

These procedures operate on observations collected by the test methods. They do not define participant tasks.

ANALYSIS PSYCHOMETRIC PROFILE

Estimate a response curve, threshold, slope, lapse rate, and uncertainty.

ANALYSIS FACTORIAL MODEL

Estimate main effects and interactions from a declared factorial design.

ANALYSIS CUE-TRADING MODEL

Estimate the relation between an imposed cue and its cancelling setting.

Analysis by test method

Test methodPrimary analysis
LIMITSEstimate separate ascending and descending transition points and their difference.
ADJUSTMENTEstimate the mean setting, within-participant dispersion, and starting-direction bias.
CONSTANTFit a psychometric function to response probability at each fixed stimulus level.
STAIRCASEEstimate threshold from the final reversal points and quantify reversal dispersion.
BAYESReport posterior estimates and intervals for threshold, slope, and lapse rate.
TRACKINGEstimate the response midpoint after reaction-delay correction and smoothing.
CONFIDENCEConvert hit and false-alarm rates to sensitivity and response criterion.
DISCRIMINATIONConvert proportion correct to sensitivity for the specified forced-choice design.
SCALINGFit the relation between physical magnitude and numerical or cross-modal response.
PAIRWISEFit a paired-comparison model with item and participant effects.
ORDERINGFit rank or rating models with item, order, and participant effects.
REPORTINGEstimate first-percept latency, state durations, transition rates, and censoring.
ANOMALYEstimate event detection, false-alarm rate, response latency, and temporal drift.
VIDEOCompare auditory-response distributions across congruent and incongruent conditions.

Saved results