Protocol

Human visual contrast 2AFC keyboard task

Human observers judge which of two visual alternatives contains the stronger or target contrast signal, commonly using keyboard or button responses. The first linked slice uses overlaid flickering gratings with probabilistic cues and points feedback. Contrast is varied to estimate psychometric threshold and lapse parameters.

Task structure

Stimulus, choice, feedback, and training as declared in the protocol record.

Stimulus

evidence_schedule
Contrast levels are sampled across trials, often with fixed grids or adaptive procedures designed to estimate a psychometric function. The first linked slice uses session-titrated target/distractor contrast differences and valid, invalid, or neutral cues.
evidence_type
adaptive
modalities
visual
notes
—
units
percent contrast, Michelson contrast
variables
contrast, stimulus interval, stimulus side, probabilistic cue, flicker frequency

Choice

action_mapping
Keyboard or button responses indicate which side or interval contained the target visual contrast.
alternatives
interval A or left, interval B or right
choice_type
2afc
notes
—
response_modalities
keyboard, button-press

Feedback

feedback_type
mixed
notes
Feedback is protocol-dependent; the first linked slice uses points feedback.
penalty
—
reward
—

Training

notes
Training is usually minimal relative to animal implementations.
stages
instructions, practice trials

Trial timing

  1. fixation
    protocol-specific

    Observer fixates before stimulus presentation.

  2. stimulus
    protocol-specific

    One or more contrast-defined visual stimuli are shown.

  3. response
    response-limited

    Observer reports the chosen interval or side.

Vertical slices

Expected analyses

  • psychometric curve by contrast
  • threshold and lapse estimates
  • bootstrap confidence intervals

Interpretive claims

  • visual sensitivity
    Confidence: medium · ref.wichmann-hill-2001

Open questions

  • Split side-based and interval-based variants if detailed datasets are linked.
  • Decide whether the Walsh SSVEP prior-cue task should eventually become a more specific concrete protocol variant.
  • Add display calibration and staircase details for source-specific records beyond the current processed stats slice.

Papers

Findings

Curves attached to this protocol on shared signed-evidence axes.

See all findings overlaid on shared axes.