Audit condition-level quality imbalance — audit_gazepoint_condition_quality

Create a publication-level audit of whether gaze, pupil, retention, or other quality metrics differ across experimental conditions.

Usage

audit_gazepoint_condition_quality_imbalance(
  data,
  condition_col = "condition",
  quality_cols = NULL,
  subject_col = NULL,
  min_units_per_condition = 1L,
  max_mean_difference = 0.1,
  max_condition_ratio = 2,
  lower_is_better = c("missing_gaze_prop", "offscreen_prop", "excluded_prop",
    "failure_prop", "artifact_prop")
)

Arguments

data: A data frame containing condition-level, unit-level, or subject-condition-level quality metrics.
condition_col: Condition column.
quality_cols: Numeric quality-metric columns. If NULL, common quality columns are detected automatically.
subject_col: Optional subject column.
min_units_per_condition: Minimum number of rows/units expected per condition.
max_mean_difference: Maximum acceptable absolute difference between condition means for each quality metric.
max_condition_ratio: Maximum acceptable ratio between the largest and smallest non-zero condition mean for each quality metric.
lower_is_better: Optional character vector naming metrics where lower values indicate better quality, such as missing-gaze or exclusion metrics.

Value

A list with class gp3_condition_quality_imbalance_audit containing overview, condition_summary, metric_summary, flagged_metrics, and settings tables.