AOI modelling workflow

This article summarises AOI-level modelling preparation.

AOI measures are visual-attention measures. They can indicate allocation and duration of gaze within defined areas, but they do not directly prove interpretation, scrutiny, preference, or comprehension.

Example workflow

aoi_summary <- summarize_gazepoint_aoi(
  all_gaze,
  group_cols = c('participant_id', 'trial_id', 'AOI')
)

aoi_glmm <- prepare_gazepoint_aoi_glmm_data(
  aoi_summary,
  outcome_col = 'dwell_time_ms',
  participant_col = 'participant_id',
  trial_col = 'trial_id'
)

aoi_logit <- transform_gazepoint_aoi_empirical_logit(
  aoi_glmm,
  success_col = 'aoi_samples',
  total_col = 'total_samples'
)

Model-readiness checks

Before fitting models, inspect zero-inflation, trial coverage, repeated-measures structure, missing AOIs, and whether the outcome is a count, duration, proportion, or transformed proportion.

Reporting note

Report AOI definitions, screen coordinate system, overlap rules, dwell/fixation definitions, and any transformation used before modelling.