Table 1.

Significance of mixed-effects models when separately accounting for each covariate

PredictorSaliencenObservers
F(1,16)pF(1,16)p
isLocTemp4.90.045.70.03
visDist14.30.00218.70.0005
visCorr18.30.000621.50.0003
visHistDist16.50.000919.70.0004
lumDiff17.20.000821.80.0003
DCNN[1… 6]minF = 12.8maxP = 0.003minF = 15.6maxP = 0.001
psdCorr170.000820.10.0004
psdDist18.60.000522.50.0002
absVolDiff18.60.0005220.0003
V1Betas9.80.00611.90.003
A1Betas16.80.000819.50.0004
isAG16.90.000819.70.0004
  • Shown are the results of the mixed-effects models separately accounting for each of the perceptual confounds and the shifts in time/location. Results are presented both for models with salience as the effect of interest and for models with nObservers as the effect of interest. All models were fitted using the following formula template (replacing <covariate> with each of the potential predictors):

    betas ∼ <covariate> + salience/nObservers + (1 participant) + (1 event boundary).