Order effects
Hackathon
2024-08-27
Goal: understand order effects.
DVs:
- we primarily care about outcomes like RT and accuracy (as we are measuring), but also care about level of “has-rt” (since looking to target at t=0 will reduce that), and do care (some) about looks before and around t=0.
for each comparison, I do
- time course graph on -500 to 2000
- 5 models – -500-0 (if data on this range); 0-250; RT; accuracy on 500-4000; has-rt
IVs:
trial order – for stimuli seen for the first time how does position in order change?
seen prev (as distractor) – for stim how does 1st time compare to seen prev as distractor?
- for models, also look at trial order x seen prev as distractor
seen prev (as target) – for stim how does 1st time compare to seen prev as target?
- for models, also look at trial order x seen prev as target
for datasets where we have it – 3 way on first; prev-dist; prev-target
Prep
## here() starts at /home/vboyce/Research/peekbank-method## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.2
## ✔ ggplot2 4.0.0 ✔ tibble 3.3.0
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.1.0
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
## Loading required package: viridisLite
##
## Loading required package: Matrix
##
##
## Attaching package: 'Matrix'
##
##
## The following objects are masked from 'package:tidyr':
##
## expand, pack, unpackTrial order
both want to look for order effects on say 1-10 and binned for longer!
group trials by 5s (labelled by the early trial of the block)
not seeing much that’s consistent trends that would be worrisome
## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_early ~ trial_order + (1 | administration_id) + (1 |
## target_label) + (1 | dataset_name)
## Data: trial_order_mod
##
## REML criterion at convergence: 28358.8
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5414 -0.9986 -0.1133 1.0506 1.5886
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.010519 0.10256
## dataset_name (Intercept) 0.001612 0.04015
## Residual 0.214378 0.46301
## Number of obs: 21713, groups:
## administration_id, 3103; target_label, 140; dataset_name, 21
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.5115196 0.0165440 30.919
## trial_order 0.0006717 0.0003977 1.689
##
## Correlation of Fixed Effects:
## (Intr)
## trial_order -0.212
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula:
## correct_late ~ trial_order + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: trial_order_mod
##
## REML criterion at convergence: 33406.2
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6087 -0.9750 -0.5558 1.0214 1.6045
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.010240 0.10120
## dataset_name (Intercept) 0.005313 0.07289
## Residual 0.236572 0.48639
## Number of obs: 23774, groups:
## administration_id, 3471; target_label, 146; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.4916199 0.0203474 24.16
## trial_order 0.0007688 0.0004134 1.86
##
## Correlation of Fixed Effects:
## (Intr)
## trial_order -0.163
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula:
## accuracy ~ trial_order + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: trial_order_mod
##
## REML criterion at convergence: 9195
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.2204 -0.6274 0.1586 0.7592 2.4132
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.007539 0.08683
## target_label (Intercept) 0.004948 0.07034
## dataset_name (Intercept) 0.006002 0.07747
## Residual 0.074600 0.27313
## Number of obs: 28713, groups:
## administration_id, 3569; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.6698855 0.0182261 36.754
## trial_order 0.0003107 0.0002215 1.403
##
## Correlation of Fixed Effects:
## (Intr)
## trial_order -0.103## Linear mixed model fit by REML ['lmerMod']
## Formula: has_rt ~ trial_order + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: trial_order_mod
##
## REML criterion at convergence: 37184.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.3346 -0.7313 -0.5962 1.3246 2.0909
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.003332 0.05773
## target_label (Intercept) 0.005551 0.07450
## dataset_name (Intercept) 0.006082 0.07798
## Residual 0.208505 0.45662
## Number of obs: 28789, groups:
## administration_id, 3569; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.2857839 0.0193559 14.765
## trial_order -0.0011606 0.0003592 -3.231
##
## Correlation of Fixed Effects:
## (Intr)
## trial_order -0.150## Linear mixed model fit by REML ['lmerMod']
## Formula: rt ~ trial_order + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: trial_order_mod
##
## REML criterion at convergence: 143207
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2285 -0.5472 -0.2610 0.2511 4.8468
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 68799 262.3
## target_label (Intercept) 12694 112.7
## dataset_name (Intercept) 20041 141.6
## Residual 300519 548.2
## Number of obs: 9168, groups:
## administration_id, 2836; target_label, 133; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 1094.1485 36.6758 29.83
## trial_order 0.4143 0.7821 0.53
##
## Correlation of Fixed Effects:
## (Intr)
## trial_order -0.171no large effects of trial order
Seen prev as distractor
## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_early ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 39740.6
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5860 -1.0157 -0.1019 1.0453 1.7481
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.012418 0.11144
## dataset_name (Intercept) 0.001301 0.03606
## Residual 0.212942 0.46146
## Number of obs: 30612, groups:
## administration_id, 3146; target_label, 146; dataset_name, 21
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.495710 0.015467 32.050
## typeprev_as_distractor 0.011804 0.006269 1.883
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_ds -0.123
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_late ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 46883.3
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6060 -0.9706 -0.5571 1.0109 1.6160
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.0000
## target_label (Intercept) 0.011912 0.1091
## dataset_name (Intercept) 0.004071 0.0638
## Residual 0.236735 0.4866
## Number of obs: 33384, groups:
## administration_id, 3518; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.480192 0.018516 25.93
## typeprev_as_distractor 0.010152 0.006426 1.58
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_ds -0.103
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: accuracy ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 13683.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.2153 -0.6234 0.1647 0.7537 2.4636
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.008531 0.09237
## target_label (Intercept) 0.004745 0.06888
## dataset_name (Intercept) 0.006566 0.08103
## Residual 0.076128 0.27591
## Number of obs: 40523, groups:
## administration_id, 3576; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.663463 0.018502 35.859
## typeprev_as_distractor 0.009176 0.003388 2.709
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_ds -0.055## Linear mixed model fit by REML ['lmerMod']
## Formula: has_rt ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 52328.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.3876 -0.7206 -0.5860 1.3319 2.0838
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.003196 0.05654
## target_label (Intercept) 0.006670 0.08167
## dataset_name (Intercept) 0.005918 0.07693
## Residual 0.208031 0.45610
## Number of obs: 40644, groups:
## administration_id, 3576; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.291475 0.018856 15.458
## typeprev_as_distractor -0.011998 0.005491 -2.185
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_ds -0.087## Linear mixed model fit by REML ['lmerMod']
## Formula: rt ~ type + (1 | administration_id) + (1 | target_label) + (1 |
## dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 199958.5
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4188 -0.5403 -0.2565 0.2312 5.0710
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 67561 259.93
## target_label (Intercept) 9650 98.24
## dataset_name (Intercept) 21825 147.73
## Residual 304245 551.58
## Number of obs: 12803, groups:
## administration_id, 3140; target_label, 140; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 1101.42 35.75 30.81
## typeprev_as_distractor -16.57 12.46 -1.33
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_ds -0.111estimates of around 1% differences in accuracy (preferring prev distractor), sometimes “significant”
## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_early ~ type * trial_order + (1 | administration_id) +
## (1 | target_label) + (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 39762.5
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5921 -1.0203 -0.1042 1.0467 1.7267
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.012290 0.11086
## dataset_name (Intercept) 0.001322 0.03636
## Residual 0.212928 0.46144
## Number of obs: 30612, groups:
## administration_id, 3146; target_label, 146; dataset_name, 21
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.4925459 0.0158878 31.002
## typeprev_as_distractor -0.0056657 0.0136853 -0.414
## trial_order 0.0004921 0.0003882 1.268
## typeprev_as_distractor:trial_order 0.0007840 0.0007606 1.031
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_ds -0.139
## trial_order -0.224 0.187
## typprv_s_:_ 0.140 -0.865 -0.428
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_late ~ type * trial_order + (1 | administration_id) +
## (1 | target_label) + (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 46900.8
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6039 -0.9712 -0.5464 1.0158 1.6161
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.00000 0.00000
## target_label (Intercept) 0.01179 0.10860
## dataset_name (Intercept) 0.00392 0.06261
## Residual 0.23670 0.48651
## Number of obs: 33384, groups:
## administration_id, 3518; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.4787467 0.0186633 25.652
## typeprev_as_distractor -0.0193517 0.0137227 -1.410
## trial_order 0.0004863 0.0004045 1.202
## typeprev_as_distractor:trial_order 0.0015056 0.0007731 1.947
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_ds -0.128
## trial_order -0.187 0.186
## typprv_s_:_ 0.129 -0.859 -0.435
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: accuracy ~ type * trial_order + (1 | administration_id) + (1 |
## target_label) + (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 13704.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.2185 -0.6238 0.1645 0.7536 2.4546
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.008465 0.09201
## target_label (Intercept) 0.004750 0.06892
## dataset_name (Intercept) 0.006383 0.07989
## Residual 0.076147 0.27595
## Number of obs: 40523, groups:
## administration_id, 3576; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 6.641e-01 1.841e-02 36.078
## typeprev_as_distractor -7.011e-03 7.244e-03 -0.968
## trial_order 9.319e-05 2.191e-04 0.425
## typeprev_as_distractor:trial_order 9.192e-04 4.115e-04 2.234
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_ds -0.070
## trial_order -0.107 0.171
## typprv_s_:_ 0.073 -0.858 -0.422## Linear mixed model fit by REML ['lmerMod']
## Formula: has_rt ~ type * trial_order + (1 | administration_id) + (1 |
## target_label) + (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 52345.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.3814 -0.7184 -0.5858 1.3321 2.0806
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.003186 0.05644
## target_label (Intercept) 0.006601 0.08125
## dataset_name (Intercept) 0.005899 0.07680
## Residual 0.208004 0.45607
## Number of obs: 40644, groups:
## administration_id, 3576; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.2981510 0.0190827 15.624
## typeprev_as_distractor -0.0009064 0.0117531 -0.077
## trial_order -0.0009351 0.0003518 -2.658
## typeprev_as_distractor:trial_order -0.0002219 0.0006657 -0.333
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_ds -0.109
## trial_order -0.160 0.181
## typprv_s_:_ 0.111 -0.859 -0.432## Linear mixed model fit by REML ['lmerMod']
## Formula:
## rt ~ type * trial_order + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_dist_mod
##
## REML criterion at convergence: 199949.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4251 -0.5400 -0.2573 0.2326 5.0568
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 67328 259.5
## target_label (Intercept) 9940 99.7
## dataset_name (Intercept) 21652 147.1
## Residual 304262 551.6
## Number of obs: 12803, groups:
## administration_id, 3140; target_label, 140; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 1093.1326 36.3956 30.035
## typeprev_as_distractor 36.9086 27.3291 1.351
## trial_order 0.3651 0.7678 0.475
## typeprev_as_distractor:trial_order -3.2753 1.5213 -2.153
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_ds -0.140
## trial_order -0.182 0.166
## typprv_s_:_ 0.138 -0.865 -0.411but it diminishes and sometimes flips sign if we take trial order into account as well. Possible faster RT to previous distractors over time.
Seen prev as target
## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_early ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 31271.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5433 -0.9966 -0.1147 1.0520 1.5444
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 4.982e-11 7.058e-06
## target_label (Intercept) 9.823e-03 9.911e-02
## dataset_name (Intercept) 1.315e-03 3.626e-02
## Residual 2.150e-01 4.637e-01
## Number of obs: 23908, groups:
## administration_id, 3103; target_label, 141; dataset_name, 21
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.52002 0.01529 34.015
## typeprev_as_target 0.01028 0.01100 0.935
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_tr -0.060
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_late ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 37126.7
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5993 -0.9607 -0.5664 1.0217 1.5955
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.010456 0.10225
## dataset_name (Intercept) 0.005189 0.07203
## Residual 0.237095 0.48692
## Number of obs: 26394, groups:
## administration_id, 3480; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.50316 0.01989 25.301
## typeprev_as_target 0.01082 0.01066 1.015
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_tr -0.037
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: accuracy ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 10791.5
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.1136 -0.6290 0.1592 0.7662 2.4394
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.007454 0.08634
## target_label (Intercept) 0.004849 0.06964
## dataset_name (Intercept) 0.005380 0.07335
## Residual 0.076472 0.27654
## Number of obs: 31777, groups:
## administration_id, 3572; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.67282 0.01733 38.824
## typeprev_as_target -0.02778 0.00570 -4.873
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_tr -0.024## Linear mixed model fit by REML ['lmerMod']
## Formula: has_rt ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 41278
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.2347 -0.7363 -0.6001 1.3292 2.0940
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.003461 0.05883
## target_label (Intercept) 0.005187 0.07202
## dataset_name (Intercept) 0.005659 0.07522
## Residual 0.209422 0.45763
## Number of obs: 31863, groups:
## administration_id, 3572; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.269222 0.018477 14.571
## typeprev_as_target -0.012363 0.009318 -1.327
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_tr -0.036## Linear mixed model fit by REML ['lmerMod']
## Formula: rt ~ type + (1 | administration_id) + (1 | target_label) + (1 |
## dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 159119.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2220 -0.5479 -0.2549 0.2442 4.8650
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 66198 257.3
## target_label (Intercept) 13774 117.4
## dataset_name (Intercept) 19992 141.4
## Residual 299980 547.7
## Number of obs: 10195, groups:
## administration_id, 2850; target_label, 136; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 1096.706 36.148 30.34
## typeprev_as_target 7.473 19.679 0.38
##
## Correlation of Fixed Effects:
## (Intr)
## typprv_s_tr -0.032-2% difference in accuracy when repeating as the target
## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_early ~ type * trial_order + (1 | administration_id) +
## (1 | target_label) + (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 31294.3
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5469 -1.0026 -0.1098 1.0584 1.5488
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.009786 0.09892
## dataset_name (Intercept) 0.001385 0.03721
## Residual 0.214991 0.46367
## Number of obs: 23908, groups:
## administration_id, 3103; target_label, 141; dataset_name, 21
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.5139209 0.0158203 32.485
## typeprev_as_target 0.0169112 0.0315571 0.536
## trial_order 0.0006782 0.0003948 1.718
## typeprev_as_target:trial_order -0.0006215 0.0013023 -0.477
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_tr -0.057
## trial_order -0.228 0.109
## typprv_s_:_ 0.073 -0.926 -0.265
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_late ~ type * trial_order + (1 | administration_id) +
## (1 | target_label) + (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 37148.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5987 -0.9651 -0.5684 1.0231 1.6012
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.0000
## target_label (Intercept) 0.010398 0.1020
## dataset_name (Intercept) 0.005198 0.0721
## Residual 0.237079 0.4869
## Number of obs: 26394, groups:
## administration_id, 3480; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.4962076 0.0201796 24.590
## typeprev_as_target 0.0230007 0.0273367 0.841
## trial_order 0.0008206 0.0004108 1.998
## typeprev_as_target:trial_order -0.0009705 0.0011875 -0.817
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_tr -0.050
## trial_order -0.170 0.103
## typprv_s_:_ 0.068 -0.907 -0.283
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: accuracy ~ type * trial_order + (1 | administration_id) + (1 |
## target_label) + (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 10815.2
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.1258 -0.6296 0.1593 0.7656 2.4373
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.007424 0.08616
## target_label (Intercept) 0.004866 0.06976
## dataset_name (Intercept) 0.005339 0.07307
## Residual 0.076481 0.27655
## Number of obs: 31777, groups:
## administration_id, 3572; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.6707926 0.0173951 38.562
## typeprev_as_target -0.0014464 0.0145883 -0.099
## trial_order 0.0001887 0.0002220 0.850
## typeprev_as_target:trial_order -0.0013054 0.0006393 -2.042
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_tr -0.032
## trial_order -0.111 0.102
## typprv_s_:_ 0.043 -0.907 -0.281## Linear mixed model fit by REML ['lmerMod']
## Formula: has_rt ~ type * trial_order + (1 | administration_id) + (1 |
## target_label) + (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 41292.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.2331 -0.7367 -0.5992 1.3271 2.0917
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.003450 0.05874
## target_label (Intercept) 0.005100 0.07142
## dataset_name (Intercept) 0.005887 0.07673
## Residual 0.209367 0.45757
## Number of obs: 31863, groups:
## administration_id, 3572; target_label, 147; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.2793934 0.0189482 14.745
## typeprev_as_target -0.0197289 0.0237756 -0.830
## trial_order -0.0012290 0.0003571 -3.442
## typeprev_as_target:trial_order 0.0009560 0.0010388 0.920
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_tr -0.045
## trial_order -0.158 0.102
## typprv_s_:_ 0.061 -0.906 -0.281## Linear mixed model fit by REML ['lmerMod']
## Formula:
## rt ~ type * trial_order + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_target_mod
##
## REML criterion at convergence: 159109.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2217 -0.5471 -0.2571 0.2500 4.8582
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 66396 257.7
## target_label (Intercept) 14055 118.6
## dataset_name (Intercept) 20871 144.5
## Residual 299691 547.4
## Number of obs: 10195, groups:
## administration_id, 2850; target_label, 136; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 1092.6406 37.2792 29.310
## typeprev_as_target -97.0160 51.7692 -1.874
## trial_order 0.6107 0.7739 0.789
## typeprev_as_target:trial_order 4.4826 2.2681 1.976
##
## Correlation of Fixed Effects:
## (Intr) typp__ trl_rd
## typprv_s_tr -0.040
## trial_order -0.171 0.106
## typprv_s_:_ 0.058 -0.912 -0.280no substantial differences
compare all three
## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_early ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 42669.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5622 -1.0159 -0.0995 1.0523 1.7398
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.00000 0.00000
## target_label (Intercept) 0.01167 0.10801
## dataset_name (Intercept) 0.00104 0.03225
## Residual 0.21352 0.46208
## Number of obs: 32807, groups:
## administration_id, 3146; target_label, 146; dataset_name, 21
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.499283 0.014673 34.027
## typeprev_as_distractor 0.011730 0.006257 1.875
## typeprev_as_target 0.005166 0.010895 0.474
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d
## typprv_s_ds -0.131
## typprv_s_tr -0.067 0.111
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_late ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 50617.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5942 -0.9640 -0.5835 1.0132 1.6125
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.011802 0.10864
## dataset_name (Intercept) 0.003945 0.06281
## Residual 0.237133 0.48696
## Number of obs: 36004, groups:
## administration_id, 3524; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.485544 0.018259 26.593
## typeprev_as_distractor 0.010141 0.006412 1.582
## typeprev_as_target 0.003285 0.010574 0.311
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d
## typprv_s_ds -0.105
## typprv_s_tr -0.047 0.126
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: accuracy ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 15249.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.2174 -0.6275 0.1669 0.7575 2.4308
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.008449 0.09192
## target_label (Intercept) 0.004758 0.06898
## dataset_name (Intercept) 0.005881 0.07669
## Residual 0.077325 0.27807
## Number of obs: 43587, groups:
## administration_id, 3579; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.663159 0.017688 37.493
## typeprev_as_distractor 0.008137 0.003401 2.393
## typeprev_as_target -0.029233 0.005692 -5.136
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d
## typprv_s_ds -0.058
## typprv_s_tr -0.027 0.108## Linear mixed model fit by REML ['lmerMod']
## Formula: has_rt ~ type + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 56429.5
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.2640 -0.7236 -0.5877 1.3360 2.0829
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.003325 0.05767
## target_label (Intercept) 0.006179 0.07861
## dataset_name (Intercept) 0.005567 0.07461
## Residual 0.208702 0.45684
## Number of obs: 43718, groups:
## administration_id, 3579; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.284675 0.018274 15.578
## typeprev_as_distractor -0.012152 0.005482 -2.217
## typeprev_as_target -0.009256 0.009233 -1.002
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d
## typprv_s_ds -0.090
## typprv_s_tr -0.040 0.119## Linear mixed model fit by REML ['lmerMod']
## Formula: rt ~ type + (1 | administration_id) + (1 | target_label) + (1 |
## dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 215865.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4186 -0.5408 -0.2536 0.2321 5.0795
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 66202 257.3
## target_label (Intercept) 10582 102.9
## dataset_name (Intercept) 21807 147.7
## Residual 303350 550.8
## Number of obs: 13830, groups:
## administration_id, 3146; target_label, 143; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 1101.75 35.81 30.763
## typeprev_as_distractor -15.15 12.39 -1.223
## typeprev_as_target 10.12 19.60 0.516
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d
## typprv_s_ds -0.109
## typprv_s_tr -0.038 0.109-3% difference in accuracy when repeating as the target
## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_early ~ type * trial_order + (1 | administration_id) +
## (1 | target_label) + (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 42702.6
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.56823 -1.02073 -0.09851 1.05215 1.71887
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.011541 0.10743
## dataset_name (Intercept) 0.001068 0.03268
## Residual 0.213509 0.46207
## Number of obs: 32807, groups:
## administration_id, 3146; target_label, 146; dataset_name, 21
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.4957309 0.0151438 32.735
## typeprev_as_distractor -0.0059683 0.0136462 -0.437
## typeprev_as_target 0.0101568 0.0313026 0.324
## trial_order 0.0005131 0.0003857 1.330
## typeprev_as_distractor:trial_order 0.0007898 0.0007595 1.040
## typeprev_as_target:trial_order -0.0004517 0.0012905 -0.350
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d typprv_s_t trl_rd typprv_s_d:_
## typprv_s_ds -0.145
## typprv_s_tr -0.060 0.051
## trial_order -0.239 0.190 0.106
## typprv_s_d:_ 0.145 -0.865 -0.053 -0.431
## typprv_s_t:_ 0.076 -0.073 -0.926 -0.264 0.130
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: correct_late ~ type * trial_order + (1 | administration_id) +
## (1 | target_label) + (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 50645.7
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5929 -0.9637 -0.5825 1.0159 1.6131
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.000000 0.00000
## target_label (Intercept) 0.011683 0.10809
## dataset_name (Intercept) 0.003822 0.06183
## Residual 0.237096 0.48693
## Number of obs: 36004, groups:
## administration_id, 3524; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.4834960 0.0184560 26.197
## typeprev_as_distractor -0.0201649 0.0136926 -1.473
## typeprev_as_target 0.0084498 0.0270039 0.313
## trial_order 0.0005360 0.0004025 1.332
## typeprev_as_distractor:trial_order 0.0015333 0.0007724 1.985
## typeprev_as_target:trial_order -0.0004896 0.0011753 -0.417
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d typprv_s_t trl_rd typprv_s_d:_
## typprv_s_ds -0.129
## typprv_s_tr -0.056 0.085
## trial_order -0.193 0.188 0.103
## typprv_s_d:_ 0.130 -0.859 -0.077 -0.437
## typprv_s_t:_ 0.074 -0.105 -0.906 -0.285 0.164
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')## Linear mixed model fit by REML ['lmerMod']
## Formula: accuracy ~ type * trial_order + (1 | administration_id) + (1 |
## target_label) + (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 15281.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.2128 -0.6265 0.1669 0.7577 2.4236
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.008398 0.09164
## target_label (Intercept) 0.004792 0.06922
## dataset_name (Intercept) 0.005820 0.07629
## Residual 0.077332 0.27809
## Number of obs: 43587, groups:
## administration_id, 3579; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 6.646e-01 1.775e-02 37.447
## typeprev_as_distractor -7.998e-03 7.276e-03 -1.099
## typeprev_as_target -5.449e-03 1.449e-02 -0.376
## trial_order -3.592e-05 2.191e-04 -0.164
## typeprev_as_distractor:trial_order 9.878e-04 4.136e-04 2.389
## typeprev_as_target:trial_order -1.072e-03 6.358e-04 -1.685
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d typprv_s_t trl_rd typprv_s_d:_
## typprv_s_ds -0.073
## typprv_s_tr -0.032 0.079
## trial_order -0.114 0.173 0.101
## typprv_s_d:_ 0.076 -0.859 -0.073 -0.423
## typprv_s_t:_ 0.043 -0.096 -0.905 -0.282 0.154## Linear mixed model fit by REML ['lmerMod']
## Formula: has_rt ~ type * trial_order + (1 | administration_id) + (1 |
## target_label) + (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 56457.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.2634 -0.7234 -0.5876 1.3353 2.0800
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 0.003318 0.05760
## target_label (Intercept) 0.006109 0.07816
## dataset_name (Intercept) 0.005601 0.07484
## Residual 0.208672 0.45681
## Number of obs: 43718, groups:
## administration_id, 3579; target_label, 151; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 0.2920061 0.0185783 15.718
## typeprev_as_distractor -0.0002779 0.0117322 -0.024
## typeprev_as_target -0.0149148 0.0234686 -0.636
## trial_order -0.0009940 0.0003501 -2.839
## typeprev_as_distractor:trial_order -0.0002415 0.0006654 -0.363
## typeprev_as_target:trial_order 0.0007599 0.0010268 0.740
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d typprv_s_t trl_rd typprv_s_d:_
## typprv_s_ds -0.112
## typprv_s_tr -0.047 0.083
## trial_order -0.168 0.183 0.102
## typprv_s_d:_ 0.114 -0.859 -0.075 -0.434
## typprv_s_t:_ 0.063 -0.102 -0.905 -0.282 0.161## Linear mixed model fit by REML ['lmerMod']
## Formula:
## rt ~ type * trial_order + (1 | administration_id) + (1 | target_label) +
## (1 | dataset_name)
## Data: prev_any_mod
##
## REML criterion at convergence: 215848.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4295 -0.5414 -0.2545 0.2342 5.0634
##
## Random effects:
## Groups Name Variance Std.Dev.
## administration_id (Intercept) 66021 256.9
## target_label (Intercept) 11101 105.4
## dataset_name (Intercept) 22140 148.8
## Residual 303193 550.6
## Number of obs: 13830, groups:
## administration_id, 3146; target_label, 143; dataset_name, 24
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 1092.8663 36.8082 29.691
## typeprev_as_distractor 36.8202 27.2081 1.353
## typeprev_as_target -95.1504 51.2296 -1.857
## trial_order 0.5556 0.7613 0.730
## typeprev_as_distractor:trial_order -3.3319 1.5145 -2.200
## typeprev_as_target:trial_order 4.5276 2.2466 2.015
##
## Correlation of Fixed Effects:
## (Intr) typprv_s_d typprv_s_t trl_rd typprv_s_d:_
## typprv_s_ds -0.139
## typprv_s_tr -0.044 0.082
## trial_order -0.182 0.169 0.103
## typprv_s_d:_ 0.137 -0.866 -0.075 -0.412
## typprv_s_t:_ 0.063 -0.098 -0.910 -0.280 0.153no substantial differences, even wipes out the -3% we were seeing without trial effects
Summary
The effects of having a target image that has been seen 1 time previously (as distractor or target), versus a target image that is being seen for the first time are (if they exist) have central estimates around 1% of the time more or less looking (both for windows around 0 and actual accuracy). These effects may well interact with trial order, although again, effects and interactions, if they exist, are small.