-
Notifications
You must be signed in to change notification settings - Fork 8
/
resubmit_exclusions.Rmd
463 lines (361 loc) · 15.5 KB
/
resubmit_exclusions.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
---
title: "Resubmit Exclusions"
output: pdf_document
---
```{r include=F}
# will need the private_metadata from the authors
library(plyr)
library(here)
library(tidyverse)
# read in the private metat data
d<-read.csv(here::here("private_metadata.csv"))
### same code as in pre-experiment to make unique IDs ###
# remove the - sign. It causes problems later
d$corpus<-revalue(d$corpus, c("Casillas-Yeli"="CasillasYeli"))
#add a column called with a unique id (the original child id pasted together with the corpus)
d <- d %>% mutate(unique_id = paste0(child_ID, corpus))
#get rid of data from the warlaumont corpus
d <- d %>% filter(corpus != c("Warlaumont"))
#make a new column with the language
d <- d %>% mutate(language = ifelse(corpus == "Seedlings",
"English",
"Non-English"))
#make a categorical age variable based on 07, 8-18, and 19-36 groupings
d <- d %>% mutate(age_groups = ifelse(age_mo_round <= 7,
"0-7",
ifelse(age_mo_round <= 18,
"8-18",
"19-36")))
#make grouping variable for age-language question
d <- d %>% mutate(age_language = paste0(age_groups, language))
#keep only canonical and non-canonical
d <- d %>% filter(Answer %in% c("Canonical", "Non-canonical"))
#### end copy of pre-experiment code #####
metadata<-d
```
This document responds to some questions raised by reviewers during the review process. You will need to get the private metadata csv file from the BabbleCor authors (and our data) in order to reproduce these results.
The reviewers expressed some confusion as to why we used the exclusion criteria that we used. Our main purpose was to create a homogeneous sample of participants in order to remove potential confounds. However, at the request of the reviewers, we include some analyses here that don't exclude based on the sociodemographic information collected.
Firstly, we can run the analyses again without any exclusion criteria. The full results are below for the reader to examine but the overall conclusions are the same as those with exclusions reported in the paper.
```{r}
##### TRIAL DATA #####
# get data before exclusions
library(here)
d <- read.csv(here::here("data","trial_data.csv"))
library(ggplot2)
# refactor the group so the order is correct in the graphs
d$stim_ageGroup<-factor(d$stim_ageGroup,levels = c("0-7","8-18","19-36"))
agg<-aggregate(correct~phase*subject_ID, data = d, mean)
t.test(x = subset(agg,agg$phase=="Age")$correct, mu = 1/3)
t.test(x = subset(agg,agg$phase=="Language")$correct, mu = 1/3)
t.test(x = subset(agg,agg$phase=="Sex")$correct, mu = 1/3)
###### Testing the Age data #######
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
###### Testing the Language data #######
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
###### Testing the Sex data #######
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
```
Next we ran the same analyses but excluding those who failed our attention check, audio check, or gave negligent responses (gave the same response for an entire phase of trials). Again, the full results are below but the conclusions are the same as before.
```{r}
##### TRIAL DATA #####
# get data before exclusions
library(here)
d <- read.csv(here::here("data","trial_data.csv"))
library(ggplot2)
sum_data <- read.csv(here::here("data","summarized_data.csv"))
attention_exc<-unique(subset(sum_data,sum_data$n_attention_checks<=5)$subject_ID)
audio_exc<-unique(subset(sum_data,sum_data$n_audio_checks<=5)$subject_ID)
neglig_exc<-unique(subset(sum_data,sum_data$var_sex<=5 & sum_data$var_lang<=5 & sum_data$var_age<=5)$subject_ID)
# refactor the group so the order is correct in the graphs
d$stim_ageGroup<-factor(d$stim_ageGroup,levels = c("0-7","8-18","19-36"))
# exclude those who failed our attention check and audio check criteria
d <- d[d$subject_ID%in%attention_exc, ]
d <- d[d$subject_ID%in%audio_exc, ]
# exclude those who clicked the same button for an entire experimental phase
d <- d[d$subject_ID%in%neglig_exc, ]
# d <- d[d$gender %in% c("Female", "Male"), ]
# d <- d[d$country %in% c("Canada", "USA"), ]
# d <- d[d$eng_first == "Yes", ]
# d <- d[d$know_corp_lang == "None", ]
agg<-aggregate(correct~phase*subject_ID, data = d, mean)
t.test(x = subset(agg,agg$phase=="Age")$correct, mu = 1/3)
t.test(x = subset(agg,agg$phase=="Language")$correct, mu = 1/3)
t.test(x = subset(agg,agg$phase=="Sex")$correct, mu = 1/3)
###### Testing the Age data #######
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Age"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
###### Testing the Language data #######
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
###### Testing the Sex data #######
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
```
Finally, we ran the some analyses but excluding based on the sociodemographic information when it made sense to do so. Once again, the full results are below but the conclusions are the same as before.
Here is the suggested analysis of the sex data with only the gender based exclusion criteria:
```{r}
##### TRIAL DATA #####
# get data before exclusions
library(here)
d <- read.csv(here::here("data","trial_data.csv"))
library(ggplot2)
# refactor the group so the order is correct in the graphs
d$stim_ageGroup<-factor(d$stim_ageGroup,levels = c("0-7","8-18","19-36"))
sum_data <- read.csv(here::here("data","summarized_data.csv"))
attention_exc<-unique(subset(sum_data,sum_data$n_attention_checks<=5)$subject_ID)
audio_exc<-unique(subset(sum_data,sum_data$n_audio_checks<=5)$subject_ID)
neglig_exc<-unique(subset(sum_data,sum_data$var_sex<=5 & sum_data$var_lang<=5 & sum_data$var_age<=5)$subject_ID)
# refactor the group so the order is correct in the graphs
d$stim_ageGroup<-factor(d$stim_ageGroup,levels = c("0-7","8-18","19-36"))
# exclude those who failed our attention check and audio check criteria
d <- d[d$subject_ID%in%attention_exc, ]
d <- d[d$subject_ID%in%audio_exc, ]
# exclude those who clicked the same button for an entire experimental phase
d <- d[d$subject_ID%in%neglig_exc, ]
d <- d[d$gender %in% c("Female", "Male"), ]
# d <- d[d$country %in% c("Canada", "USA"), ]
# d <- d[d$eng_first == "Yes", ]
# d <- d[d$know_corp_lang == "None", ]
agg<-aggregate(correct~phase*subject_ID, data = d, mean)
#t.test(x = subset(agg,agg$phase=="Age")$correct, mu = 1/3)
#t.test(x = subset(agg,agg$phase=="Language")$correct, mu = 1/3)
t.test(x = subset(agg,agg$phase=="Sex")$correct, mu = 1/3)
# testing the sex data
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Sex"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
```
We also ran the analyses on the language data while only excluding those who knew any of the corpus languages with the same results as before:
```{r}
##### TRIAL DATA #####
# get data before exclusions
library(here)
d <- read.csv(here::here("data","trial_data.csv"))
library(ggplot2)
# refactor the group so the order is correct in the graphs
d$stim_ageGroup<-factor(d$stim_ageGroup,levels = c("0-7","8-18","19-36"))
sum_data <- read.csv(here::here("data","summarized_data.csv"))
attention_exc<-unique(subset(sum_data,sum_data$n_attention_checks<=5)$subject_ID)
audio_exc<-unique(subset(sum_data,sum_data$n_audio_checks<=5)$subject_ID)
neglig_exc<-unique(subset(sum_data,sum_data$var_sex<=5 & sum_data$var_lang<=5 & sum_data$var_age<=5)$subject_ID)
# refactor the group so the order is correct in the graphs
d$stim_ageGroup<-factor(d$stim_ageGroup,levels = c("0-7","8-18","19-36"))
# exclude those who failed our attention check and audio check criteria
d <- d[d$subject_ID%in%attention_exc, ]
d <- d[d$subject_ID%in%audio_exc, ]
# exclude those who clicked the same button for an entire experimental phase
d <- d[d$subject_ID%in%neglig_exc, ]
# d <- d[d$gender %in% c("Female", "Male"), ]
# d <- d[d$country %in% c("Canada", "USA"), ]
# d <- d[d$eng_first == "Yes", ]
d <- d[d$know_corp_lang == "list()", ]
agg<-aggregate(correct~phase*subject_ID, data = d, mean)
#t.test(x = subset(agg,agg$phase=="Age")$correct, mu = 1/3)
t.test(x = subset(agg,agg$phase=="Language")$correct, mu = 1/3)
#t.test(x = subset(agg,agg$phase=="Sex")$correct, mu = 1/3)
# testing the sex data
# Random intercept Model: Baseline
library(lme4)
model1<- glmer(correct~1+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model1)
# Adding level 1 predictors: childcare & caregiver
model2<- glmer(correct~1+childcare+caregiver+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model2)
# Compare two models
anova(model1,model2)
#additional model with only childcare
model4<- glmer(correct~1+childcare+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model4)
#compare to baseline model
anova(model1,model4)
# Adding level 2 predictor: Participant's gender
model3<- glmer(correct~1+childcare+caregiver+gender+(1|subject_ID),
data = subset(d,d$phase=="Language"),
family = binomial(link=logit))
summary(model3)
# Compare two models
anova(model2,model3)
```