/
data_processing.Rmd
executable file
·277 lines (191 loc) · 14.1 KB
/
data_processing.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
---
title: "R Notebook"
output: html_notebook
---
```{r message=FALSE, warning=FALSE, paged.print=FALSE}
# PREAMBLE
# This section loads the required packages for this notebook.
# All the required packages are already installed in the container
library(tidyverse)
library(magrittr)
library(glue)
```
# Introduction
This *R Markdown* notebook contains all the code for doing the data cleaning and data collation steps for the Fischer et al RRR project.
To use this notebook simply run each of the code chucks (either by running each chunk individually or by Run All from the run dropdown menu).
## Participant exclusions
```{r}
# Define some parameters
CATCH.TRIAL.REJECT = 5
```
Participants are exluded is their error rate is more than `r CATCH.TRIAL.REJECT`%. Participants are also excluded if they did not complete all the measures, or if the exit questionnaire determined that they were not naive to the purpose of the experiment.
```{r}
# Perform the exclusions
dataFolder = here::here("data/")
generated.dataFolder = here::here("data/processed_data/")
# Load the data
load(file.path(generated.dataFolder,"TidyFileList.sav")) # the file list contains experimenter determined exclusions. For example, crashing computers, technical errors etc
#load("generated_data/TidyFileList.sav")
Nodes = names(FileList)
for (node in Nodes) {
load(glue('{generated.dataFolder}{node}-ResponseData.dat'))
load(glue('{generated.dataFolder}{node}-SecondaryData.dat'))
load(glue('{generated.dataFolder}{node}-EyeTrackerData.dat'))
map_df(names(ResponseData),
function(x) tibble(`Subject Code` = x,
`Catch Trial Reject` = ((1- mean(ResponseData[[x]] %>% pull(correct))) * 100) > CATCH.TRIAL.REJECT)) %>%
full_join(FileList[[node]], by = "Subject Code") -> FileList[[node]]
map_df(SecondaryData %>% as.list(),
function(x) tibble(`Subject Code` = x$code, `Guess` = ifelse(x$guess == "N", FALSE,TRUE))) %>%
full_join(FileList[[node]], by = "Subject Code") -> FileList[[node]]
}
rm(list = c("node","Nodes")) # tidying up!
rm(list = c("ResponseData","SecondaryData","EyeTrackerData")) # tidying up!
cat("The exclusions have been marked.")
```
```{r}
# Make a table of the exclude counts
makeExcludeTable<-function(x){
# This peforms the exclusions in a specific order. This is to prevent double counting (for example, a participant that both guessed the purpose of the experiment and had a high error rate wouldn't be counted as two)
all.Counts = x %>% summarise(`Exclude Data` = sum(`Exclude Data`), `Catch Trial Reject` = sum(`Catch Trial Reject`), Guess = sum(Guess), total = n())
TotalSample = dim(x)[1]
# Remove excluded
x %>% filter(`Exclude Data` == TRUE) %>% dim() %>% t() %>% as.tibble() %>% pull(V1) -> Step1.r
x %>% filter(`Exclude Data` == FALSE) -> x
Step1.n = dim(x)[1]
# Remove catch trial errors
x %>% filter(`Catch Trial Reject` == TRUE) %>% dim() %>% t() %>% as.tibble() %>% pull(V1) -> Step2.r
x %>% filter(`Catch Trial Reject` == FALSE) -> x
Step2.n = dim(x)[1]
# Remove participants that have guessed
x %>% filter(Guess == TRUE) %>% dim() %>% t() %>% as.tibble() %>% pull(V1) -> Step3.r
x %>% filter(Guess == FALSE) -> x
Step3.n = dim(x)[1]
return.df = tibble(
node = unique(x$Node),
`Total sample` = all.Counts$total,
`After exclusions` = Step3.n,
`Other exclusions` = Step1.r,
`Error rate exclusions` = Step2.r,
`Guessed purpose` = Step3.r
)
return(return.df)
}
map_df(FileList, function (x) makeExcludeTable(x)) -> Exclusions.Table
# I'll use the exclusions table later, so I'll save it for now
ReplicationProjectTools::GetLabNames()
Exclusions.Table %>% rename(Lab = node, Total = `Total sample`, Analysed = `After exclusions`, Other = `Other exclusions`, Errors = `Error rate exclusions`, Purpose = `Guessed purpose`) %>% inner_join(tibble(Lab = Nodes, Names = LabNames), by = "Lab") %>% mutate(Lab = Names) %>% select(-Names) %>% arrange(Lab) %>% readr::write_csv(path = file.path(generated.dataFolder,"ExclusionsTable.csv"))
```
```{r}
generated.dataFolder = here::here("data/processed_data/")
map_df(Nodes, function(x) {load(file = glue("{generated.dataFolder}/{x}-HasEyeTracker.dat"))
tibble(Node = x, Code = HasEyeTracker %>% as.list() %@% "names", HasEye = HasEyeTracker %>% as.list() %>% unlist())}) %>% write_csv(path = here::here("data/processed_data/EyeTrackerTable.csv"))
map(Nodes,function(x) {load(file = glue("{generated.dataFolder}/{x}-ResponseData.dat"))
ResponseData %>% as.list() %>% Reduce(f = function(x,y) rbind(x,y))}) %>% Reduce(f = function(x,y) rbind(x,y)) -> AllResponseData
map(Nodes,function(x) {load(file = glue("{generated.dataFolder}/{x}-EyeTrackerData.dat"))
EyeTrackerData %>% as.list() %>% Reduce(f = function(x,y) rbind(x,y))}) %>% Reduce(f = function(x,y) rbind(x,y)) -> AllEyeTrackerData
CombinedData = merge.data.frame(AllResponseData, AllEyeTrackerData, by = c("node","code","trial"))
rm(list = c("AllResponseData","AllEyeTrackerData"))
map(Nodes,function(x) {load(file = glue("{generated.dataFolder}/{x}-FingerData.dat"))
FingerData %>% as.list() %>% Reduce(f = function(x,y) rbind(x,y))}) %>% Reduce(f = function(x,y) rbind(x,y)) -> AllFingerData
map(Nodes,function(x) {load(file = glue("{generated.dataFolder}/{x}-SecondaryData.dat"))
SecondaryData %>% as.list() %>% Reduce(f = function(x,y) rbind(x,y))}) %>% Reduce(f = function(x,y) rbind(x,y)) -> AllSecondaryData
merge.data.frame(AllSecondaryData, AllFingerData, by = c("node","code")) %>%
mutate(fingers = paste0(substr(fc1,1,1),substr(fc2,1,1),substr(fc3,1,1),substr(fc4,1,1))) %>% # count left and right finger uses
mutate(Lef = str_count(fingers,"L")) %>% mutate(Rig = str_count(fingers,"R")) %>%
rename(Lab = node, Code = code) %>% as.tibble() -> Moderators
Moderators %>% mutate(FingerGroup = # Add the finger counting group. Change this is you want to change how the groups are determined
case_when(Rig >= 3 & (Rig + Lef >= 3) ~ "RS",
Lef >= 3 & (Rig + Lef >= 3) ~ "LS",
Rig == 2 & (Rig + Lef == 2) ~ "RA",
Lef == 2 & (Rig + Lef == 2) ~ "LA",
Rig == 2 & (Rig + Lef == 3) ~ "RA",
Lef == 2 & (Rig + Lef == 3) ~ "LA",
TRUE ~ "NG")) %>% # Add the handedness group
mutate(Hand = ifelse(handedness < 0, "LH","RH")) %>% # Add the reading direction group
mutate(ReadingDir = ifelse(language == 1, "LTR","NLR")) %>% # trim it down
select(Lab,Code,FingerGroup,Hand,ReadingDir,mathTest,amas) %>% # do some tidying up
rename(MathTest = mathTest, AMAS = amas) -> Moderators
Reduce(FileList, f = function(x,y) rbind(x,y)) -> FileList
names(FileList)[1] = "Code"
names(CombinedData)[2] <- "Code"
FileList %>% full_join(Moderators, by = "Code") %>% full_join(CombinedData, by = "Code") -> AllData
AllData %>% write_csv(here::here("data/final_data_csv/AllData.csv"))
# Now process the different data partitions
# This isn't the easiest for creating arbitrary partitions
# The code below would probably need to be edited.
# In a later version I might change this to make it more user friendly
GUESS.FILTER = c(FALSE,FALSE,TRUE) # Keep participants that correctly guessed purpose? FALSE = drop them; TRUE = select only them
REJECT.FILTER = c(0,1,0) # Drop eye-movement contaminated trials? 0 = drop them; 1 = select only them
CSV.NAME = c(here::here("data/final_data_csv/AllFischerData.csv"),
here::here("data/final_data_csv/ContaminatedFischerData.csv"),
here::here("data/final_data_csv/RejectedFischerData.csv")) # where to save the file?
pmap(list(GUESS.FILTER = GUESS.FILTER, REJECT.FILTER = REJECT.FILTER, CSV.NAME = CSV.NAME),
function(GUESS.FILTER, REJECT.FILTER,CSV.NAME)
AllData %>% filter(Guess == GUESS.FILTER, `Catch Trial Reject` == FALSE, `Exclude Data` == FALSE) %>% filter(correct == 1, Reject == REJECT.FILTER) %>% select(Lab,Code,delayDur,targetLocation,cue,RT,FingerGroup,Hand,ReadingDir,MathTest,AMAS) %>% mutate(Magnitude = ifelse(cue < 5, "low","high")) %>% filter(targetLocation != "c") %>%
mutate(RT = RT * 1000) %>% group_by(Code,Magnitude,delayDur,targetLocation) %>%
summarise(RT = mean(RT),
Lab = unique(Lab),
FingerGroup = unique(FingerGroup),
Hand = unique(Hand),
ReadingDir = unique(ReadingDir),
MathTest = unique(MathTest),
AMAS = unique(AMAS)) %>%
mutate(Congruency = case_when(
targetLocation == "l" & Magnitude == "low" ~ "Con", # left targets low number = congurent
targetLocation == "l" & Magnitude == "high" ~ "Inc", # left targets and high numbers = incongruent
targetLocation == "r" & Magnitude == "low" ~ "Inc", # right targets low number = incongruent
targetLocation == "r" & Magnitude == "high" ~ "Con")) %>%
group_by(Code,delayDur,Congruency) %>% summarise(Lab = unique(Lab),
FingerGroup = unique(FingerGroup),
Hand = unique(Hand),
ReadingDir = unique(ReadingDir),
MathTest = unique(MathTest),
AMAS = unique(AMAS), RT = mean(RT)) %>% spread(Congruency,RT) %>% mutate(RT = Inc - Con) %>% select(-Con,-Inc) %>% spread(delayDur,RT) %>% drop_na() %>% rename(d250 = `250`, d500 = `500`, d750 = `750`, d1000 = `1000`) %>% select(Lab,Code,d250,d500,d750,d1000,FingerGroup,Hand,ReadingDir,MathTest,AMAS)) -> DataPartitions
DataPartitions[[1]] %>% arrange(Lab,Code) %>% write_csv(path = CSV.NAME[1])
DataPartitions[[2]] %>% arrange(Lab) %>% select(-FingerGroup,-Hand,-ReadingDir,-MathTest,-AMAS) %>% write_csv(path = CSV.NAME[2])
height<-function(x){dim(x)[1]}
DataPartitions[[3]] %>% arrange(Lab) %>% ungroup() %>% group_by(Lab) %>% nest() %>% mutate(Ss = map_int(data,height)) %>% filter(Ss > 4) %>% unnest()%>% arrange(Lab,Code) %>% select(-Ss) %>% write_csv(path = CSV.NAME[3]) # make sure to only select labs with more than 5 participants
map(Nodes,function(x) {load(file = glue("{generated.dataFolder}/{x}-HasEyeTracker.dat"))
tibble(Code = names(HasEyeTracker), HasEye = unlist(HasEyeTracker %>% as.list()))}) %>% Reduce(f = function(x,y) rbind(x,y)) -> AllHasEyeTracker
#AllData %>% filter(Guess == FALSE, `Catch Trial Reject` == FALSE, `Exclude Data` == FALSE) %>% #full_join(AllHasEyeTracker, by = "Code") %>% filter(HasEye == T) %>% mutate(Magnitude = ifelse(cue < 5, #"low","high")) %>% filter(targetLocation != "c", correct == 1) %>% mutate(Congruency = case_when(
# targetLocation == "l" & Magnitude == "low" ~ "Con", # left targets low number = congurent
# targetLocation == "l" & Magnitude == "high" ~ "Inc", # left targets and high numbers = #incongruent
# targetLocation == "r" & Magnitude == "low" ~ "Inc", # right targets low number = incongruent
# targetLocation == "r" & Magnitude == "high" ~ "Con")) %>% group_by(Lab,Code,Congruency,delayDur) %>% #summarise(Rejected = mean(Reject) * 100) %>% ungroup() %>% group_by(Lab,Congruency,delayDur) %>% #summarise(M = mean(Rejected), SD = sd(Rejected)) %>% inner_join(tibble(Lab = Nodes, Name = LabNames), by = #"Lab") %>% ungroup() %>% select(Name,Congruency,delayDur,M,SD) %>% mutate(val = glue_data(list(M = M,SD = #SD), "{papaja::printnum(M)} ({papaja::printnum(SD)})")) %>% select(-M,-SD) %>% spread(delayDur,val) %>% #mutate(Congruency = case_when(Congruency == "Con" ~ "congruent", Congruency == "Inc" ~ "incongruent")) %>% #write_csv("generated_data/EyeTrackerDetail.csv")
```
```{r}
require(papaja)
ReplicationProjectTools::GetLabNames()
LabCodes = tibble(Name = LabNames, Lab = Nodes)
AllData %>% filter(targetLocation != "c",
correct == 1,
`Exclude Data` == FALSE,
`Catch Trial Reject` == FALSE,
Guess == FALSE) %>%
mutate(Magnitude = ifelse(cue < 5, "low","high")) %>%
mutate(Congruency = case_when(
targetLocation == "l" & Magnitude == "low" ~ "Con",
targetLocation == "l" & Magnitude == "high" ~ "Inc",
targetLocation == "r" & Magnitude == "low" ~ "Inc",
targetLocation == "r" & Magnitude == "high" ~ "Con")) %>%
group_by(Lab,Code,Congruency,delayDur) %>% summarise(Reject = sum(Reject)) %>%
filter(Reject > 0) %>% select(Lab,Code,Congruency,delayDur,Reject) -> Eye.tmp
# Trials in Analysis
Eye.tmp %>% unite("ISI.Con", c("Congruency","delayDur")) %>% spread(ISI.Con,Reject) %>% drop_na() %>% gather(ISI.Con,Reject,-Lab,-Code) %>% group_by(Lab,ISI.Con) %>% summarise(Reject = sum(Reject), N = unique(Code) %>% length()) %>%
separate(ISI.Con, c("Congruency","delayDur")) %>% ungroup() -> TrialsInAnalysis.Model1c #%>% spread(delayDur,Reject)
# Total Trials
Eye.tmp %>% unite("ISI.Con", c("Congruency","delayDur")) %>% group_by(Lab,ISI.Con) %>% summarise(Reject = sum(Reject)) %>%
separate(ISI.Con,c("Congruency","delayDur")) %>% ungroup() -> TotalTrails.Model1c #%>% spread(delayDur,Reject)
AllData %>% filter(maxHorzDeviation > 0) %>% group_by(Lab) %>% summarise(Total = unique(Code) %>% length()) -> Total.Ns
merge.data.frame(TrialsInAnalysis.Model1c,TotalTrails.Model1c, by = c("Lab","Congruency","delayDur"), suffixes = c(".Analysed",".Total")) %>% mutate(Value = paste0(Reject.Analysed," (",Reject.Total,")")) %>% select(Lab,N,Congruency,delayDur,Value) %>% spread(delayDur,Value) %>% inner_join(LabCodes) %>% inner_join(Total.Ns) %>%
select(Name,Total,N,Congruency,`250`,`500`,`750`,`1000`) %>%
arrange(Name) -> Eye.TrialNumbers
names(Eye.TrialNumbers) <- c("Lab","Subjects","Analysed","Trial Type",
"250 ms","500 ms","750 ms", "1000 ms")
Eye.TrialNumbers %>% mutate(`Trial Type` =
case_when(`Trial Type` == "Con" ~ "Congruent",
`Trial Type` == "Inc" ~ "Incongruent")) %>%
write_csv(here::here("data/processed_data/EyeTrackerDetail.csv"))
```
You can load the [Analysis notebook](./processing/analysis.Rmd).