1_MScProject_getASdata.Rmd

---
title: "Preparation: calculate and get AS in data collection file"
author: "MScStudent"
date: "20/08/2021"
output: html_document
---

```{r setup, include=FALSE}

#check working directory
knitr::opts_chunk$set(echo = TRUE)
getwd()
```

###Prepare the data files###

#1st Step: Turn requencies in German databases into Associative Strengths (AS)
*divide the frequency of the target-response pair in the database by the sum of all frequencies of one token*  

```{r}
# read in all the German files
#German databases
Assocs <- read.csv("Assocs.csv", sep=",", header=TRUE) # English
compound_NN_AMT <- read.csv("compound_NN.assoc-AMT.csv", sep="\t", header=FALSE, check.names=FALSE) # deutsch
nag.assoc_second <- read.csv("nag.assoc-second.csv", sep="\t", header=FALSE) # deutsch
nag.assoc_first <- read.table("nag.assoc-first.csv", quote = "", header=TRUE) # deutsch
for (index in (1:nrow(nag.assoc_first))){
  nag.assoc_first[index, 1] <- substr(nag.assoc_first[index, 1], start = 2, stop = nchar(nag.assoc_first[index, 1]))
  nag.assoc_first[index, 3] <- substr(nag.assoc_first[index, 3], start = 1, stop = nchar(nag.assoc_first[index, 3])-1)
}

#data collection file
Experiment_data_latestversion_german_csv <- read.csv("Experiment_data_latestversion_german_csv.csv", sep="\t", header=TRUE, check.names=FALSE, encoding="UTF-8") # clean
Experiment_data_latestversion_csv <- read.csv("Experiment_data_latestversion_csv.csv", sep="\t", header=TRUE, check.names=FALSE, encoding="UTF-8") # clean

#which database to use: targets&databases.csv
targets_databases <- read.csv("targets&databases.csv", sep=";", header=TRUE)
targets_databases <- read.table("targets&databases.csv", sep=";", header=TRUE)

#manual correction of Umlaut:
targets_databases[14, 1] <- "Rüstung"
targets_databases[20, 1] <- "Brücke"

#for German: get associative strength:
df_all_targets <- NULL
for (index in (1:nrow(targets_databases))){
  target_word <- targets_databases[index ,"target"]
  print(target_word)
  file <- targets_databases[index ,"database"]
  print(file)
  if (file == "NAG assoc 1st"){
    file_name <- nag.assoc_first
  }
  if (file == "compound.NN.assoc-AMT"){
    file_name <- compound_NN_AMT
  }
  if (file == "nag assoc second"){
    file_name <- nag.assoc_second
  }
  #make all counts in files numeric
  file_name[,3] <- as.integer(file_name[,3])
  
  #calculate forward associative strengths; first forward AS, then backward AS
  subset_forward <- file_name[file_name[, 1] == target_word, ]
  total_guesses <- sum(as.integer(subset_forward[, 3]))
  results_forward <- subset_forward[, 3]/total_guesses
  results_backward <- c()
  for (inner_index in (1:nrow(subset_forward))){
    new_target <- subset_forward[inner_index, 2]
    new_subset <- file_name[file_name[, 1] == new_target, ]
    backwards_count <- new_subset[new_subset[, 2]== target_word, 3]
    if (length(backwards_count) == 0){
      resuls_to_add <- NA
    } else {
      resuls_to_add <- backwards_count/sum(as.integer(new_subset[, 3]))
    }
    results_backward <- append(results_backward, resuls_to_add)
  }
  new_df <- cbind(subset_forward[,1:2], results_forward, results_backward)
  print(new_df)
  if (is.null(df_all_targets)){
    df_all_targets <- new_df
  } else{
    colnames(new_df) <- colnames(df_all_targets)
    df_all_targets <- rbind(df_all_targets, new_df)
  }
}

#save as csv file

write.csv(df_all_targets, file="df_all_targets_german.csv")
```



#2nd Step: Tranfer AS from databases into data collection files

*for the German data*

```{r}
#read in German database with the AS (instead of frequencies)
german_database <- read.csv('all_targets_german.csv', sep=",", header=TRUE) #Doc incl. corrections for Umlaut and FAS/BAS as col names

#sanity checks
str(german_database)
head(german_database)
summary(german_database)

#read in German data: data collection document
germandata <- read.csv('germandata.csv', sep=",", header=TRUE) #cleanded up without lemmatisation

#sanity checks
str(germandata)
head(germandata)
summary(germandata)
```

```{r}
#ensure that types (strings vs. numbers) are stored correctly

german_database$target <- as.factor(german_database$target)
german_database$cue <- as.factor(german_database$cue)
germandata$target <- as.factor(germandata$target)


german_database$FAS <- as.numeric(german_database$FAS)
german_database$BAS <- as.numeric(german_database$BAS)
germandata$fas_tcfirst <- as.numeric(germandata$fas_tcfirst)
```


Abbreviations in the code:
AS=associative strength
FAS = forward AS
BAS = backward AS
tc = refers to target-cue/cue-target associations (by the signaller) #
(cr = referst to response-cue/cue-response response asociations (by the guesser))

*Insert German forward AS (FAS) from database in data collection file*

```{r}
###AS to AS 

#for all forward AS (FAS):
#for target --> cues 1-5

#target -> cue 1; fas_tcfirst

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$first_cue[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_tcfirst[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_tcfirst[i] <- NA
  }
}


#for fas_tcsecond

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$second_cue[i]))) #second cue
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_tcsecond[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_tcsecond[i] <- NA
  }
}

#for fas_tcthird

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$third_cue[i]))) #second cue
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_tcthird[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_tcthird[i] <- NA
  }
}

#for fas_tcfourth

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$fourth_cue[i]))) #second cue
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_tcfourth[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_tcfourth[i] <- NA
  }
} 

#for fas_tcfifth

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$fifth_cue[i]))) #second cue
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_tcfifth[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_tcfifth[i] <- NA
  }
}


#for cue 1-5 --> responses 1-5
#cue 1 --> response 1

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$first_cue[i]))&(as.character(german_database$cue) == as.character(germandata$first_response[i]))) 
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_crfirst[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_crfirst[i] <- NA
  }
}

#cue 2 --> response 2

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$second_cue[i]))&(as.character(german_database$cue) == as.character(germandata$second_response[i]))) 
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_crsecond[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_crsecond[i] <- NA
  }
}


#cue 3 --> response 3

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$third_cue[i]))&(as.character(german_database$cue) == as.character(germandata$third_response[i]))) 
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_crthird[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_crthird[i] <- NA
  }
}


#cue 4 --> response 4

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$fourth_cue[i]))&(as.character(german_database$cue) == as.character(germandata$fourth_response[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_crfourth[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_crfourth[i] <- NA
  }
}

#cue 5 --> response 5

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$fifth_cue[i]))&(as.character(german_database$cue) == as.character(germandata$fifth_response[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$fas_crfifth[i] <- german_database$FAS[matching_row_in_german_database]
  } else {
    germandata$fas_crfifth[i] <- NA
  }
}
```


*Insert German backward AS (BAS) from database in data collection file*

```{r}
#for all the German BAS
#cue 1-5 to response 1-5

#cue 1 --> target


for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$first_cue[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_tcfirst[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_tcfirst[i] <- NA
  }
}

#cue 2 --> target

for (i in (1:nrow(germandata))){
 matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$second_cue[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_tcsecond[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_tcsecond[i] <- NA
  }
}

#cue 3 --> target

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$third_cue[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_tcthird[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_tcthird[i] <- NA
  }
}

#cue 4 --> target

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$fourth_cue[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_tcfourth[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_tcfourth[i] <- NA
  }
}


#cue 5 --> target

for (i in (1:nrow(germandata))){
 matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$target[i]))&(as.character(german_database$cue) == as.character(germandata$fifth_cue[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_tcfifth[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_tcfifth[i] <- NA
  }
}

```

```{r}
#response 1-5 --> cue 1-5
#response 1 --> cue 1

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$first_cue[i]))&(as.character(german_database$cue) == as.character(germandata$first_response[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_crfirst[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_crfirst[i] <- NA
  }
}

#response 2 --> cue 2

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$tsecond_cue[i]))&(as.character(german_database$cue) == as.character(germandata$second_response[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_crsecond[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_crsecond[i] <- NA
  }
}

#response 3 --> cue 3

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$third_cue[i]))&(as.character(german_database$cue) == as.character(germandata$third_response[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_crthird[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_crthird[i] <- NA
  }
}


#response 4 --> cue 4

for (i in (1:nrow(germandata))){
 matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$fourth_cue[i]))&(as.character(german_database$cue) == as.character(germandata$fourth_response[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_crfourth[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_crfourth[i] <- NA
  }
}

#response 5 --> cue 5

for (i in (1:nrow(germandata))){
  matching_row_in_german_database <- which((as.character(german_database$target) == as.character(germandata$fifth_cue[i]))&(as.character(german_database$cue) == as.character(germandata$fifth_response[i])))
  if (length(matching_row_in_german_database) == 1){
    germandata$bas_crfifth[i] <- german_database$BAS[matching_row_in_german_database]
  } else {
    germandata$bas_crfifth[i] <- NA
  }
}
```


```{r}
#replace all empty cells with NA

germandata[germandata == ""] <- NA
```

```{r}
#save amended germandata with the added AS from the database in folder

write.csv(germandata, file="germandata_withAS.csv")

```






*same for the English data*

The English Small World of Words (SWOW) database already has AS (not frequencies), so I can directly start with reading in the the database and inserting the AS from SWOW into my English data collection file

```{r}
#read in English database (=SWOW)
swow <- read.csv('Assocs.csv', sep=",", header=TRUE)

#sanity checks
str(swow)
head(swow)
summary(swow)

#read in English data: data collection document
englishdata <- read.csv('englishdata.csv', sep=",", header=TRUE) #cleanded up without lemmatisation

#sanity checks
str(englishdata)
head(englishdata)
summary(englishdata)
```

```{r}
#ensure that types stored correctly

swow$ï..cue <- as.factor(swow$ï..cue)
swow$response <- as.factor(swow$response)
englishdata$target <- as.factor(englishdata$target)


swow$R1.Strength <- as.numeric(swow$R1.Strength)
englishdata$fas_tcfirst <- as.numeric(englishdata$fas_tcfirst)
```


*Insert English FAS from database in data collection file*


```{r}
 
#for all English forward AS (FAS):

#for target --> cues 1-5

#for target --> cue 1; fas_tcfirst

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$target[i]))&(as.character(swow$response) == as.character(englishdata$first_cue[i])))
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_tcfirst[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_tcfirst[i] <- NA
  }
}


#for target --> cue 2; fas_tcsecond

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$target[i]))&(as.character(swow$response)  == as.character(englishdata$second_cue[i])))  # for second cue!
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_tcsecond[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_tcsecond[i] <- NA
  }
}


#for target --> cue 3; fas_tcthird

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$target[i]))&(as.character(swow$response)  == as.character(englishdata$third_cue[i])))  # for third cue!
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_tcthird[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_tcthird[i] <- NA
  }
}

#for fas_tcfourth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$target[i]))&(as.character(swow$response)  == as.character(englishdata$fourth_cue[i])))  # for fourth cue!
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_tcfourth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_tcfourth[i] <- NA
  }
}

#for fas_tcfifth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$target[i]))&(as.character(swow$response)  == as.character(englishdata$fifth_cue[i])))  # for fifth cue!
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_tcfifth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_tcfifth[i] <- NA
  }
}


#same for cues 1-5 -->responses 1-5
#for cue 1--> response 1; fas_crfirst

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$first_cue[i]))&(as.character(swow$response)  == as.character(englishdata$first_response[i])))  # for first response
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_crfirst[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_crfirst[i] <- NA
  }
}
#for cue 2 --> response 2;  fas_crsecond

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$second_cue[i]))&(as.character(swow$response)  == as.character(englishdata$second_response[i])))  # for second response
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_crsecond[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_crsecond[i] <- NA
  }
}

#for fas_crthird

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$third_cue[i]))&(as.character(swow$response)  == as.character(englishdata$third_response[i])))  # for third response
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_crthird[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_crthird[i] <- NA
  }
}

#for fas_crfourth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$fourth_cue[i]))&(as.character(swow$response)  == as.character(englishdata$fourth_response[i])))  # for fourth response
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_crfourth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_crfourth[i] <- NA
  }
}

#for fas_crfifth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$fifth_cue[i]))&(as.character(swow$response)  == as.character(englishdata$fifth_response[i])))  # for fifth response
  if (length(matching_row_in_swow) == 1){
    englishdata$fas_crfifth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$fas_crfifth[i] <- NA
  }
}


```


*Insert English BAS from database in data collection file*

```{r}
#for all English backward AS (BAS)

#for cue 1-5 --> target
#cue1 --> target; bas_tcfirst

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$first_cue[i]))&(as.character(swow$response)  == as.character(englishdata$target[i])))  # bas for first cue
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_tcfirst[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_tcfirst[i] <- NA
  }
}


#cue2 --> target; bas_tcsecond


for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$second_cue[i]))&(as.character(swow$response)  == as.character(englishdata$target[i])))  # bas for second cue
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_tcsecond[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_tcsecond[i] <- NA
  }
}


#cue3 --> target; bas_tcthird

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$third_cue[i]))&(as.character(swow$response)  == as.character(englishdata$target[i])))  #bas for third cue
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_tcthird[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_tcthird[i] <- NA
  }
}


#cue4 --> target; bas_tcfourth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$fourth_cue[i]))&(as.character(swow$response)  == as.character(englishdata$target[i])))  # bas for fourth cue
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_tcfourth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_tcfourth[i] <- NA
  }
}


#cue5 --> target; bas_tcfifth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$fifth_cue[i]))&(as.character(swow$response)  == as.character(englishdata$target[i])))  #bas for fifth cue
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_tcfifth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_tcfifth[i] <- NA
  }
}


#same for responses 1-5 --> cue1-5 
#response1 --> cue1; bas_crfirst

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$first_response[i]))&(as.character(swow$response)  == as.character(englishdata$first_cue[i])))  
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_crfirst[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_crfirst[i] <- NA
  }
}


#response2 --> cue2; bas_crsecond

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$second_response[i]))&(as.character(swow$response)  == as.character(englishdata$second_cue[i])))  
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_crsecond[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_crsecond[i] <- NA
  }
}

#response3 --> cue3; bas_crsecond

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$third_response[i]))&(as.character(swow$response)  == as.character(englishdata$third_cue[i])))  #bas for fifth cue
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_crthird[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_crthird[i] <- NA
  }
}


#response4 --> cue4; bas_crfourth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$fourth_response[i]))&(as.character(swow$response)  == as.character(englishdata$fourth_cue[i])))  
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_crfourth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_crfourth[i] <- NA
  }
}


#response5 --> cue5; bas_crfifth

for (i in (1:nrow(englishdata))){
  matching_row_in_swow <- which((as.character(swow$ï..cue) == as.character(englishdata$fifth_response[i]))&(as.character(swow$response)  == as.character(englishdata$fifth_cue[i])))  
  if (length(matching_row_in_swow) == 1){
    englishdata$bas_crfifth[i] <- swow$R1.Strength[matching_row_in_swow]
  } else {
    englishdata$bas_crfifth[i] <- NA
  }
}
```


```{r}
#replace all empty cells with NA

englishdata[englishdata == ""] <- NA

```

```{r}
#save edited englishdata with AS from the SWOW database as a csv file

write.csv(englishdata, file="englishdata_withAS.csv")
```