analyses.Rmd

---
title: "pwdm_analyses"
output: html_document
---

Copyright © 2022 Peter Mayer, Collins W. Munyendo, Michelle L. Mazurek, and Adam J. Aviv

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

# Packages
library(dplyr)
library(AICcmodavg)

# Working Directory
knitr::opts_knit$set(root.dir = file.path("./"))
```


```{r read in data and exclusion of invalid cases}
# Read in data
data_complete <- read.csv("data_cleaned.csv",header=TRUE,sep=";",stringsAsFactors=TRUE,check.names = FALSE)

# Exlcude answers
excluded_IDs_invalid_responses <- c("R_3fvnZ2tMIK8CTq3", "R_2coM9bJevimmYpG", "R_2ZVPjMUZF2CTvGk")
excluded_IDs_empty_cells <- c("R_2TnSUF7noksmvvo")
excluded_IDs <- c(excluded_IDs_invalid_responses, excluded_IDs_empty_cells)
data_post_exclusion <- data_complete[!data_complete$ResponseId %in% excluded_IDs,]
```


```{r prepare STEFDATT variables}
# Security
##########
data_post_exclusion$STEFDATT_security_num[data_post_exclusion$Q11_1 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_security_num[data_post_exclusion$Q11_1 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_security_num[data_post_exclusion$Q11_1 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_security_num[data_post_exclusion$Q11_1 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_security_num[data_post_exclusion$Q11_1 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_security_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin security data
data_post_exclusion$STEFDATT_security_binned[data_post_exclusion$STEFDATT_security_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_security_binned[data_post_exclusion$STEFDATT_security_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_security_binned.f <- factor(data_post_exclusion$STEFDATT_security_binned)
data_post_exclusion$STEFDATT_security_binned.f <- relevel(data_post_exclusion$STEFDATT_security_binned.f, ref = "Not Agree")


# Tranquility
#############
data_post_exclusion$STEFDATT_tranquility_num[data_post_exclusion$Q11_2 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_tranquility_num[data_post_exclusion$Q11_2 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_tranquility_num[data_post_exclusion$Q11_2 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_tranquility_num[data_post_exclusion$Q11_2 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_tranquility_num[data_post_exclusion$Q11_2 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_tranquility_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin tranquility data
data_post_exclusion$STEFDATT_tranquility_binned[data_post_exclusion$STEFDATT_tranquility_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_tranquility_binned[data_post_exclusion$STEFDATT_tranquility_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_tranquility_binned.f <- factor(data_post_exclusion$STEFDATT_tranquility_binned)
data_post_exclusion$STEFDATT_tranquility_binned.f <- relevel(data_post_exclusion$STEFDATT_tranquility_binned.f, ref = "Not Agree")


# Fun
#####
data_post_exclusion$STEFDATT_fun_num[data_post_exclusion$Q11_3 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_fun_num[data_post_exclusion$Q11_3 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_fun_num[data_post_exclusion$Q11_3 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_fun_num[data_post_exclusion$Q11_3 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_fun_num[data_post_exclusion$Q11_3 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_fun_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin fun data
data_post_exclusion$STEFDATT_fun_binned[data_post_exclusion$STEFDATT_fun_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_fun_binned[data_post_exclusion$STEFDATT_fun_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_fun_binned.f <- factor(data_post_exclusion$STEFDATT_fun_binned)
data_post_exclusion$STEFDATT_fun_binned.f <- relevel(data_post_exclusion$STEFDATT_fun_binned.f, ref = "Not Agree")


# Easy-to-use
#############
data_post_exclusion$STEFDATT_easy_num[data_post_exclusion$Q11_4 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_easy_num[data_post_exclusion$Q11_4 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_easy_num[data_post_exclusion$Q11_4 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_easy_num[data_post_exclusion$Q11_4 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_easy_num[data_post_exclusion$Q11_4 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_easy_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin easy-to-use data
data_post_exclusion$STEFDATT_easy_binned[data_post_exclusion$STEFDATT_easy_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_easy_binned[data_post_exclusion$STEFDATT_easy_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_easy_binned.f <- factor(data_post_exclusion$STEFDATT_easy_binned)
data_post_exclusion$STEFDATT_easy_binned.f <- relevel(data_post_exclusion$STEFDATT_easy_binned.f, ref = "Not Agree")


# Difficult-to-use
##################
data_post_exclusion$STEFDATT_difficult_num[data_post_exclusion$Q11_5 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_difficult_num[data_post_exclusion$Q11_5 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_difficult_num[data_post_exclusion$Q11_5 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_difficult_num[data_post_exclusion$Q11_5 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_difficult_num[data_post_exclusion$Q11_5 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_difficult_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin difficult-to-use data
data_post_exclusion$STEFDATT_difficult_binned[data_post_exclusion$STEFDATT_difficult_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_difficult_binned[data_post_exclusion$STEFDATT_difficult_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_difficult_binned.f <- factor(data_post_exclusion$STEFDATT_difficult_binned)
data_post_exclusion$STEFDATT_difficult_binned.f <- relevel(data_post_exclusion$STEFDATT_difficult_binned.f, ref = "Not Agree")


# Annoying
##########
data_post_exclusion$STEFDATT_annoying_num[data_post_exclusion$Q11_6 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_annoying_num[data_post_exclusion$Q11_6 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_annoying_num[data_post_exclusion$Q11_6 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_annoying_num[data_post_exclusion$Q11_6 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_annoying_num[data_post_exclusion$Q11_6 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_annoying_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin annoying data
data_post_exclusion$STEFDATT_annoying_binned[data_post_exclusion$STEFDATT_annoying_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_annoying_binned[data_post_exclusion$STEFDATT_annoying_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_annoying_binned.f <- factor(data_post_exclusion$STEFDATT_annoying_binned)
data_post_exclusion$STEFDATT_annoying_binned.f <- relevel(data_post_exclusion$STEFDATT_annoying_binned.f, ref = "Not Agree")


# Transparency
##############
data_post_exclusion$STEFDATT_transparency_num[data_post_exclusion$Q11_7 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_transparency_num[data_post_exclusion$Q11_7 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_transparency_num[data_post_exclusion$Q11_7 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_transparency_num[data_post_exclusion$Q11_7 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_transparency_num[data_post_exclusion$Q11_7 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_transparency_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin transparency data
data_post_exclusion$STEFDATT_transparency_binned[data_post_exclusion$STEFDATT_transparency_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_transparency_binned[data_post_exclusion$STEFDATT_transparency_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_transparency_binned.f <- factor(data_post_exclusion$STEFDATT_transparency_binned)
data_post_exclusion$STEFDATT_transparency_binned.f <- relevel(data_post_exclusion$STEFDATT_transparency_binned.f, ref = "Not Agree")

# Trust
#######
data_post_exclusion$STEFDATT_trust_num[data_post_exclusion$Q11_8 == "Strongly Disagree"] <- 1
data_post_exclusion$STEFDATT_trust_num[data_post_exclusion$Q11_8 == "Somewhat Disagree"] <- 2
data_post_exclusion$STEFDATT_trust_num[data_post_exclusion$Q11_8 == "Neither agree nor disagree"] <- 3
data_post_exclusion$STEFDATT_trust_num[data_post_exclusion$Q11_8 == "Somewhat Agree"] <- 4
data_post_exclusion$STEFDATT_trust_num[data_post_exclusion$Q11_8 == "Strongly Agree"] <- 5
#hist(data_post_exclusion$STEFDATT_trust_num, breaks = c(0, 1, 2, 3, 4, 5))

# Bin trust data
data_post_exclusion$STEFDATT_trust_binned[data_post_exclusion$STEFDATT_trust_num %in% c(1,2,3)] <- "Not Agree"
data_post_exclusion$STEFDATT_trust_binned[data_post_exclusion$STEFDATT_trust_num %in% c(4,5)] <- "Agree"
data_post_exclusion$STEFDATT_trust_binned.f <- factor(data_post_exclusion$STEFDATT_trust_binned)
data_post_exclusion$STEFDATT_trust_binned.f <- relevel(data_post_exclusion$STEFDATT_trust_binned.f, ref = "Not Agree")
```


```{r prepare role variable}
# Binned into student and non-student; used in analyses
data_post_exclusion$role[data_post_exclusion$A2 %in% c("Student")] <- "Student"
data_post_exclusion$role[!(data_post_exclusion$A2 %in% c("Student"))] <- "Non-Student"
data_post_exclusion$role.f <- factor(data_post_exclusion$role)

# Binned into student, faculty/leadership, staff, and other/prefer not to disclose; used for demographics
data_post_exclusion$role_ext[data_post_exclusion$A2 %in% c("Student")] <- "Student"
data_post_exclusion$role_ext[(data_post_exclusion$A2 %in% c("Full-time Faculty", "Part-time/Adjunct Faculty", "Leadership (provost/dean/president’s office)"))] <- "Faculty/Leadership"
data_post_exclusion$role_ext[(data_post_exclusion$A2 %in% c("Staff (non-faculty/administrative)"))] <- "Staff (non-faculty/administrative)"
data_post_exclusion$role_ext[(data_post_exclusion$A2 %in% c("Prefer not to disclose", "Other"))] <- "Other/Prefer not to disclose"
data_post_exclusion$role_ext.f <- factor(data_post_exclusion$role_ext)
```


```{r prepare SA-6 variable}
# Item 1
data_post_exclusion$A71_num[data_post_exclusion$A7_1 == "Strongly Disagree"] <- 1
data_post_exclusion$A71_num[data_post_exclusion$A7_1 == "Disagree"] <- 2
data_post_exclusion$A71_num[data_post_exclusion$A7_1 == "Somewhat Disagree"] <- 3
data_post_exclusion$A71_num[data_post_exclusion$A7_1 == "Neither agree nor disagree"] <- 4
data_post_exclusion$A71_num[data_post_exclusion$A7_1 == "Somewhat Agree"] <- 5
data_post_exclusion$A71_num[data_post_exclusion$A7_1 == "Agree"] <- 6
data_post_exclusion$A71_num[data_post_exclusion$A7_1 == "Strongly Agree"] <- 7

# Item 2
data_post_exclusion$A72_num[data_post_exclusion$A7_2 == "Strongly Disagree"] <- 1
data_post_exclusion$A72_num[data_post_exclusion$A7_2 == "Disagree"] <- 2
data_post_exclusion$A72_num[data_post_exclusion$A7_2 == "Somewhat Disagree"] <- 3
data_post_exclusion$A72_num[data_post_exclusion$A7_2 == "Neither agree nor disagree"] <- 4
data_post_exclusion$A72_num[data_post_exclusion$A7_2 == "Somewhat Agree"] <- 5
data_post_exclusion$A72_num[data_post_exclusion$A7_2 == "Agree"] <- 6
data_post_exclusion$A72_num[data_post_exclusion$A7_2 == "Strongly Agree"] <- 7

# Item 3
data_post_exclusion$A73_num[data_post_exclusion$A7_3 == "Strongly Disagree"] <- 1
data_post_exclusion$A73_num[data_post_exclusion$A7_3 == "Disagree"] <- 2
data_post_exclusion$A73_num[data_post_exclusion$A7_3 == "Somewhat Disagree"] <- 3
data_post_exclusion$A73_num[data_post_exclusion$A7_3 == "Neither agree nor disagree"] <- 4
data_post_exclusion$A73_num[data_post_exclusion$A7_3 == "Somewhat Agree"] <- 5
data_post_exclusion$A73_num[data_post_exclusion$A7_3 == "Agree"] <- 6
data_post_exclusion$A73_num[data_post_exclusion$A7_3 == "Strongly Agree"] <- 7


# Item 4
data_post_exclusion$A74_num[data_post_exclusion$A7_4 == "Strongly Disagree"] <- 1
data_post_exclusion$A74_num[data_post_exclusion$A7_4 == "Disagree"] <- 2
data_post_exclusion$A74_num[data_post_exclusion$A7_4 == "Somewhat Disagree"] <- 3
data_post_exclusion$A74_num[data_post_exclusion$A7_4 == "Neither agree nor disagree"] <- 4
data_post_exclusion$A74_num[data_post_exclusion$A7_4 == "Somewhat Agree"] <- 5
data_post_exclusion$A74_num[data_post_exclusion$A7_4 == "Agree"] <- 6
data_post_exclusion$A74_num[data_post_exclusion$A7_4 == "Strongly Agree"] <- 7

# Item 5
data_post_exclusion$A75_num[data_post_exclusion$A7_5 == "Strongly Disagree"] <- 1
data_post_exclusion$A75_num[data_post_exclusion$A7_5 == "Disagree"] <- 2
data_post_exclusion$A75_num[data_post_exclusion$A7_5 == "Somewhat Disagree"] <- 3
data_post_exclusion$A75_num[data_post_exclusion$A7_5 == "Neither agree nor disagree"] <- 4
data_post_exclusion$A75_num[data_post_exclusion$A7_5 == "Somewhat Agree"] <- 5
data_post_exclusion$A75_num[data_post_exclusion$A7_5 == "Agree"] <- 6
data_post_exclusion$A75_num[data_post_exclusion$A7_5 == "Strongly Agree"] <- 7

# Item 6
data_post_exclusion$A76_num[data_post_exclusion$A7_6 == "Strongly Disagree"] <- 1
data_post_exclusion$A76_num[data_post_exclusion$A7_6 == "Disagree"] <- 2
data_post_exclusion$A76_num[data_post_exclusion$A7_6 == "Somewhat Disagree"] <- 3
data_post_exclusion$A76_num[data_post_exclusion$A7_6 == "Neither agree nor disagree"] <- 4
data_post_exclusion$A76_num[data_post_exclusion$A7_6 == "Somewhat Agree"] <- 5
data_post_exclusion$A76_num[data_post_exclusion$A7_6 == "Agree"] <- 6
data_post_exclusion$A76_num[data_post_exclusion$A7_6 == "Strongly Agree"] <- 7

# Combine it
for (i in 1:277) {
  data_post_exclusion$SA6[i] <- mean(c(data_post_exclusion$A71_num[i], data_post_exclusion$A72_num[i], data_post_exclusion$A73_num[i], data_post_exclusion$A74_num[i], data_post_exclusion$A75_num[i], data_post_exclusion$A76_num[i]))
}

# Overall
mean(data_post_exclusion$SA6)
sd(data_post_exclusion$SA6)

# per Role
tapply(data_post_exclusion$SA6, data_post_exclusion$role_ext.f, mean)
tapply(data_post_exclusion$SA6, data_post_exclusion$role_ext.f, sd)
```


```{r prepare Web Skills Index variable}
# Advanced Search
data_post_exclusion$A91_num[data_post_exclusion$A9_1 == "No understanding"] <- 1
data_post_exclusion$A91_num[data_post_exclusion$A9_1 == "Low understanding"] <- 2
data_post_exclusion$A91_num[data_post_exclusion$A9_1 == "Medium understanding"] <- 3
data_post_exclusion$A91_num[data_post_exclusion$A9_1 == "High understanding"] <- 4
data_post_exclusion$A91_num[data_post_exclusion$A9_1 == "Full Understanding"] <- 5

# PDF
data_post_exclusion$A92_num[data_post_exclusion$A9_2 == "No understanding"] <- 1
data_post_exclusion$A92_num[data_post_exclusion$A9_2 == "Low understanding"] <- 2
data_post_exclusion$A92_num[data_post_exclusion$A9_2 == "Medium understanding"] <- 3
data_post_exclusion$A92_num[data_post_exclusion$A9_2 == "High understanding"] <- 4
data_post_exclusion$A92_num[data_post_exclusion$A9_2 == "Full Understanding"] <- 5

# Spyware
data_post_exclusion$A93_num[data_post_exclusion$A9_3 == "No understanding"] <- 1
data_post_exclusion$A93_num[data_post_exclusion$A9_3 == "Low understanding"] <- 2
data_post_exclusion$A93_num[data_post_exclusion$A9_3 == "Medium understanding"] <- 3
data_post_exclusion$A93_num[data_post_exclusion$A9_3 == "High understanding"] <- 4
data_post_exclusion$A93_num[data_post_exclusion$A9_3 == "Full Understanding"] <- 5

# Wiki
data_post_exclusion$A94_num[data_post_exclusion$A9_4 == "No understanding"] <- 1
data_post_exclusion$A94_num[data_post_exclusion$A9_4 == "Low understanding"] <- 2
data_post_exclusion$A94_num[data_post_exclusion$A9_4 == "Medium understanding"] <- 3
data_post_exclusion$A94_num[data_post_exclusion$A9_4 == "High understanding"] <- 4
data_post_exclusion$A94_num[data_post_exclusion$A9_4 == "Full Understanding"] <- 5

# Cache
data_post_exclusion$A95_num[data_post_exclusion$A9_5 == "No understanding"] <- 1
data_post_exclusion$A95_num[data_post_exclusion$A9_5 == "Low understanding"] <- 2
data_post_exclusion$A95_num[data_post_exclusion$A9_5 == "Medium understanding"] <- 3
data_post_exclusion$A95_num[data_post_exclusion$A9_5 == "High understanding"] <- 4
data_post_exclusion$A95_num[data_post_exclusion$A9_5 == "Full Understanding"] <- 5

# Phishing
data_post_exclusion$A96_num[data_post_exclusion$A9_6 == "No understanding"] <- 1
data_post_exclusion$A96_num[data_post_exclusion$A9_6 == "Low understanding"] <- 2
data_post_exclusion$A96_num[data_post_exclusion$A9_6 == "Medium understanding"] <- 3
data_post_exclusion$A96_num[data_post_exclusion$A9_6 == "High understanding"] <- 4
data_post_exclusion$A96_num[data_post_exclusion$A9_6 == "Full Understanding"] <- 5

for (i in 1:277) {
  data_post_exclusion$WebSkillIndex[i] <- mean(c(data_post_exclusion$A91_num[i], data_post_exclusion$A92_num[i], data_post_exclusion$A93_num[i], data_post_exclusion$A94_num[i], data_post_exclusion$A95_num[i], data_post_exclusion$A96_num[i]))
}

# Overall
mean(data_post_exclusion$WebSkillIndex)
sd(data_post_exclusion$WebSkillIndex)

# per Role
tapply(data_post_exclusion$WebSkillIndex, data_post_exclusion$role_ext.f, mean)
tapply(data_post_exclusion$WebSkillIndex, data_post_exclusion$role_ext.f, sd)
```


```{r prepare GW Password Security variable}
data_post_exclusion$GWpwdSec[data_post_exclusion$Q6 == "Much less secure"] <- 1
data_post_exclusion$GWpwdSec[data_post_exclusion$Q6 == "Somewhat less secure"] <- 2
data_post_exclusion$GWpwdSec[data_post_exclusion$Q6 == "About equally secure"] <- 3
data_post_exclusion$GWpwdSec[data_post_exclusion$Q6 == "Somewhat less secure"] <- 4
data_post_exclusion$GWpwdSec[data_post_exclusion$Q6 == "Much more secure"] <- 5

# "Unsure" responses are excluded
data_post_exclusion$GWpwdSec_binned[data_post_exclusion$GWpwdSec %in% c(1,2,3)] <- "Not more secure"
data_post_exclusion$GWpwdSec_binned[data_post_exclusion$GWpwdSec %in% c(4,5)] <- "More secure"
data_post_exclusion$GWpwdSec_binned.f <- factor(data_post_exclusion$GWpwdSec_binned)
data_post_exclusion$GWpwdSec_binned.f <- relevel(data_post_exclusion$GWpwdSec_binned.f, ref = "Not more secure")
```


```{r prepare password manager use variable}
# Dichotomous
data_post_exclusion$pwdm_use[data_post_exclusion$Q8 %in% c("I do not use a password manager.")] <- "No"
data_post_exclusion$pwdm_use[!(data_post_exclusion$Q8 %in% c("I do not use a password manager."))] <- "Yes"
data_post_exclusion$pwdm_use.f <- factor(data_post_exclusion$pwdm_use)
data_post_exclusion$pwdm_use.f <- relevel(data_post_exclusion$pwdm_use.f, ref = "No")
table(data_post_exclusion$pwdm_use.f)

# Details retained
data_post_exclusion$pwdm_use_binned[data_post_exclusion$Q8 %in% c("I do not use a password manager.")] <- "Do not use"
data_post_exclusion$pwdm_use_binned[data_post_exclusion$Q8 %in% c("I save my passwords in the browser (for example, passwords saved in Chrome).")] <- "Browser"
data_post_exclusion$pwdm_use_binned[data_post_exclusion$Q8 %in% c("I use a system provided password manager (for example, Apple's Keychain).")] <- "OS"
data_post_exclusion$pwdm_use_binned[data_post_exclusion$Q8 %in% c("I use a third-party manager (for example, Lastpass or Onepass).")] <- "Third-party"
data_post_exclusion$pwdm_use_binned[!(data_post_exclusion$Q8 %in% c("I do not use a password manager.", "I save my passwords in the browser (for example, passwords saved in Chrome).", "I use a system provided password manager (for example, Apple's Keychain).", "I use a third-party manager (for example, Lastpass or Onepass)."))] <- "Multiple"
data_post_exclusion$pwdm_use_binned.f <- factor(data_post_exclusion$pwdm_use_binned)

data_post_exclusion$pwdm_browser <- sapply(data_post_exclusion$Q8, function(x) grepl("browser", x, fixed = TRUE))
data_post_exclusion$pwdm_system <- sapply(data_post_exclusion$Q8, function(x) grepl("system", x, fixed = TRUE))
data_post_exclusion$pwdm_third_party <- sapply(data_post_exclusion$Q8, function(x) grepl("third-party", x, fixed = TRUE))
```


```{r prepare password generated by password manager variable}
data_post_exclusion$pwdm_generated[data_post_exclusion$Q21 %in% c("")] <- NA
data_post_exclusion$pwdm_generated[data_post_exclusion$Q21 %in% c("I let the password manager create and store the password")] <- "Yes"
data_post_exclusion$pwdm_generated[!(data_post_exclusion$Q21 %in% c("I let the password manager create and store the password", ""))] <- "No"
data_post_exclusion$pwdm_generated.f <- factor(data_post_exclusion$pwdm_generated)
data_post_exclusion$pwdm_generated.f <- relevel(data_post_exclusion$pwdm_generated.f, ref = "No")
```


```{r prepare pwdm use in organisation variable}
data_post_exclusion$PWDMuseInOrg[data_post_exclusion$Q24 == "Extremely unlikely"] <- 1
data_post_exclusion$PWDMuseInOrg[data_post_exclusion$Q24 == "Moderately unlikely"] <- 2
data_post_exclusion$PWDMuseInOrg[data_post_exclusion$Q24 == "Slightly unlikely"] <- 3
data_post_exclusion$PWDMuseInOrg[data_post_exclusion$Q24 == "Neither likely nor unlikely"] <- 4
data_post_exclusion$PWDMuseInOrg[data_post_exclusion$Q24 == "Slightly likely"] <- 5
data_post_exclusion$PWDMuseInOrg[data_post_exclusion$Q24 == "Moderately likely"] <- 6
data_post_exclusion$PWDMuseInOrg[data_post_exclusion$Q24 == "Extremely likely"] <- 7
data_post_exclusion$PWDMuseInOrg.f <- factor(data_post_exclusion$Q24, levels=c("Extremely unlikely", "Moderately unlikely", "Slightly unlikely", "Neither likely nor unlikely", "Slightly likely", "Moderately likely", "Extremely likely"), ordered=TRUE)

# Binning with "neither likely nor unlikely" sorted into "no" to not overestimate effet
data_post_exclusion$PWDMuseInOrg_binned.f[data_post_exclusion$PWDMuseInOrg %in% c(1,2,3,4)] <- "No"
data_post_exclusion$PWDMuseInOrg_binned.f[data_post_exclusion$PWDMuseInOrg %in% c(5,6,7)] <- "Yes"
data_post_exclusion$PWDMuseInOrg_binned.f <- factor(data_post_exclusion$PWDMuseInOrg_binned.f)
```


```{r prepare pwdm awareness variable}
data_post_exclusion$pwdm_awareness[data_post_exclusion$Q10 %in% c("I first heard about it in this study")] <- "No"
data_post_exclusion$pwdm_awareness[!(data_post_exclusion$Q10 %in% c("I first heard about it in this study"))] <- "Yes"
data_post_exclusion$pwdm_awareness.f <- factor(data_post_exclusion$pwdm_awareness)
data_post_exclusion$pwdm_awareness.f <- relevel(data_post_exclusion$pwdm_awareness.f, ref = "No")
```


```{r prepare gender variable}
# Drop prefer not to disclose & non-binary due to low counts
data_post_exclusion$gender[data_post_exclusion$GENDER %in% c("Man")] <- "Man"
data_post_exclusion$gender[data_post_exclusion$GENDER %in% c("Woman")] <- "Woman"
data_post_exclusion$gender.f <- factor(data_post_exclusion$gender, ordered = FALSE)

table(data_post_exclusion$GENDER, data_post_exclusion$role_ext.f, dnn = c("Gender", "Role"))/277
```


```{r prepare school variable}
data_post_exclusion$school.f <- factor(data_post_exclusion$A1, ordered = FALSE)
```


```{r prepare ethnicity variables}
data_post_exclusion$eth_hispanic[data_post_exclusion$ETH_HISP %in% c("No, not of Hispanic, Latino/a/x, or Spanish origin")] <- "No"
data_post_exclusion$eth_hispanic[data_post_exclusion$ETH_HISP %in% c("Prefer not to disclose")] <- "Undisclosed"
data_post_exclusion$eth_hispanic[!(data_post_exclusion$ETH_HISP %in% c("No, not of Hispanic, Latino/a/x, or Spanish origin", "Prefer not to disclose"))] <- "Yes"
data_post_exclusion$eth_hispanic.f <- factor(data_post_exclusion$eth_hispanic, ordered = FALSE)

table(data_post_exclusion$eth_hispanic.f,data_post_exclusion$role_ext.f)

# Combine small counts into one "other" variable
data_post_exclusion$eth_race[data_post_exclusion$ETH_R %in% c("White")] <- "White"
data_post_exclusion$eth_race[data_post_exclusion$ETH_R %in% c("Black or African American")] <- "Black or African American"
data_post_exclusion$eth_race[data_post_exclusion$ETH_R %in% c("Prefer not to disclose")] <- "Prefer not to disclose"
data_post_exclusion$eth_race[!(data_post_exclusion$ETH_R %in% c("White", "Black or African American", "Prefer not to disclose"))] <- "Other"
data_post_exclusion$eth_race.f <- factor(data_post_exclusion$eth_race, ordered = FALSE)

data_eth_nohis <- subset(data_post_exclusion, eth_hispanic.f != "Yes")
table(data_eth_nohis$eth_race.f,data_eth_nohis$role_ext.f)
# n_faculty = 58
# n_staff = 118
```


```{r binning of age variable}
data_post_exclusion$age_binned[data_post_exclusion$AGE_NUM %in% c(18, 19, 20, 21, 22, 23, 24, 25)] <- "18-25"
data_post_exclusion$age_binned[data_post_exclusion$AGE_NUM %in% c(26, 27, 28, 29, 30, 31, 32, 33, 34, 35)] <- "26-35"
data_post_exclusion$age_binned[data_post_exclusion$AGE_NUM %in% c(36, 37, 38, 39, 40, 41, 42, 43, 44, 45)] <- "36-45"
data_post_exclusion$age_binned[data_post_exclusion$AGE_NUM %in% c(46, 47, 48, 49, 50, 51, 52, 53, 54, 55)] <- "46-55"
data_post_exclusion$age_binned[data_post_exclusion$AGE_NUM %in% c(56, 57, 58, 59, 60, 61, 63, 63, 64, 65)] <- "56-65"
data_post_exclusion$age_binned[data_post_exclusion$AGE_NUM %in% c(66, 67, 68, 71, 76, 82)] <- ">65"
data_post_exclusion$age_binned[data_post_exclusion$AGE_ITEM %in% c("Prefer not to disclose")] <- "Prefer not to disclose"
data_post_exclusion$age_binned.f <- factor(data_post_exclusion$age_binned)

table(data_post_exclusion$age_binned.f, data_post_exclusion$role_ext.f, dnn = c("Age", "Role"))
```


```{r RQ-1 pwdm awareness regression}
glm_pwdm_awareness <- glm(pwdm_awareness.f ~ SA6
                                    + WebSkillIndex
                                    + gender.f
                                    + AGE_NUM,
                               data=data_post_exclusion, family="binomial"
)
summary(glm_pwdm_awareness)
exp(summary(glm_pwdm_awareness)$coefficients[,1]["(Intercept)"])
exp(summary(glm_pwdm_awareness)$coefficients[,1]["SA6"])
exp(summary(glm_pwdm_awareness)$coefficients[,1]["WebSkillIndex"])
exp(summary(glm_pwdm_awareness)$coefficients[,1]["gender.fWoman"])
exp(summary(glm_pwdm_awareness)$coefficients[,1]["AGE_NUM"])
exp(confint(glm_pwdm_awareness))
```


```{r RQ-1 PM awareness of pm types}
pm_types_data <- select(data_post_exclusion, ResponseId, RB, Q8)

pm_types_data$pre_browser <- sapply(pm_types_data$RB, function(x) grepl("I saved my passwords in the browser (for example, passwords saved in Chrome)", x, fixed = TRUE))
pm_types_data$pre_third_party <- sapply(pm_types_data$RB, function(x) grepl("I used a third-party password manager (for example, Lastpass or Onepass)", x, fixed = TRUE))
pm_types_data$pre_system <- sapply(pm_types_data$RB, function(x) grepl("I used a system-provided password manager (for example, Apple's Keychain)", x, fixed = TRUE))

pm_types_data$post_browser <- sapply(pm_types_data$Q8, function(x) grepl("I save my passwords in the browser (for example, passwords saved in Chrome).", x, fixed = TRUE))
pm_types_data$post_third_party <- sapply(pm_types_data$Q8, function(x) grepl("I use a third-party manager (for example, Lastpass or Onepass).", x, fixed = TRUE))
pm_types_data$post_system <- sapply(pm_types_data$Q8, function(x) grepl("I use a system provided password manager (for example, Apple's Keychain).", x, fixed = TRUE))

# Changes in stated usage
pm_types_data$browser_change <- with(pm_types_data, ifelse(pre_browser != post_browser, ifelse(pre_browser, "Pre only", "Post only"), ifelse(pre_browser,"No change (use)","No change (non-use)")))
table(pm_types_data$browser_change)

pm_types_data$third_party_change <- with(pm_types_data, ifelse(pre_third_party != post_third_party, ifelse(pre_third_party, "Pre only", "Post only"), ifelse(pre_third_party,"No change (use)","No change (non-use)")))
table(pm_types_data$third_party_change)

pm_types_data$system_change <- with(pm_types_data, ifelse(pre_system != post_system, ifelse(pre_system, "Pre only", "Post only"), ifelse(pre_system,"No change (use)","No change (non-use)")))
table(pm_types_data$system_change)

pm_types_data$any_change <- with(pm_types_data, browser_change != "No change" | third_party_change != "No change" | system_change != "No change")
table(pm_types_data$any_change)

pm_types_data$num_changes <- with(pm_types_data, (browser_change != "No change") + (third_party_change != "No change") + (system_change != "No change"))
table(pm_types_data$num_changes)
```


```{r RQ-2 password management strategies}
general_strategies_data_merged <- select(data_post_exclusion, ResponseId, RB, Q8)

general_strategies_data_merged$memory <- sapply(general_strategies_data_merged$RB, function(x) grepl("I remembered my passwords without writing them down or storing them digitally", x, fixed = TRUE))
general_strategies_data_merged$reset <- sapply(general_strategies_data_merged$RB, function(x) grepl("I reset my password every time I log in rather than remembering my password", x, fixed = TRUE))
general_strategies_data_merged$file <- sapply(general_strategies_data_merged$RB, function(x) grepl("I stored my passwords in a digital file or files", x, fixed = TRUE))
general_strategies_data_merged$physical <- sapply(general_strategies_data_merged$RB, function(x) grepl("I wrote my passwords down on paper or other physical media.", x, fixed = TRUE))
general_strategies_data_merged$browser <- sapply(general_strategies_data_merged$Q8, function(x) grepl("I save my passwords in the browser (for example, passwords saved in Chrome).", x, fixed = TRUE))
general_strategies_data_merged$third_party <- sapply(general_strategies_data_merged$Q8, function(x) grepl("I use a third-party manager (for example, Lastpass or Onepass).", x, fixed = TRUE))
general_strategies_data_merged$system <- sapply(general_strategies_data_merged$Q8, function(x) grepl("I use a system provided password manager (for example, Apple's Keychain).", x, fixed = TRUE))
general_strategies_data_merged$none <- sapply(general_strategies_data_merged$Q8, function(x) grepl("None of the above", x, fixed = TRUE))

general_strategies_data_merged$any_pm <- with(general_strategies_data_merged, browser | third_party | system)

sprintf("Memorize %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$memory), sum(general_strategies_data_merged$memory)/277*100)
sprintf("Reset %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$reset), sum(general_strategies_data_merged$reset)/277*100)
sprintf("File %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$file), sum(general_strategies_data_merged$file)/277*100)
sprintf("Physical %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$physical), sum(general_strategies_data_merged$physical)/277*100)
sprintf("Browser %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$browser), sum(general_strategies_data_merged$browser)/277*100)
sprintf("Third-party %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$third_party), sum(general_strategies_data_merged$third_party)/277*100)
sprintf("System %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$system), sum(general_strategies_data_merged$system)/277*100)
sprintf("Any PM %1$1.0f (%2$2.1f%%)", sum(general_strategies_data_merged$any_pm), sum(general_strategies_data_merged$any_pm)/277*100)

# Which PMs

general_strategies_data_merged$pm_name_other <-data_post_exclusion$Q19_12_TEXT
# R_33kW4DWtY4YMotC uses Edge -> Browser
# R_3qJcsl9WKi0Sc1w uses RoboForm -> Third-party
# R_25A74UJr5HYyc6o uses Edge, Safari -> Browser
# R_3J1TlT9YyW2zE0d uses Keeper -> Third-party
# R_cCHIV1eOzQBtMRP uses Roboform -> Third-party
# R_3dJSttncu8L5sbf uses Passwordsafe -> Third-party
# R_3R1S0QqNzzZ8JYM uses Passwordsafe -> Third-party
# R_3I66B9kzz5QfRpD uses Safari -> Browser
# R_cwlAXy5XmFBeDp7 uses Bitwarden -> Third-party
# R_12x4LuIhxOzKfgK uses Norton/Lifelock -> Third-party
# R_ZKnL98YqiLLVgTn uses "My own in the google sheets apple. It is connected to my gmail and I can see on my phone."

# Close-up on reset
general_strategies_data_merged$num_stra_wo_reset <- with(general_strategies_data_merged, memory + file + physical + browser + third_party + system)
table(general_strategies_data_merged$num_stra_wo_reset, general_strategies_data_merged$reset)

# General numbers
general_strategies_data_merged$num_stra <- with(general_strategies_data_merged, memory + file + physical + browser + third_party + system + reset)
mean(general_strategies_data_merged$num_stra)
sd(general_strategies_data_merged$num_stra)
table(general_strategies_data_merged$num_stra)

```


```{r RQ-2 usage of sync}
rq2_sync_data <- dplyr::select(data_post_exclusion, ResponseId, DF1, DF2, SB1, TP1, OS1)

# file
table(rq2_sync_data$DF1, rq2_sync_data$DF2, dnn = c("manual", "tool"))

# os
table(rq2_sync_data$OS1)

# browser
table(rq2_sync_data$SB1)

# 3rd party
table(rq2_sync_data$TP1)
```


```{r RQ-2 reuse}
table(data_post_exclusion$CB3)
```


```{r RQ-2 PM generation}
pwd_generation_data <- select(data_post_exclusion, ResponseId, Q21, Q8)

pwd_generation_data$browser <- sapply(pwd_generation_data$Q8, function(x) grepl("I save my passwords in the browser (for example, passwords saved in Chrome).", x, fixed = TRUE))
pwd_generation_data$third_party <- sapply(pwd_generation_data$Q8, function(x) grepl("I use a third-party manager (for example, Lastpass or Onepass).", x, fixed = TRUE))
pwd_generation_data$system <- sapply(pwd_generation_data$Q8, function(x) grepl("I use a system provided password manager (for example, Apple's Keychain).", x, fixed = TRUE))

table(pwd_generation_data$Q21)
table(pwd_generation_data$Q21,pwd_generation_data$browser)
table(pwd_generation_data$Q21,pwd_generation_data$third_party)
table(pwd_generation_data$Q21,pwd_generation_data$system)
```


```{r RQ-2 regression password generated by pwdm}
glm_pwd_generated <- glm(pwdm_generated.f ~ SA6
                                    + WebSkillIndex
                                    + STEFDATT_security_binned.f 
                                    + STEFDATT_tranquility_binned.f 
                                    + STEFDATT_fun_binned.f 
                                    + STEFDATT_easy_binned.f
                                    + STEFDATT_difficult_binned.f
                                    + STEFDATT_annoying_binned.f
                                    + STEFDATT_transparency_binned.f
                                    + STEFDATT_trust_binned.f
                                    + GWpwdSec_binned.f,
                                  data=data_post_exclusion, family="binomial"
)
summary(glm_pwd_generated)
exp(summary(glm_pwd_generated)$coefficients[,1]["(Intercept)"])
exp(summary(glm_pwd_generated)$coefficients[,1]["SA6"])
exp(summary(glm_pwd_generated)$coefficients[,1]["WebSkillIndex"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_security_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_tranquility_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_fun_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_easy_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_difficult_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_annoying_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_transparency_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["STEFDATT_trust_binned.fAgree"])
exp(summary(glm_pwd_generated)$coefficients[,1]["GWpwdSec_binned.fMore secure"])
exp(confint(glm_pwd_generated))
```


```{r RQ-2 regression pwdm use}
glm_pwd_use <- glm(pwdm_use.f ~ SA6
                                    + WebSkillIndex
                                    + STEFDATT_security_binned.f 
                                    + STEFDATT_tranquility_binned.f 
                                    + STEFDATT_fun_binned.f 
                                    + STEFDATT_easy_binned.f
                                    + STEFDATT_difficult_binned.f
                                    + STEFDATT_annoying_binned.f
                                    + STEFDATT_transparency_binned.f
                                    + STEFDATT_trust_binned.f
                                    + GWpwdSec_binned.f
                                    + role.f, #Role
                        data=data_post_exclusion, family="binomial"
)
summary(glm_pwd_use)
exp(summary(glm_pwd_use)$coefficients[,1]["(Intercept)"])
exp(summary(glm_pwd_use)$coefficients[,1]["SA6"])
exp(summary(glm_pwd_use)$coefficients[,1]["WebSkillIndex"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_security_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_tranquility_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_fun_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_easy_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_difficult_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_annoying_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_transparency_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["STEFDATT_trust_binned.fAgree"])
exp(summary(glm_pwd_use)$coefficients[,1]["GWpwdSec_binned.fMore secure"])
exp(summary(glm_pwd_use)$coefficients[,1]["role.fStudent"])
exp(confint(glm_pwd_use))
```


```{r RQ-2 regression browser pwdm use}
glm_pwd_browser <- glm(pwdm_browser ~ SA6
                                    + WebSkillIndex
                                    + STEFDATT_security_binned.f 
                                    + STEFDATT_tranquility_binned.f 
                                    + STEFDATT_fun_binned.f 
                                    + STEFDATT_easy_binned.f
                                    + STEFDATT_difficult_binned.f
                                    + STEFDATT_annoying_binned.f
                                    + STEFDATT_transparency_binned.f
                                    + STEFDATT_trust_binned.f
                                    + GWpwdSec_binned.f
                                    + role.f
                                    + gender.f,
                   data=data_post_exclusion, family="binomial"
)
summary(glm_pwd_browser)
exp(summary(glm_pwd_browser)$coefficients[,1]["(Intercept)"])
exp(summary(glm_pwd_browser)$coefficients[,1]["SA6"])
exp(summary(glm_pwd_browser)$coefficients[,1]["WebSkillIndex"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_security_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_tranquility_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_fun_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_easy_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_difficult_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_annoying_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_transparency_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["STEFDATT_trust_binned.fAgree"])
exp(summary(glm_pwd_browser)$coefficients[,1]["GWpwdSec_binned.fMore secure"])
exp(summary(glm_pwd_browser)$coefficients[,1]["role.fStudent"])
exp(summary(glm_pwd_browser)$coefficients[,1]["gender.fWoman"])
exp(confint(glm_pwd_browser))
```


```{r RQ-2 regression third-party pwdm use}
glm_pwd_third_party <- glm(pwdm_third_party ~ SA6
                                    + WebSkillIndex
                                    + STEFDATT_security_binned.f 
                                    + STEFDATT_tranquility_binned.f 
                                    + STEFDATT_fun_binned.f 
                                    + STEFDATT_easy_binned.f
                                    + STEFDATT_difficult_binned.f
                                    + STEFDATT_annoying_binned.f
                                    + STEFDATT_transparency_binned.f
                                    + STEFDATT_trust_binned.f
                                    + GWpwdSec_binned.f
                                    + role.f, #Role
                        data=data_post_exclusion, family="binomial"
)
summary(glm_pwd_third_party)
exp(abs(summary(glm_pwd_third_party)$coefficients[,1]["(Intercept)"]))
exp(summary(glm_pwd_third_party)$coefficients[,1]["SA6"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["WebSkillIndex"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_security_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_tranquility_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_fun_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_easy_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_difficult_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_annoying_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_transparency_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["STEFDATT_trust_binned.fAgree"])
exp(summary(glm_pwd_third_party)$coefficients[,1]["GWpwdSec_binned.fMore secure"])
exp(abs(summary(glm_pwd_third_party)$coefficients[,1]["role.fStudent"]))
exp(confint(glm_pwd_third_party))
```


```{r RQ-2 regression system pwdm use}
glm_pwd_system <- glm(pwdm_system ~ SA6
                                    + WebSkillIndex
                                    + STEFDATT_security_binned.f 
                                    + STEFDATT_tranquility_binned.f 
                                    + STEFDATT_fun_binned.f 
                                    + STEFDATT_easy_binned.f
                                    + STEFDATT_difficult_binned.f
                                    + STEFDATT_annoying_binned.f
                                    + STEFDATT_transparency_binned.f
                                    + STEFDATT_trust_binned.f
                                    + role.f, #Role
                        data=data_post_exclusion, family="binomial"
)
summary(glm_pwd_system)
exp(summary(glm_pwd_system)$coefficients[,1]["(Intercept)"])
exp(summary(glm_pwd_system)$coefficients[,1]["SA6"])
exp(summary(glm_pwd_system)$coefficients[,1]["WebSkillIndex"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_security_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_tranquility_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_fun_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_easy_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_difficult_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_annoying_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_transparency_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["STEFDATT_trust_binned.fAgree"])
exp(summary(glm_pwd_system)$coefficients[,1]["role.fStudent"])
exp(confint(glm_pwd_system))
```


```{r RQ-3 autofill for uni account}
table(data_post_exclusion$Q5)
```


```{r RQ-4 motivators and barriers descriptives}
# n=212

# motivators
sort(table(data_post_exclusion$Q22)/212, decreasing = TRUE)

# barriers
sort(table(data_post_exclusion$Q23)/212, decreasing = TRUE)
```