CHER’s 2022 Health Equity Research Intensive
Josephine McKelvy
2022-09-01
R Markdown
This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see http://rmarkdown.rstudio.com.
When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document. You can embed an R code chunk like this:
###################
# READ THIS FIRST #
###################
# Right-click the R icon and "Run as administrator" > Tools > Check for Updates. Session > Restart R and Clear Output.
# Create new GitHub repository. Open New Project in RStudio w/repo URL.
# Qualtrics: Data & Analysis/Export & Import/Export Data/CSV/Use Numeric values. Extract files.
# Identify column numbers in Excel. Rename csv being read in.Overview (Steps)
- import all responses (as *.csv & numeric values)
- subset only the data needed per chart
- sort CATA by session, rename columns, bind
- factor the categorical variables
- generate average item scores for dumbbell plots
- plot the dumbbell/diverging bar charts & save
###################
# Import the data #
###################
# Import data, subset and name the variables in the dataframes, remove blanks (by language)
raw.df <- read.csv("UNC Center for Health Equity Research - Health Equity Summer Intensive Evaluation_October 7, 2022_06.34.csv", stringsAsFactors = TRUE)
raw.df <- raw.df[3:nrow(raw.df),] #cut rows 2-3 with question wording & "ImportID"
raw.df$Q3 <- as.character(raw.df$Q3) #convert CATA to character and not factor
#############
# Libraries #
#############
library(data.table) #for %like%## Warning: package 'data.table' was built under R version 4.1.3library(dplyr) #for %>% ?## Warning: package 'dplyr' was built under R version 4.1.3##
## Attaching package: 'dplyr'## The following objects are masked from 'package:data.table':
##
## between, first, last## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(ggplot2)## Warning: package 'ggplot2' was built under R version 4.1.3#install.packages("ggalt")
library(ggalt) #for geom_dumbbell## Warning: package 'ggalt' was built under R version 4.1.3## Registered S3 methods overwritten by 'ggalt':
## method from
## grid.draw.absoluteGrob ggplot2
## grobHeight.absoluteGrob ggplot2
## grobWidth.absoluteGrob ggplot2
## grobX.absoluteGrob ggplot2
## grobY.absoluteGrob ggplot2library(likert)## Loading required package: xtable##
## Attaching package: 'likert'## The following object is masked from 'package:dplyr':
##
## recode#install.packages("writexl")
library(writexl)## Warning: package 'writexl' was built under R version 4.1.3#########################
# Wrangle Dumbbell Data #
#########################
#slice Loop & Merge dataset by sessions for dumbbell
ksa01.df <- raw.df[raw.df$Q3 %like% "1",c(21,22:25)] #will also catch 10,11,12,14
ksa02.df <- raw.df[raw.df$Q3 %like% "2",c(21,90:93)] #will also catch 12
ksa03.df <- raw.df[raw.df$Q3 %like% "3",c(21,158:161)]
ksa04.df <- raw.df[raw.df$Q3 %like% "4",c(21,226:229)] #will also catch 14
ksa05.df <- raw.df[raw.df$Q3 %like% "5",c(21,294:297)]
ksa06.df <- raw.df[raw.df$Q3 %like% "6",c(21,362:365)]
ksa07.df <- raw.df[raw.df$Q3 %like% "7",c(21,430:433)]
ksa08.df <- raw.df[raw.df$Q3 %like% "8",c(21,498:501)]
ksa09.df <- raw.df[raw.df$Q3 %like% "9",c(21,566:569)]
ksa10.df <- raw.df[raw.df$Q3 %like% "10",c(21,634:637)]
ksa11.df <- raw.df[raw.df$Q3 %like% "11",c(21,702:705)]
ksaws.df <- raw.df[raw.df$Q3 %like% "12",c(21,770:773)]
#rename the session no. for CATA obs & listwise delete missingness
ksa01.df$Q3[ksa01.df$Q3 != "1"] <- "1" #this is key for the CATA
ksa01.df[ksa01.df==""] <- NA
ksa01.df <- na.omit(ksa01.df) #listwise delete missingness, else can't compute means
ksa02.df$Q3[ksa02.df$Q3 != "2"] <- "2"
ksa02.df[ksa02.df==""] <- NA
ksa02.df <- na.omit(ksa02.df)
ksa03.df$Q3[ksa03.df$Q3 != "3"] <- "3"
ksa03.df[ksa03.df==""] <- NA
ksa03.df <- na.omit(ksa03.df)
ksa04.df$Q3[ksa04.df$Q3 != "4"] <- "4"
ksa04.df[ksa04.df==""] <- NA
ksa04.df <- na.omit(ksa04.df) #listwise delete missingness, else can't compute means
ksa05.df$Q3[ksa05.df$Q3 != "5"] <- "5"
ksa05.df[ksa05.df==""] <- NA
ksa05.df <- na.omit(ksa05.df)
ksa06.df$Q3[ksa06.df$Q3 != "6"] <- "6"
ksa06.df[ksa06.df==""] <- NA
ksa06.df <- na.omit(ksa06.df) #listwise delete missingness, else can't compute means
ksa07.df$Q3[ksa07.df$Q3 != "7"] <- "7"
ksa07.df[ksa07.df==""] <- NA
ksa07.df <- na.omit(ksa07.df)
ksa08.df$Q3[ksa08.df$Q3 != "8"] <- "8"
ksa08.df[ksa08.df==""] <- NA
ksa08.df <- na.omit(ksa08.df)
ksa09.df$Q3[ksa09.df$Q3 != "9"] <- "9"
ksa09.df[ksa09.df==""] <- NA
ksa09.df <- na.omit(ksa09.df)
ksa10.df$Q3[ksa10.df$Q3 != "10"] <- "10"
ksa10.df[ksa10.df==""] <- NA
ksa10.df <- na.omit(ksa10.df) #listwise delete missingness, else can't compute means
ksa11.df$Q3[ksa11.df$Q3 != "11"] <- "11"
ksa11.df[ksa11.df==""] <- NA
ksa11.df <- na.omit(ksa11.df)
ksaws.df$Q3[ksaws.df$Q3 != "12"] <- "12"
ksaws.df[ksaws.df==""] <- NA
ksaws.df <- na.omit(ksaws.df)
# rename all the columns in these dumbbell dataframes
rename_ksa <- lapply(list(ksa01.df,
ksa02.df,
ksa03.df,
ksa04.df,
ksa05.df,
ksa06.df,
ksa07.df,
ksa08.df,
ksa09.df,
ksa10.df,
ksa11.df,
ksaws.df),function(x) {
names(x)<- c("session",
"prek","prea","postk","posta")
x})
names(rename_ksa) <- c("ksa01.df",
"ksa02.df",
"ksa03.df",
"ksa04.df",
"ksa05.df",
"ksa06.df",
"ksa07.df",
"ksa08.df",
"ksa09.df",
"ksa10.df",
"ksa11.df",
"ksaws.df")
list2env(rename_ksa, envir = .GlobalEnv)## <environment: R_GlobalEnv>#bind the dumbbell dataframes into one
ksa.df <- rbind(ksa01.df,
ksa02.df,
ksa03.df,
ksa04.df,
ksa05.df,
ksa06.df,
ksa07.df,
ksa08.df,
ksa09.df,
ksa10.df,
ksa11.df,
ksaws.df)
#convert session name to factor
ksa.df$session <- factor(ksa.df$session,
levels = c(1,2,3,4,5,6,7,8,9,10,11,12),
labels = c("Concept of Trust",
"Structural Racism",
"Participatory Budgeting",
"Social Determinants of Health",
"Data is More than Numbers",
"Equitable Partnerships",
"Applying an Equity Lens",
"Race & Racism in Healthcare",
"Intervention Mapping",
"Culturally Responsive Evaluation",
"Co-creating Data Visualization",
"Co-Lab(orative) Learning Workshop"))
#convert pre/post scores to numeric type
ksa.df$prek <- as.numeric(as.character(ksa.df$prek))
ksa.df$prea <- as.numeric(as.character(ksa.df$prea))
ksa.df$postk <- as.numeric(as.character(ksa.df$postk))
ksa.df$posta <- as.numeric(as.character(ksa.df$posta))
#create a knowledge data frame with generated average scores for dumbbells
ksa.df %>%
group_by(session) %>%
summarise_at(vars("prek","postk"), mean)-> prepostk.df
prepostk.df$id <- "knowledge" #this is key for aggregated dataset
names(prepostk.df)<- c("session", "pre","post","id")
#create an ability data frame with generated average scores for dumbbells
ksa.df %>%
group_by(session) %>%
summarise_at(vars("prea","posta"), mean)-> preposta.df
preposta.df$id <- "ability" #this is key for aggregated dataset
names(preposta.df)<- c("session", "pre","post","id")
#create one data frame for knowledge & ability by session: https://r-graph-gallery.com/web-extended-dumbbell-plot-ggplot2.html
prepost.df <- rbind(prepostk.df,preposta.df)
#########################
# Wrangle Lollipop Data #
#########################
overall01.df <- raw.df[raw.df$Q3 %like% "1",c(21,70)] #will also catch 10,11,12,14
overall02.df <- raw.df[raw.df$Q3 %like% "2",c(21,134)] #will also catch 12
overall03.df <- raw.df[raw.df$Q3 %like% "3",c(21,198)]
overall04.df <- raw.df[raw.df$Q3 %like% "4",c(21,262)] #will also catch 14
overall05.df <- raw.df[raw.df$Q3 %like% "5",c(21,326)]
overall06.df <- raw.df[raw.df$Q3 %like% "6",c(21,390)]
overall07.df <- raw.df[raw.df$Q3 %like% "7",c(21,464)]
overall08.df <- raw.df[raw.df$Q3 %like% "8",c(21,518)]
overall09.df <- raw.df[raw.df$Q3 %like% "9",c(21,582)]
overall10.df <- raw.df[raw.df$Q3 %like% "10",c(21,646)]
overall11.df <- raw.df[raw.df$Q3 %like% "11",c(21,710)]
overallws.df <- raw.df[raw.df$Q3 %like% "12",c(21,774)]
#rename the session no. for CATA obs & listwise delete missingness
overall01.df$Q3[overall01.df$Q3 != "1"] <- "1" #this is key for the CATA
overall01.df[overall01.df==""] <- NA
overall01.df <- na.omit(overall01.df)
overall02.df$Q3[overall02.df$Q3 != "2"] <- "2"
overall02.df[overall02.df==""] <- NA
overall02.df <- na.omit(overall02.df)
overall03.df$Q3[overall03.df$Q3 != "3"] <- "3"
overall03.df[overall03.df==""] <- NA
overall03.df <- na.omit(overall03.df)
overall04.df$Q3[overall04.df$Q3 != "4"] <- "4"
overall04.df[overall04.df==""] <- NA
overall04.df <- na.omit(overall04.df)
overall05.df$Q3[overall05.df$Q3 != "5"] <- "5"
overall05.df[overall05.df==""] <- NA
overall05.df <- na.omit(overall05.df)
overall06.df$Q3[overall06.df$Q3 != "6"] <- "6"
overall06.df[overall06.df==""] <- NA
overall06.df <- na.omit(overall06.df)
overall07.df$Q3[overall07.df$Q3 != "7"] <- "7"
overall07.df[overall07.df==""] <- NA
overall07.df <- na.omit(overall07.df)
overall08.df$Q3[overall08.df$Q3 != "8"] <- "8"
overall08.df[overall08.df==""] <- NA
overall08.df <- na.omit(overall08.df)
overall09.df$Q3[overall09.df$Q3 != "9"] <- "9"
overall09.df[overall09.df==""] <- NA
overall09.df <- na.omit(overall09.df)
overall10.df$Q3[overall10.df$Q3 != "10"] <- "10"
overall10.df[overall10.df==""] <- NA
overall10.df <- na.omit(overall10.df)
overall11.df$Q3[overall11.df$Q3 != "11"] <- "11"
overall11.df[overall11.df==""] <- NA
overall11.df <- na.omit(overall11.df)
overallws.df$Q3[overallws.df$Q3 != "12"] <- "12"
overallws.df[overallws.df==""] <- NA
overallws.df <- na.omit(overallws.df)
# rename all the columns in these dumbbell dataframes
rename_overall <- lapply(list(overall01.df,
overall02.df,
overall03.df,
overall04.df,
overall05.df,
overall06.df,
overall07.df,
overall08.df,
overall09.df,
overall10.df,
overall11.df,
overallws.df),function(x) {
names(x)<- c("session","rating")
x})
names(rename_overall) <- c("overall01.df",
"overall02.df",
"overall03.df",
"overall04.df",
"overall05.df",
"overall06.df",
"overall07.df",
"overall08.df",
"overall09.df",
"overall10.df",
"overall11.df",
"overallws.df")
list2env(rename_overall, envir = .GlobalEnv)## <environment: R_GlobalEnv>#bind the overall dataframes into one
overall.df <- rbind(overall01.df,
overall02.df,
overall03.df,
overall04.df,
overall05.df,
overall06.df,
overall07.df,
overall08.df,
overall09.df,
overall10.df,
overall11.df,
overallws.df)
#convert session name to factor
overall.df$session <- factor(overall.df$session,
levels = c(1,2,3,4,5,6,7,8,9,10,11,12),
labels = c("Concept of Trust",
"Structural Racism",
"Participatory Budgeting",
"Social Determinants of Health",
"Data is More than Numbers",
"Equitable Partnerships",
"Applying an Equity Lens",
"Race & Racism in Healthcare",
"Intervention Mapping",
"Culturally Responsive Evaluation",
"Co-creating Data Visualization",
"Co-Lab(orative) Learning Workshop"))
#convert overall rating score to numeric type
overall.df$rating <- as.numeric(as.character(overall.df$rating))
#generate average scores for dumbbells: overall quality for all sessions
overall.df %>%
group_by(session) %>%
summarise_at(vars("rating"), mean)-> overall.df
#generate average scores for overall Q
overalltot <- mean(overall.df$rating)
#######################
# Wrangle Likert Data #
#######################
#subset Likert scale items by session
likert01.df <- raw.df[raw.df$Q3 %like% "1",c(21,71:74)] #will also catch 10,11,12,14
likert02.df <- raw.df[raw.df$Q3 %like% "2",c(21,135:138)] #will also catch 12
likert03.df <- raw.df[raw.df$Q3 %like% "3",c(21,199:202)]
likert04.df <- raw.df[raw.df$Q3 %like% "4",c(21,263:266)] #will also catch 14
likert05.df <- raw.df[raw.df$Q3 %like% "5",c(21,327:330)]
likert06.df <- raw.df[raw.df$Q3 %like% "6",c(21,391:394)]
likert07.df <- raw.df[raw.df$Q3 %like% "7",c(21,455:458)]
likert08.df <- raw.df[raw.df$Q3 %like% "8",c(21,519:522)]
likert09.df <- raw.df[raw.df$Q3 %like% "9",c(21,583:586)]
likert10.df <- raw.df[raw.df$Q3 %like% "10",c(21,647:650)]
likert11.df <- raw.df[raw.df$Q3 %like% "11",c(21,711:714)]
likertws.df <- raw.df[raw.df$Q3 %like% "12",c(21,779:783)]
#rename the Likert session no. for CATA obs & listwise delete missingness
likert01.df$Q3[likert01.df$Q3 != "1"] <- "1" #this is key for CATA
likert01.df[likert01.df==""] <- NA
likert01.df <- na.omit(likert01.df)
likert02.df$Q3[likert02.df$Q3 != "2"] <- "2"
likert02.df[likert02.df==""] <- NA
likert02.df <- na.omit(likert02.df)
likert03.df$Q3[likert03.df$Q3 != "3"] <- "3"
likert03.df[likert03.df==""] <- NA
likert03.df <- na.omit(likert03.df)
likert04.df$Q3[likert04.df$Q3 != "4"] <- "4"
likert04.df[likert04.df==""] <- NA
likert04.df <- na.omit(likert04.df)
likert05.df$Q3[likert05.df$Q3 != "5"] <- "5"
likert05.df[likert05.df==""] <- NA
likert05.df <- na.omit(likert05.df)
likert06.df$Q3[likert06.df$Q3 != "6"] <- "6"
likert06.df[likert06.df==""] <- NA
likert06.df <- na.omit(likert06.df)
likert07.df$Q3[likert07.df$Q3 != "7"] <- "7"
likert07.df[likert07.df==""] <- NA
likert07.df <- na.omit(likert07.df)
likert08.df$Q3[likert08.df$Q3 != "8"] <- "8"
likert08.df[likert08.df==""] <- NA
likert08.df <- na.omit(likert08.df)
likert09.df$Q3[likert09.df$Q3 != "9"] <- "9"
likert09.df[likert09.df==""] <- NA
likert09.df <- na.omit(likert09.df)
likert10.df$Q3[likert10.df$Q3 != "10"] <- "10"
likert10.df[likert10.df==""] <- NA
likert10.df <- na.omit(likert10.df)
likert11.df$Q3[likert11.df$Q3 != "11"] <- "11"
likert11.df[likert11.df==""] <- NA
likert11.df <- na.omit(likert11.df)
likertws.df$Q3[likertws.df$Q3 != "12"] <- "12"
likertws.df[likertws.df==""] <- NA
likertws.df <- na.omit(likertws.df)
# rename all the columns in these Likert dataframes
rename_likert <- lapply(list(likert01.df,
likert02.df,
likert03.df,
likert04.df,
likert05.df,
likert06.df,
likert07.df,
likert08.df,
likert09.df,
likert10.df,
likert11.df),function(x) {
names(x)<- c("session",
"effective presenter",
"knowledgeable presenter",
"insightful information",
"relevant techniques")
x})
names(rename_likert) <- c("likert01.df",
"likert02.df",
"likert03.df",
"likert04.df",
"likert05.df",
"likert06.df",
"likert07.df",
"likert08.df",
"likert09.df",
"likert10.df",
"likert11.df")
list2env(rename_likert, envir = .GlobalEnv)## <environment: R_GlobalEnv>names(likertws.df) <- c("session",
"effective guidance",
"welcoming space",
"structured activities",
"insightful information",
"relevant techniques")
#bind the Days 1-2 session dataframes into one
likert.df <- rbind(likert01.df,
likert02.df,
likert03.df,
likert04.df,
likert05.df,
likert06.df,
likert07.df,
likert08.df,
likert09.df,
likert10.df,
likert11.df)
#convert variables to factor
likert.df$session <- factor(likert.df$session,
levels = c(1,2,3,4,5,6,7,8,9,10,11),
labels = c("Concept of Trust",
"Structural Racism",
"Participatory Budgeting",
"Social Determinants of Health",
"Data is More than Numbers",
"Equitable Partnerships",
"Applying an Equity Lens",
"Race & Racism in Healthcare",
"Intervention Mapping",
"Culturally Responsive Evaluation",
"Co-creating Data Visualization"))
likertws.df$session <- factor(likertws.df$session,
levels = c(12),
labels = c("Co-Lab(orative) Learning Workshop"))
likertlabels <- c("Strongly Disagree",
"Somewhat Disagree",
"Neither",
"Somewhat Agree",
"Strongly Agree")
likert.df$`effective presenter` <- factor(likert.df$`effective presenter`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likert.df$`knowledgeable presenter` <- factor(likert.df$`knowledgeable presenter`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likert.df$`insightful information` <- factor(likert.df$`insightful information`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likert.df$`relevant techniques` <- factor(likert.df$`relevant techniques`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likertws.df$`effective guidance` <- factor(likertws.df$`effective guidance`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likertws.df$`welcoming space` <- factor(likertws.df$`welcoming space`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likertws.df$`structured activities` <- factor(likertws.df$`structured activities`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likertws.df$`insightful information` <- factor(likertws.df$`insightful information`,
levels = c(1,2,3,4,5),
labels = likertlabels)
likertws.df$`relevant techniques` <- factor(likertws.df$`relevant techniques`,
levels = c(1,2,3,4,5),
labels = likertlabels)
#######################
# Wrangle Likert Data #
#######################
#subset likelihood items by session
likely01.df <- raw.df[raw.df$Q3 %like% "1",c(21,80:81)] #will also catch 10,11,12,14
likely02.df <- raw.df[raw.df$Q3 %like% "2",c(21,144:145)] #will also catch 12
likely03.df <- raw.df[raw.df$Q3 %like% "3",c(21,208:209)]
likely04.df <- raw.df[raw.df$Q3 %like% "4",c(21,272:273)] #will also catch 14
likely05.df <- raw.df[raw.df$Q3 %like% "5",c(21,336:337)]
likely06.df <- raw.df[raw.df$Q3 %like% "6",c(21,400:401)]
likely07.df <- raw.df[raw.df$Q3 %like% "7",c(21,464:465)]
likely08.df <- raw.df[raw.df$Q3 %like% "8",c(21,528:529)]
likely09.df <- raw.df[raw.df$Q3 %like% "9",c(21,592:593)]
likely10.df <- raw.df[raw.df$Q3 %like% "10",c(21,656:657)]
likely11.df <- raw.df[raw.df$Q3 %like% "11",c(21,720:721)]
likelyws.df <- raw.df[raw.df$Q3 %like% "12",c(21,784:785)]
#rename the likely session no. for CATA obs & listwise delete missingness
likely01.df$Q3[likely01.df$Q3 != "1"] <- "1" #this is key for CATA
likely01.df[likely01.df==""] <- NA
likely01.df <- na.omit(likely01.df)
likely02.df$Q3[likely02.df$Q3 != "2"] <- "2"
likely02.df[likely02.df==""] <- NA
likely02.df <- na.omit(likely02.df)
likely03.df$Q3[likely03.df$Q3 != "3"] <- "3"
likely03.df[likely03.df==""] <- NA
likely03.df <- na.omit(likely03.df)
likely04.df$Q3[likely04.df$Q3 != "4"] <- "4"
likely04.df[likely04.df==""] <- NA
likely04.df <- na.omit(likely04.df)
likely05.df$Q3[likely05.df$Q3 != "5"] <- "5"
likely05.df[likely05.df==""] <- NA
likely05.df <- na.omit(likely05.df)
likely06.df$Q3[likely06.df$Q3 != "6"] <- "6"
likely06.df[likely06.df==""] <- NA
likely06.df <- na.omit(likely06.df)
likely07.df$Q3[likely07.df$Q3 != "7"] <- "7"
likely07.df[likely07.df==""] <- NA
likely07.df <- na.omit(likely07.df)
likely08.df$Q3[likely08.df$Q3 != "8"] <- "8"
likely08.df[likely08.df==""] <- NA
likely08.df <- na.omit(likely08.df)
likely09.df$Q3[likely09.df$Q3 != "9"] <- "9"
likely09.df[likely09.df==""] <- NA
likely09.df <- na.omit(likely09.df)
likely10.df$Q3[likely10.df$Q3 != "10"] <- "10"
likely10.df[likely10.df==""] <- NA
likely10.df <- na.omit(likely10.df)
likely11.df$Q3[likely11.df$Q3 != "11"] <- "11"
likely11.df[likely11.df==""] <- NA
likely11.df <- na.omit(likely11.df)
likelyws.df$Q3[likelyws.df$Q3 != "12"] <- "12"
likelyws.df[likelyws.df==""] <- NA
likelyws.df <- na.omit(likelyws.df)
# rename all the columns in these likelihood dataframes
rename_likely <- lapply(list(likely01.df,
likely02.df,
likely03.df,
likely04.df,
likely05.df,
likely06.df,
likely07.df,
likely08.df,
likely09.df,
likely10.df,
likely11.df,
likelyws.df),function(x) {
names(x)<- c("session",
"use what you learned",
"recommend this session to others")
x})
names(rename_likely) <- c("likely01.df",
"likely02.df",
"likely03.df",
"likely04.df",
"likely05.df",
"likely06.df",
"likely07.df",
"likely08.df",
"likely09.df",
"likely10.df",
"likely11.df",
"likelyws.df")
list2env(rename_likely, envir = .GlobalEnv)## <environment: R_GlobalEnv>#bind the likelihood dataframes into one
likely.df <- rbind(likely01.df,
likely02.df,
likely03.df,
likely04.df,
likely05.df,
likely06.df,
likely07.df,
likely08.df,
likely09.df,
likely10.df,
likely11.df,
likelyws.df)
#convert likelihood variables to factor
likely.df$session <- factor(likely.df$session,
levels = c(1,2,3,4,5,6,7,8,9,10,11,12),
labels = c("Concept of Trust",
"Structural Racism",
"Participatory Budgeting",
"Social Determinants of Health",
"Data is More than Numbers",
"Equitable Partnerships",
"Applying an Equity Lens",
"Race & Racism in Healthcare",
"Intervention Mapping",
"Culturally Responsive Evaluation",
"Co-creating Data Visualization",
"Co-Lab(orative) Learning Workshop"))
likelylabels <- c("Not at all Likely",
"Somewhat Unlikely",
"Neutral",
"Somewhat Likely",
"Extremely Likely")
likely.df$`use what you learned` <- factor(likely.df$`use what you learned`,
levels = c(1,2,3,4,5),
labels = likelylabels)
likely.df$`recommend this session to others` <- factor(likely.df$`recommend this session to others`,
levels = c(1,2,3,4,5),
labels = likelylabels)
####################
# Wrangle Feedback #
####################
feedback01.df <- raw.df[raw.df$Q3 %like% "1",c(19,21,83,85)] #will also catch 10,11,12,14
feedback02.df <- raw.df[raw.df$Q3 %like% "2",c(19,21,147,149)] #will also catch 12
feedback03.df <- raw.df[raw.df$Q3 %like% "3",c(19,21,211,213)]
feedback04.df <- raw.df[raw.df$Q3 %like% "4",c(19,21,275,277)] #will also catch 14
feedback05.df <- raw.df[raw.df$Q3 %like% "5",c(19,21,339,341)]
feedback06.df <- raw.df[raw.df$Q3 %like% "6",c(19,21,402,403,405)]
feedback07.df <- raw.df[raw.df$Q3 %like% "7",c(19,21,467,469)]
feedback08.df <- raw.df[raw.df$Q3 %like% "8",c(19,21,531,533)]
feedback09.df <- raw.df[raw.df$Q3 %like% "9",c(19,21,595,597)]
feedback10.df <- raw.df[raw.df$Q3 %like% "10",c(19,21,659,661)]
feedback11.df <- raw.df[raw.df$Q3 %like% "11",c(19,21,723,725)]
feedbackws.df <- raw.df[raw.df$Q3 %like% "12",c(19,21,788,789)]
feedbackheri.df <- raw.df[raw.df$Q3 %like% "14", c(19,21,854:856)]
#rename the likelihood session no. for CATA obs & listwise delete missingness
feedback01.df$Q3[feedback01.df$Q3 != "1"] <- "1" #this is key for CATA
feedback01.df[feedback01.df==""] <- NA
feedback02.df$Q3[feedback02.df$Q3 != "2"] <- "2"
feedback02.df[feedback02.df==""] <- NA
feedback03.df$Q3[feedback03.df$Q3 != "3"] <- "3"
feedback03.df[feedback03.df==""] <- NA
feedback04.df$Q3[feedback04.df$Q3 != "4"] <- "4"
feedback04.df[feedback04.df==""] <- NA
feedback05.df$Q3[feedback05.df$Q3 != "5"] <- "5"
feedback05.df[feedback05.df==""] <- NA
feedback06.df$Q3[feedback06.df$Q3 != "6"] <- "6"
feedback06.df[feedback06.df==""] <- NA
feedback07.df$Q3[feedback07.df$Q3 != "7"] <- "7"
feedback07.df[feedback07.df==""] <- NA
feedback08.df$Q3[feedback08.df$Q3 != "8"] <- "8"
feedback08.df[feedback08.df==""] <- NA
feedback09.df$Q3[feedback09.df$Q3 != "9"] <- "9"
feedback09.df[feedback09.df==""] <- NA
feedback10.df$Q3[feedback10.df$Q3 != "10"] <- "10"
feedback10.df[feedback10.df==""] <- NA
feedback11.df$Q3[feedback11.df$Q3 != "11"] <- "11"
feedback11.df[feedback11.df==""] <- NA
feedbackws.df$Q3[feedbackws.df$Q3 != "12"] <- "12"
feedbackws.df[feedbackws.df==""] <- NA
feedbackheri.df$Q3[feedbackheri.df$Q3 != "14"] <- "14"
feedbackheri.df[feedbackheri.df==""] <- NA
# rename all the columns in these Likert dataframes
rename_feedback <- lapply(list(feedback01.df,
feedback02.df,
feedback03.df,
feedback04.df,
feedback05.df,
# feedback06.df,
feedback07.df,
feedback08.df,
feedback09.df,
feedback10.df,
feedback11.df,
feedbackws.df),function(x) {
names(x)<- c("role","session",
"session improvements",
"session relevance")
x})
names(rename_feedback) <- c("feedback01.df",
"feedback02.df",
"feedback03.df",
"feedback04.df",
"feedback05.df",
# "feedback06.df",
"feedback07.df",
"feedback08.df",
"feedback09.df",
"feedback10.df",
"feedback11.df",
"feedbackws.df")
list2env(rename_feedback, envir = .GlobalEnv)## <environment: R_GlobalEnv>names(feedback06.df) <- c("role","session",
"use in work",
"session improvements",
"session relevance")
names(feedbackheri.df) <- c("role","session",
"networking",
"event improvements",
"event successes")
#bind the dataframes into one
feedback.df <- rbind(feedback01.df,
feedback02.df,
feedback03.df,
feedback04.df,
feedback05.df,
# feedback06.df,
feedback07.df,
feedback08.df,
feedback09.df,
feedback10.df,
feedback11.df,
feedbackws.df)
#convert feedback variables to factor
feedback.df$session <- factor(feedback.df$session,
levels = c(1,2,3,4,5,6,7,8,9,10,11,12),
labels = c("Concept of Trust",
"Structural Racism",
"Participatory Budgeting",
"Social Determinants of Health",
"Data is More than Numbers",
"Equitable Partnerships",
"Applying an Equity Lens",
"Race & Racism in Healthcare",
"Intervention Mapping",
"Culturally Responsive Evaluation",
"Co-creating Data Visualization",
"Co-Lab(orative) Learning Workshop"))
feedback06.df$session <- factor(feedback06.df$session,
levels = c(6),
labels = c("Equitable Partnerships"))
feedbackheri.df$session <- factor(feedbackheri.df$session,
levels = c(14),
labels = c("HERI"))
feedback.df$role <- factor(feedback.df$role,
levels = c(1,2,7,5,4,6),
labels = c("Student","Faculty","Research Staff","Corporate/Industry","Community Member","Other"))
#write_xlsx(feedback.df,"C:\\Users\\josep\\Documents\\HERI\\feedback-mmdd.xlsx") #create a spreadsheet out of the dataframe; will stop knitting
#write_xlsx(feedback06.df,"C:\\Users\\josep\\Documents\\HERI\\feedbackS6-mmdd.xlsx")
#write_xlsx(feedbackheri.df,"C:\\Users\\josep\\Documents\\HERI\\feedbackheri-mmdd.xlsx")
#####################
# Dumbbell Settings #
#####################
# Version 1: https://r-graph-gallery.com/web-extended-dumbbell-plot-ggplot2.html (too many dots from full dataset)
# Version 2: from IIN (just dots plus legend with generated averages)
# Version 3: https://datavizpyr.com/dumbbell-plot-in-r-with-ggplot2/(did not print)
# Version 4: https://www.geeksforgeeks.org/how-to-make-dumbbell-plot-in-r-with-ggplot2/ (works with generated averages, sorted by session name)
# Version 5: https://www.youtube.com/watch?v=USNOB-5ou8k&ab_channel=BusinessScience at 2:1 (works with generated averages, plus guiding lines, sorted by greatest difference)
# Version 6: https://www.connorrothschild.com/post/dumbbell-plots/ with difference scores (works but can't get difference scores to print)
gold <- "#d8b365"
teal <- "#5ab4ab"
xsubtitle <- "Average Ratings (out of 5)"
caption <- "Source: 2022 HERI Evaluation Survey"
###########################
# K for all sessions #
###########################
prepost.df [prepost.df$id=="knowledge",] %>% #runs code if these cells have a certain value
mutate(difference = (post-pre), #generates a difference score
position = post - (0.5 * difference)) %>% #half-way between pre and post, from the post dot
ggplot() +
aes(x = pre,
xend = post,
# y = session) + #default: sorted by session name
y = reorder(session, pre), group = session) + #sorted by greatest pre score; could replace with difference
geom_dumbbell(
size_x = 5, colour_x = gold, #orange pre dot
size_xend = 5, colour_xend = teal, #green post dot
size = 1, #black line
dot_guide = TRUE,
dot_guide_size = 0.15,
dot_guide_colour = "grey60") +
scale_x_continuous(breaks = seq (1,5,1),limits = c(1,5)) + #show x axis from 1 to 5 in intervals of 1, even if no data
geom_text(data=filter(prepost.df, session=="Race & Racism in Healthcare"),
aes(x=pre, y=session, label = "Before"),
color = gold, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(data=filter(prepost.df, session=="Race & Racism in Healthcare"),
aes(x=post, y=session, label = "After"),
color = teal, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(aes(x=pre,y=session,label = round(pre, digits = 1)), color = "white", size = 2) + #insert pre-score in dot, rounding to 1 decimal place
geom_text(aes(x=post,y=session,label = round(post, digits = 1)), color = "white", size = 2) + #insert post-score in dot, rounding to 1 decimal place
geom_text(aes(x=position,y=session,label = round(difference, digits = 1)), nudge_y = 0.25, size = 1.5)+ #insert difference over line
labs(title = "Knowledge Ratings Before & After Each Session", y = NULL, x = xsubtitle, caption = caption) +
theme(panel.border = element_blank(), #no gap
panel.background = element_rect(fill = 'transparent',color = NA), #transparent bkgd
panel.grid.minor = element_blank(),
panel.grid.major.y = element_blank(),
panel.grid.major.x = element_line(),
axis.ticks = element_blank()) #no ticks## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database
ggsave("Session All - Knowledge.png", height = 4, width = 7)## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database######################
# A for all sessions #
######################
prepost.df [prepost.df$id == "ability",] %>%
mutate(difference = (post-pre),
position = post - (0.5 * difference)) %>%
ggplot() +
aes(x = pre,
xend = post,
# y = session) + #default: sorted by session name
y = reorder(session, pre), group = session) +
geom_dumbbell(
size_x = 5, colour_x = gold,
size_xend = 5, colour_xend = teal,
size = 1,
dot_guide = TRUE,
dot_guide_size = 0.15,
dot_guide_colour = "grey60") +
scale_x_continuous(breaks = seq (1,5,1),limits = c(1,5)) +
geom_text(data=filter(prepost.df, session=="Race & Racism in Healthcare"),
aes(x=pre, y=session, label = "Before"),
color = gold, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(data=filter(prepost.df, session=="Race & Racism in Healthcare"),
aes(x=post, y=session, label = "After"),
color = teal, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(aes(x=pre,y=session,label = round(pre, digits = 1)), color = "white", size = 2) +
geom_text(aes(x=post,y=session,label = round(post, digits = 1)), color = "white", size = 2) +
geom_text(aes(x=position,y=session,label = round(difference, digits = 1)), nudge_y = 0.25, size = 1.5)+
labs(title = "Ability Ratings Before & After Each Session", y = NULL, x = xsubtitle, caption = caption) +
theme(panel.border = element_blank(),
panel.background = element_rect(fill = 'transparent',color = NA),
panel.grid.minor = element_blank(),
panel.grid.major.y = element_blank(),
panel.grid.major.x = element_line(),
axis.ticks = element_blank()) ## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database
ggsave("Session All - Ability.png", height = 4, width = 7)## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database############################
# Overall for all sessions #
############################
overall.df %>%
ggplot() +
aes(x = rating,
xend = rating,
y = reorder(session, rating), group = session) +
geom_dumbbell(
size_x = 5, colour_x = gold,
size_xend = 5, colour_xend = teal,
size = 1,
dot_guide = TRUE,
dot_guide_size = 0.2,
dot_guide_colour = "black") +
scale_x_continuous(breaks = seq (1,5,1),limits = c(1,5)) +
geom_text(aes(label = round(rating, digits = 1)), color = "white", size = 2) + #insert rating in lollipop, rounding to 1 decimal place
labs(title = "Overall Ratings for Each Session", y = NULL, x = "average ratings (from very poor to excellent)",
caption = caption) +
theme(panel.border = element_blank(),
panel.background = element_rect(fill = 'transparent',color = NA),
panel.grid.minor = element_blank(),
panel.grid.major.y = element_blank(),
panel.grid.major.x = element_line(),
axis.ticks = element_blank()) 
ggsave("Session All - Overall Rating.png", height = 3.5, width = 7)
#########################################
# Session Settings: Dumbbell AND Likert #
#########################################
# CHANGE thissession, GGSAVEs; RERUN for each session
thissession <- "Concept of Trust"
#thissession <- "Structural Racism"
#thissession <- "Participatory Budgeting"
#thissession <- "Social Determinants of Health"
#thissession <- "Data is More than Numbers"
#thissession <- "Equitable Partnerships"
#thissession <- "Applying an Equity Lens"
#thissession <- "Race & Racism in Healthcare"
#thissession <- "Intervention Mapping"
#thissession <- "Culturally Responsive Evaluation"
#thissession <- "Co-creating Data Visualization"
#thissession <- "Co-Lab(orative) Learning Workshop"
######################
# K and A by session #
######################
prepost.df [prepost.df$session==thissession,] %>%
ggplot() +
aes(x = pre,
xend = post,
y = id) +
geom_dumbbell(
size_x = 2, colour_x = gold,
size_xend = 2, colour_xend = teal,
size = 1,
dot_guide = TRUE,
dot_guide_size = 0.15,
dot_guide_colour = "grey60") +
scale_x_continuous(breaks = seq (1,5,1),limits = c(1,5)) +
#geom_text(data=filter(prepost.df, id=="knowledge"),
# aes(x=pre, y=session, label="Before"),
# color = gold, size=2, vjust=-1.5, fontface="bold",family="Lato") +
#geom_text(data=filter(prepost.df, id=="knowledge"),
# aes(x=post, y=session, label = "After"),
# color = teal, size=2, vjust=-1.5, fontface="bold",family="Lato") +
labs(title = thissession, y = NULL, x = xsubtitle,
subtitle = "Changes in knowledge & ability ratings",
caption = caption) +
theme(
panel.grid.minor = element_blank(),
panel.grid.major.y = element_blank(),
panel.grid.major.x = element_line(),
axis.ticks = element_blank(),
panel.border = element_blank(),
panel.background = element_rect(fill = 'transparent',color = NA)
)
ggsave("Session ## - K and A.png", height = 1.5, width = 5)
#####################
# Likert by session #
#####################
quant <- likert(likert.df[likert.df$session==thissession,2:5])
plot(quant, positive.order = TRUE) +
labs(title = thissession)
ggsave("Session ## - Quant.png", width = 9, height = 2.5)
#quantws <- likert(likertws.df[likertws.df$session==thissession,2:6])
#plot(quantws, positive.order = TRUE) +
# labs(title = thissession)
#ggsave("Session 12 - Quant.png", width = 9, height = 2.5)
#########################
# Likelihood by session #
#########################
likely <- likert(likely.df[likely.df$session==thissession,2:3])
plot(likely, positive.order = TRUE) +
labs(title = thissession, subtitle = "How likely are you to...")
ggsave("Session ## - Likely.png", width = 9, height = 2)