CHER’s 2023 Health Equity Research Intensive
Josephine McKelvy
2023-10-11
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. Create New Project > Version Control > repo URL (github.com/josephine-mckelvy/repo-name).
# Qualtrics: Data & Analysis/Export & Import/Export Data/CSV/Use Numeric values. Extract files.
# Qualtrics codebook: Survey/Tools/Export Survey to Word (show logic)
# 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("2023 Health Equity Research Intensive_October 18, 2023_10.51.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%
library(dplyr) #for %>% ?##
## 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)
#install.packages("ggalt")
library(ggalt) #for geom_dumbbell## 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)
library(tidyr)
#########################
# Wrangle Dumbbell Data #
#########################
#slice Loop & Merge dataset by sessions for dumbbell; double-check the column #s for the pre/post questions
ksa01.df <- raw.df[raw.df$Q3 %like% "1",c(21, 22:25)] #will also catch sessions 10,11,12,14
ksa02.df <- raw.df[raw.df$Q3 %like% "2",c(21,242:245)] #will also catch session 12
ksa03.df <- raw.df[raw.df$Q3 %like% "3",c(21,138:141)]
ksa04.df <- raw.df[raw.df$Q3 %like% "4",c(21,196:199)] #will also catch session 14
ksa05.df <- raw.df[raw.df$Q3 %like% "5",c(21, 92:95)]
ksa06.df <- raw.df[raw.df$Q3 %like% "6",c(21,312:315)]
ksa07.df <- raw.df[raw.df$Q3 %like% "7",c(21,370:373)]
ksa08.df <- raw.df[raw.df$Q3 %like% "8",c(21,428:431)]
ksa09.df <- raw.df[raw.df$Q3 %like% "9",c(21,486:489)]
ksa10.df <- raw.df[raw.df$Q3 %like% "10",c(21,544:547)]
ksa11.df <- raw.df[raw.df$Q3 %like% "11",c(21,602:605)]
#ksaws.df <- raw.df[raw.df$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)
# 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),function(x) {
names(x)<- c("session",
"pre_k","pre_a","post_k","post_a")
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")
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)
#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),
labels = c("Trust & Trustworthiness",
"Equitable Partnerships",
"Nuts & Bolts",
"Community-Based Participatory Research",
"Participatory Budgeting",
"Role of Research(ers)",
"Communicating Research",
"Photovoice",
"Groundwater Approach",
"Climate Change & Environmental Health",
"Co-Lab(orative) Learning Workshop"))
#convert long-form dataset, combining all pre and all post into new variables, sorted by learning objective, keeping only the pivoted columns
longksa.df <- melt(setDT(ksa.df), measure=patterns("pre","post"), value.name = c("pre","post"), variable.name = "objectives")
longksa.df <- longksa.df[,c("session","pre","post")]
#convert pre/post scores to numeric type
longksa.df$pre <- as.numeric(as.character(longksa.df$pre))
longksa.df$post <- as.numeric(as.character(longksa.df$post))
ksa.df$pre_k <- as.numeric(as.character(ksa.df$pre_k))
ksa.df$pre_a <- as.numeric(as.character(ksa.df$pre_a))
ksa.df$post_k <- as.numeric(as.character(ksa.df$post_k))
ksa.df$post_a <- as.numeric(as.character(ksa.df$post_a))
#create a data frame with generated average scores for dumbbells
longksa.df %>%
group_by(session) %>%
summarise_at(vars("pre","post"), mean)-> prepost_long.df
#create a knowledge data frame with generated average scores for dumbbells
ksa.df %>%
group_by(session) %>%
summarise_at(vars("pre_k","post_k"), mean)-> prepost_k.df
prepost_k.df$id <- "LO1" #this is key for aggregated dataset
names(prepost_k.df)<- c("session", "pre","post","id")
#create an ability data frame with generated average scores for dumbbells
ksa.df %>%
group_by(session) %>%
summarise_at(vars("pre_a","post_a"), mean)-> prepost_a.df
prepost_a.df$id <- "LO2" #this is key for aggregated dataset
names(prepost_a.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(prepost_k.df,prepost_a.df)
###########################
# Wrangle 5 content scale #
###########################
#session content only
ses01.df <- raw.df[raw.df$Q3 %like% "1",c(21, 66:70)] #will also catch sessions 10,11,12,14
ses02.df <- raw.df[raw.df$Q3 %like% "2",c(21,282:286)] #will also catch session 12
ses03.df <- raw.df[raw.df$Q3 %like% "3",c(21,174:178)]
ses04.df <- raw.df[raw.df$Q3 %like% "4",c(21,228:232)] #will also catch session 14
ses05.df <- raw.df[raw.df$Q3 %like% "5",c(21,120:124)]
ses06.df <- raw.df[raw.df$Q3 %like% "6",c(21,336:340)]
ses07.df <- raw.df[raw.df$Q3 %like% "7",c(21,390:394)]
ses08.df <- raw.df[raw.df$Q3 %like% "8",c(21,444:448)]
ses09.df <- raw.df[raw.df$Q3 %like% "9",c(21,498:502)]
ses10.df <- raw.df[raw.df$Q3 %like% "10",c(21,552:556)]
ses11.df <- raw.df[raw.df$Q3 %like% "11",c(21,606:610)]
#rename the session no. for CATA obs & listwise delete missingness
ses01.df$Q3[ses01.df$Q3 != "1"] <- "1" #this is key for the CATA
ses01.df[ses01.df==""] <- NA
ses01.df <- na.omit(ses01.df) #listwise delete missingness, else can't compute means
ses02.df$Q3[ses02.df$Q3 != "2"] <- "2"
ses02.df[ses02.df==""] <- NA
ses02.df <- na.omit(ses02.df)
ses03.df$Q3[ses03.df$Q3 != "3"] <- "3"
ses03.df[ses03.df==""] <- NA
ses03.df <- na.omit(ses03.df)
ses04.df$Q3[ses04.df$Q3 != "4"] <- "4"
ses04.df[ses04.df==""] <- NA
ses04.df <- na.omit(ses04.df) #listwise delete missingness, else can't compute means
ses05.df$Q3[ses05.df$Q3 != "5"] <- "5"
ses05.df[ses05.df==""] <- NA
ses05.df <- na.omit(ses05.df)
ses06.df$Q3[ses06.df$Q3 != "6"] <- "6"
ses06.df[ses06.df==""] <- NA
ses06.df <- na.omit(ses06.df) #listwise delete missingness, else can't compute means
ses07.df$Q3[ses07.df$Q3 != "7"] <- "7"
ses07.df[ses07.df==""] <- NA
ses07.df <- na.omit(ses07.df)
ses08.df$Q3[ses08.df$Q3 != "8"] <- "8"
ses08.df[ses08.df==""] <- NA
ses08.df <- na.omit(ses08.df)
ses09.df$Q3[ses09.df$Q3 != "9"] <- "9"
ses09.df[ses09.df==""] <- NA
ses09.df <- na.omit(ses09.df)
ses10.df$Q3[ses10.df$Q3 != "10"] <- "10"
ses10.df[ses10.df==""] <- NA
ses10.df <- na.omit(ses10.df) #listwise delete missingness, else can't compute means
ses11.df$Q3[ses11.df$Q3 != "11"] <- "11"
ses11.df[ses11.df==""] <- NA
ses11.df <- na.omit(ses11.df)
# rename all the columns in these Likert dataframes
rename_ses <- lapply(list(ses01.df,
ses02.df,
ses03.df,
ses04.df,
ses05.df,
ses06.df,
ses07.df,
ses08.df,
ses09.df,
ses10.df,
ses11.df),function(x) {
names(x)<- c("session",
"length","delivery","participation", "useful","recommend")
x})
names(rename_ses) <- c("ses01.df",
"ses02.df",
"ses03.df",
"ses04.df",
"ses05.df",
"ses06.df",
"ses07.df",
"ses08.df",
"ses09.df",
"ses10.df",
"ses11.df")
list2env(rename_ses, envir = .GlobalEnv)## <environment: R_GlobalEnv>#bind the Days 1-2 session dataframes into one
session.df <- rbind(ses01.df,
ses02.df,
ses03.df,
ses04.df,
ses05.df,
ses06.df,
ses07.df,
ses08.df,
ses09.df,
ses10.df,
ses11.df)
#convert variables to factor
sessionlabels <- c("Trust & Trustworthiness",
"Equitable Partnerships",
"Nuts & Bolts",
"Community-Based Participatory Research",
"Participatory Budgeting",
"Role of Research(ers)",
"Communicating Research",
"Photovoice",
"Groundwater Approach",
"Climate Change & Environmental Health",
"Co-Lab(orative) Learning Workshop")
session.df$session <- factor(session.df$session,
levels = c(1,2,3,4,5,6,7,8,9,10,11),
labels = sessionlabels)
likertlabels <- c("Strongly Disagree",
"Somewhat Disagree",
"Neither",
"Somewhat Agree",
"Strongly Agree")
session.df$`length` <- factor(session.df$`length`,
levels = c(1,2,3,4,5),
labels = likertlabels)
session.df$`delivery` <- factor(session.df$`delivery`,
levels = c(1,2,3,4,5),
labels = likertlabels)
session.df$`participation` <- factor(session.df$`participation`,
levels = c(1,2,3,4,5),
labels = likertlabels)
session.df$`useful` <- factor(session.df$`useful`,
levels = c(1,2,3,4,5),
labels = likertlabels)
session.df$`recommend` <- factor(session.df$`recommend`,
levels = c(1,2,3,4,5),
labels = likertlabels)
###########################
# Wrangle 3 domain scale #
###########################
#session content and HERI domains only
dom01.df <- raw.df[raw.df$Q3 %like% "1",c(21, 71:73)] #will also catch sessions 10,11,12,14
dom02.df <- raw.df[raw.df$Q3 %like% "2",c(21,287:289)] #will also catch session 12
dom03.df <- raw.df[raw.df$Q3 %like% "3",c(21,179:181)]
dom04.df <- raw.df[raw.df$Q3 %like% "4",c(21,233:235)] #will also catch session 14
dom05.df <- raw.df[raw.df$Q3 %like% "5",c(21,125:127)]
dom06.df <- raw.df[raw.df$Q3 %like% "6",c(21,341:343)]
dom07.df <- raw.df[raw.df$Q3 %like% "7",c(21,395:397)]
dom08.df <- raw.df[raw.df$Q3 %like% "8",c(21,449:451)]
dom09.df <- raw.df[raw.df$Q3 %like% "9",c(21,503:505)]
dom10.df <- raw.df[raw.df$Q3 %like% "10",c(21,557:559)]
dom11.df <- raw.df[raw.df$Q3 %like% "11",c(21,611:613)]
#rename the session no. for CATA obs & listwise delete missingness
dom01.df$Q3[dom01.df$Q3 != "1"] <- "1" #this is key for the CATA
dom01.df[dom01.df==""] <- NA
dom01.df <- na.omit(dom01.df) #listwise delete missingness, else can't compute means
dom02.df$Q3[dom02.df$Q3 != "2"] <- "2"
dom02.df[dom02.df==""] <- NA
dom02.df <- na.omit(dom02.df)
dom03.df$Q3[dom03.df$Q3 != "3"] <- "3"
dom03.df[dom03.df==""] <- NA
dom03.df <- na.omit(dom03.df)
dom04.df$Q3[dom04.df$Q3 != "4"] <- "4"
dom04.df[dom04.df==""] <- NA
dom04.df <- na.omit(dom04.df) #listwise delete missingness, else can't compute means
dom05.df$Q3[dom05.df$Q3 != "5"] <- "5"
dom05.df[dom05.df==""] <- NA
dom05.df <- na.omit(dom05.df)
dom06.df$Q3[dom06.df$Q3 != "6"] <- "6"
dom06.df[dom06.df==""] <- NA
dom06.df <- na.omit(dom06.df) #listwise delete missingness, else can't compute means
dom07.df$Q3[dom07.df$Q3 != "7"] <- "7"
dom07.df[dom07.df==""] <- NA
dom07.df <- na.omit(dom07.df)
dom08.df$Q3[dom08.df$Q3 != "8"] <- "8"
dom08.df[dom08.df==""] <- NA
dom08.df <- na.omit(dom08.df)
dom09.df$Q3[dom09.df$Q3 != "9"] <- "9"
dom09.df[dom09.df==""] <- NA
dom09.df <- na.omit(dom09.df)
dom10.df$Q3[dom10.df$Q3 != "10"] <- "10"
dom10.df[dom10.df==""] <- NA
dom10.df <- na.omit(dom10.df) #listwise delete missingness, else can't compute means
dom11.df$Q3[dom11.df$Q3 != "11"] <- "11"
dom11.df[dom11.df==""] <- NA
dom11.df <- na.omit(dom11.df)
# rename all the columns in these Likert dataframes
rename_dom <- lapply(list(dom01.df,
dom02.df,
dom03.df,
dom04.df,
dom05.df,
dom06.df,
dom07.df,
dom08.df,
dom09.df,
dom10.df,
dom11.df),function(x) {
names(x)<- c("session",
"research","implementation","engagement")
x})
names(rename_dom) <- c("dom01.df",
"dom02.df",
"dom03.df",
"dom04.df",
"dom05.df",
"dom06.df",
"dom07.df",
"dom08.df",
"dom09.df",
"dom10.df",
"dom11.df")
list2env(rename_dom, envir = .GlobalEnv)## <environment: R_GlobalEnv>#bind the Days 1-2 session dataframes into one
domain.df <- rbind(dom01.df,
dom02.df,
dom03.df,
dom04.df,
dom05.df,
dom06.df,
dom07.df,
dom08.df,
dom09.df,
dom10.df,
dom11.df)
#convert variables to factor
domain.df$session <- factor(domain.df$session,
levels = c(1,2,3,4,5,6,7,8,9,10,11),
labels = sessionlabels)
domain.df$`research` <- factor(domain.df$`research`,
levels = c(1,2,3,4,5),
labels = likertlabels)
domain.df$`implementation` <- factor(domain.df$`implementation`,
levels = c(1,2,3,4,5),
labels = likertlabels)
domain.df$`engagement` <- factor(domain.df$`engagement`,
levels = c(1,2,3,4,5),
labels = likertlabels)
####################
# Wrangle Feedback #
####################
com01.df <- raw.df[raw.df$Q3 %like% "1",c(21, 74:75)] #will also catch sessions 10,11,12,14
com02.df <- raw.df[raw.df$Q3 %like% "2",c(21,290:291)] #will also catch session 12
com03.df <- raw.df[raw.df$Q3 %like% "3",c(21,182:183)]
com04.df <- raw.df[raw.df$Q3 %like% "4",c(21,236:237)] #will also catch session 14
com05.df <- raw.df[raw.df$Q3 %like% "5",c(21,128:129)]
com06.df <- raw.df[raw.df$Q3 %like% "6",c(21,344:345)]
com07.df <- raw.df[raw.df$Q3 %like% "7",c(21,398:399)]
com08.df <- raw.df[raw.df$Q3 %like% "8",c(21,452:453)]
com09.df <- raw.df[raw.df$Q3 %like% "9",c(21,506:507)]
com10.df <- raw.df[raw.df$Q3 %like% "10",c(21,560:561)]
com11.df <- raw.df[raw.df$Q3 %like% "11",c(21,614:615)]
#comheri.df <- raw.df[,c(616:618)]
#comheri.df$Q37 <- as.numeric(as.character(comheri.df$Q37))
#rename the likelihood session no. for CATA obs & listwise delete missingness
com01.df$Q3[com01.df$Q3 != "1"] <- "1" #this is key for CATA
com01.df[com01.df==""] <- NA
com02.df$Q3[com02.df$Q3 != "2"] <- "2"
com02.df[com02.df==""] <- NA
com03.df$Q3[com03.df$Q3 != "3"] <- "3"
com03.df[com03.df==""] <- NA
com04.df$Q3[com04.df$Q3 != "4"] <- "4"
com04.df[com04.df==""] <- NA
com05.df$Q3[com05.df$Q3 != "5"] <- "5"
com05.df[com05.df==""] <- NA
com06.df$Q3[com06.df$Q3 != "6"] <- "6"
com06.df[com06.df==""] <- NA
com07.df$Q3[com07.df$Q3 != "7"] <- "7"
com07.df[com07.df==""] <- NA
com08.df$Q3[com08.df$Q3 != "8"] <- "8"
com08.df[com08.df==""] <- NA
com09.df$Q3[com09.df$Q3 != "9"] <- "9"
com09.df[com09.df==""] <- NA
com10.df$Q3[com10.df$Q3 != "10"] <- "10"
com10.df[com10.df==""] <- NA
com11.df$Q3[com11.df$Q3 != "11"] <- "11"
com11.df[com11.df==""] <- NA
com11.df$Q3[com11.df$Q3 != "11"] <- "11"
com11.df[com11.df==""] <- NA
#comheri.df$Q37[comheri.df$Q37 != "2"] <- NA
# rename all the columns in these Likert dataframes
rename_com <- lapply(list(com01.df,
com02.df,
com03.df,
com04.df,
com05.df,
com06.df,
com07.df,
com08.df,
com09.df,
com10.df,
com11.df),function(x) {
names(x)<- c("session",
"session strength",
"session relevance")
x})
names(rename_com) <- c("com01df",
"com02.df",
"com03.df",
"com04.df",
"com05.df",
"com06.df",
"com07.df",
"com08.df",
"com09.df",
"com10.df",
"com11.df")
list2env(rename_com, envir = .GlobalEnv)## <environment: R_GlobalEnv>#bind the dataframes into one
#feedback.df <- rbind(com01.df,
# com02.df,
# com03.df,
# com04.df,
# com05.df,
# com06.df,
# com07.df,
# com08.df,
# com09.df,
# com10.df,
# com11.df)
#convert feedback variables to factor
#feedback.df$session <- factor(feedback.df$session,
# levels = c(1,2,3,4,5,6,7,8,9,10,11),
# labels = sessionlabels)
#feedback.df$role <- factor(feedback.df$role,
# levels = c(1,2,7,5,4,6),
# labels = c("Student","Faculty","Research Staff","Corporate/Industry","Community Member","Other"))
#write_xlsx(feedback.df,"C:\\Users\\josep\\Documents\\HERI\\feedback-mmdd.xlsx") #create a spreadsheet out of the dataframe; will stop knitting
#write_xlsx(com06.df,"C:\\Users\\josep\\Documents\\HERI\\feedbackS6-mmdd.xlsx")
#write_xlsx(feedbackheri.df,"C:\\Users\\josep\\Documents\\HERI\\feedbackheri-mmdd.xlsx")
#####################
# Dumbbell Settings #
#####################
# Version 1: https://r-graph-gallery.com/web-extended-dumbbell-plot-ggplot2.html (too many dots from full dataset)
# Version 2: from IIN (just dots plus legend with generated averages)
# Version 3: https://datavizpyr.com/dumbbell-plot-in-r-with-ggplot2/(did not print)
# Version 4: https://www.geeksforgeeks.org/how-to-make-dumbbell-plot-in-r-with-ggplot2/ (works with generated averages, sorted by session name)
# Version 5: https://www.youtube.com/watch?v=USNOB-5ou8k&ab_channel=BusinessScience at 2:1 (works with generated averages, plus guiding lines, sorted by greatest difference)
# Version 6: https://www.connorrothschild.com/post/dumbbell-plots/ with difference scores (works but can't get difference scores to print)
gold <- "#d8b365"
teal <- "#5ab4ab"
xsubtitle <- "Average Ratings (out of 5)"
caption <- "Source: 2023 HERI Evaluation Survey"
###########################
# K for all sessions #
###########################
prepost.df [prepost.df$id=="LO1",] %>% #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=="Trust & Trustworthiness"),
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=="Trust & Trustworthiness"),
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: Using the `size` aesthietic with geom_segment was deprecated in ggplot2 3.4.0.
## ℹ Please use the `linewidth` aesthetic instead.## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database
ggsave("Session All - LO1.png", height = 4, width = 7)## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database######################
# A for all sessions #
######################
prepost.df [prepost.df$id == "LO2",] %>%
mutate(difference = (post-pre),
position = post - (0.5 * difference)) %>%
ggplot() +
aes(x = pre,
xend = post,
# y = session) + #default: sorted by session name
y = reorder(session, pre), group = session) +
geom_dumbbell(
size_x = 5, colour_x = gold,
size_xend = 5, colour_xend = teal,
size = 1,
dot_guide = TRUE,
dot_guide_size = 0.15,
dot_guide_colour = "grey60") +
scale_x_continuous(breaks = seq (1,5,1),limits = c(1,5)) +
geom_text(data=filter(prepost.df, session=="Co-Lab(orative) Learning Workshop"),
aes(x=pre, y=session, label = "Before"),
color = gold, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(data=filter(prepost.df, session=="Co-Lab(orative) Learning Workshop"),
aes(x=post, y=session, label = "After"),
color = teal, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(aes(x=pre,y=session,label = round(pre, digits = 1)), color = "white", size = 2) +
geom_text(aes(x=post,y=session,label = round(post, digits = 1)), color = "white", size = 2) +
geom_text(aes(x=position,y=session,label = round(difference, digits = 1)), nudge_y = 0.25, size = 1.5)+
labs(title = "Ability Ratings Before & After Each Session", y = NULL, x = xsubtitle, caption = caption) +
theme(panel.border = element_blank(),
panel.background = element_rect(fill = 'transparent',color = NA),
panel.grid.minor = element_blank(),
panel.grid.major.y = element_blank(),
panel.grid.major.x = element_line(),
axis.ticks = element_blank()) ## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database
ggsave("Session All - LO2.png", height = 4, width = 7)## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database#############################
# Combined for all sessions #
#############################
prepost_long.df %>%
mutate(difference = (post-pre),
position = post - (0.5 * difference)) %>%
ggplot() +
aes(x = pre,
xend = post,
# y = session) + #default: sorted by session name
y = reorder(session, pre), group = session) +
geom_dumbbell(
size_x = 5, colour_x = gold,
size_xend = 5, colour_xend = teal,
size = 1,
dot_guide = TRUE,
dot_guide_size = 0.15,
dot_guide_colour = "grey60") +
scale_x_continuous(breaks = seq (1,5,1),limits = c(1,5)) +
geom_text(data=filter(prepost_long.df, session=="Groundwater Approach"),
aes(x=pre, y=session, label = "Before"),
color = gold, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(data=filter(prepost_long.df, session=="Groundwater Approach"),
aes(x=post, y=session, label = "After"),
color = teal, size=2, vjust=-1.5, fontface="bold",family="Lato")+
geom_text(aes(x=pre,y=session,label = round(pre, digits = 1)), color = "white", size = 2) +
geom_text(aes(x=post,y=session,label = round(post, digits = 1)), color = "white", size = 2) +
geom_text(aes(x=position,y=session,label = round(difference, digits = 1)), nudge_y = 0.25, size = 1.5)+
labs(title = "Combined Ratings Before & After Each Session", y = NULL, x = xsubtitle, caption = caption) +
theme(panel.border = element_blank(),
panel.background = element_rect(fill = 'transparent',color = NA),
panel.grid.minor = element_blank(),
panel.grid.major.y = element_blank(),
panel.grid.major.x = element_line(),
axis.ticks = element_blank()) ## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database
ggsave("Session All - combined.png", height = 4, width = 7)## Warning in grid.Call.graphics(C_text, as.graphicsAnnot(x$label), x$x, x$y, :
## font family not found in Windows font database#####################
# Likert by session #
#####################
#thissession <- "Trust & Trustworthiness"
#thissession <- "Equitable Partnerships"
#thissession <- "Nuts & Bolts"
#thissession <- "Community-Based Participatory Research"
#thissession <- "Participatory Budgeting"
#thissession <- "Role of Research(ers)"
#thissession <- "Communicating Research"
#thissession <- "Photovoice"
#thissession <- "Groundwater Approach"
#thissession <- "Climate Change & Environmental Health"
thissession <- "Co-Lab(orative) Learning Workshop"
sessionlikert <- likert(session.df[session.df$session==thissession,2:6])
plot(sessionlikert, positive.order = TRUE) +
labs(title = thissession)
ggsave("Session 11 - Session.png", width = 9, height = 2.5)
domainlikert <- likert(domain.df[domain.df$session==thissession,2:4])
plot(domainlikert, positive.order = TRUE) +
labs(title = thissession)
ggsave("Session 11 - Domain.png", width = 9, height = 2.5)