Data Solutions

Data-driven solutions:

        • Vehicle to increase membership & member involvement

        • Help leadership at the national, state, and chapter-levels act based on feedback from bargaining units, chapter leadership, membership trends, etc.

        • Improve our interaction with our chapters to give them what they need

        • Mechanism for sharing best practices

        • Identify where people have overcome issues and share how they overcame that barrier

Scope: UFF project or joint partnership with AFT and/or NEA

Find the present document at the following link:

https://rpubs.com/samanthashep/datasolutions

Three proposed projects & their phases:

Project #1: Climate Survey

        • focus groups/staff feedback

        • survey design

        • data collection

        • data analysis

        • data visualization

Project #2: Membership Data Analyses

        • focus groups/staff feedback

        • data analysis

        • data visualization

Project #3: Bargaining Survey

        • focus groups/staff feedback

        • survey design

        • data collection

        • data analysis

        • data visualization

Software

Qualtrics will be used for survey design and collection.

GitHub will be used for storing data and analyses.

R will be used for data analysis and visualization.

What is R?

R is a free, open-source programming language used for statistical computing and data visualization.

What is R Markdown?

This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. R Markdown also helps keeps data analyses reproducible.

The present document will present example plots and the R code I used to generate them. We would receive this data output from the surveys we give bargaining units, and we can analyze this data to generate easy-to-understand data visualizations.

Below, on the right-hand side of each plot, click “Show” to view the code I used to generate them.

What is Reproducible Research?

“A data analysis is reproducible if all the information (data, files, etc.) required is available for someone else to re-do your entire analysis. This includes:

Data available
All code for cleaning raw data
All code and software (specific versions, packages) for analysis”

From: R Programming for Research - Brooke Anderson, Rachel Severson, and Nicholas Good; Colorado State University

Link here: https://geanders.github.io/RProgrammingForResearch/reproducible-research-1.html

Nationwide Replicability

These projects, and others like them, can be replicated nationwide.

• Example: pilot ideas first in Florida, co-create with Florida chapters, then use them elsewhere

Data Analysis & Visualization examples

The following are examples of visualizations we can get from each project (using fake data).

We can customize these data visualizations, track changes over time, combine to visualize associations, etc.

Project #1: Climate Survey

The following questions are example questions; we can further develop the questions based on focus groups, staff feedback, etc.

The survey can also be used as an organizing tool to:

        • sign up non-members

        • increase involvement among existing members

        • initiate follow-up contact

        • inform bargaining unit of importance of keeping their union

Click link for example survey: https://bit.ly/UFFclimate

Project #1

The next few examples will use items from the survey we just viewed.

Example #1

The following plot represents how the sample responded regarding changing their teaching due to recent legislation.

library(tidyverse)

climate1 <- read.csv("~/GGPLOT/climate1.csv")
View(climate1)

my_ggp <- climate1%>%
  ggplot(aes(x = answer, y = count))+
  geom_col(aes(fill = answer))+
  scale_fill_gradient2(low = "white", 
                       high = "purple") + 
  xlab("Responses")+
  ylab("Count")+
  labs(title="'I have felt compelled to change the way I teach",
       subtitle = " due to recent higher education legislation.'")+
  coord_flip()+
  theme_set(theme_minimal())+
  theme_dark(base_size = 14)+
  theme(
    plot.title = element_text(size = 16.5),
    plot.subtitle = element_text(size = 15),
    axis.title.x = element_text(size = 15),
    axis.title.y = element_text(size = 15),
    axis.text=element_text(size=10)
  )+

scale_x_discrete(limits = c("1", "2", "3", 
                            "4", "5", "6", 
                            "7"),
                 labels = c("Strongly disagree", 
                            "Disagree", 
                            "Slightly disagree", "Neither agree nor disagree", "Slightly Agree", 
                            "Agree", "Strongly Agree"))

  my_ggp

Out of 1,140 respondents:

• 78.9% reported they agreed with this survey item.

• 17.5% reported they strongly agreed with this item.

Project #1

Example #2

We can look at how the sample responded regarding considering moving out of state due to recent legislation.

climate3 <- read.csv("~/GGPLOT/climate3.csv")
View(climate3)


my_ggp <- climate3%>%
  ggplot(aes(x = answer, y = count))+
  geom_col(aes(fill = answer))+
  scale_fill_gradient2(low = "white", 
                       high = "purple") + 
  xlab("Responses")+
  ylab("Count")+
  labs(title="'I am considering moving out of state",
       subtitle = "due to recent higher education legislation.'")+
  coord_flip()+
  theme_set(theme_minimal())+
  theme_dark(base_size = 14)+
  theme(
    plot.title = element_text(size = 18),
    plot.subtitle = element_text(size = 15),
    axis.title.x = element_text(size = 15),
    axis.title.y = element_text(size = 15),
    axis.text=element_text(size=12)
  )+
  
  scale_x_discrete(limits = c("1", "2", "3", 
                              "4", "5", "6", 
                              "7"),
                   labels = c("Strongly disagree", 
                              "Disagree", 
                              "Slightly disagree", "Neither agree nor disagree", "Slightly Agree", 
                              "Agree", "Strongly Agree"))

my_ggp

Out of 1,140 respondents:

• 31.6% of respondents agreed that they were considering moving out of state due to recent legislation.

• 10.5% respondents strongly agreed that they were considering moving out of state due to recent legislation.

Project #1

Example #3

We can also analyze the relationship between 2 variables of the survey - for example, the 2 previous items that we just plotted.

library(smplot2)
library(tidyverse)


climate2 <- read.csv("~/GGPLOT/climate2.csv")

my_ggp <- climate2%>%
ggplot(aes(x=changeT, y=moveF)) +
  geom_point(size=3)+
  geom_smooth(method=lm, linetype="dashed",
              color="black", fill="purple")+

xlab("'Compelled to change Teaching' agreement")+
  ylab("'Considering Moving States' agreement")+
  labs(title = "Association between 'change teaching' & 'considering moving out of state'",
       subtitle = "(Correlation between the previous 2 survey items)")+
  
theme(plot.title = element_text(size = 30), 
        plot.subtitle = element_text(size = 20) 
        )+
  
scale_x_discrete(limits = c("1", "2", "3", 
                            "4", "5", "6", 
                            "7"),
                 labels = c("1", 
                            "2", 
                            "3", "4", "5", 
                            "6", "7"))+
  
  scale_y_discrete(limits = c("1", "2", "3", 
                              "4", "5", "6", 
                              "7"),
                   labels = c("1", 
                              "2", 
                              "3", "4", "5", 
                              "6", "7"))+
  

sm_statCorr(corr_method = "spearman",
              linetype = "dashed"
)

my_ggp

• This tells us that there is a positive correlation between the first two survey questions. The greater extent that respondents reported feeling compelled to change the way they teach predicts higher reports of considering moving out of state.

Project # 2: Membership Data

We can also learn a lot just from existing data, like membership data (i.e., this project does not consist of survey design or data collection). We can look at both chapter densities and membership recruitment statistics.

Project #2

Example #1

We can look at membership densities and compare side-by-side with each chapter.

chpdensity.uni <- read.csv("~/GGPLOT/chpdensity.uni.csv")
View(chpdensity.uni)


my_ggp <- chpdensity.uni%>%
  ggplot(aes(x = density, y = chapter))+
           geom_col(fill = "lightblue")+
  theme_dark(base_size = 14)+
  geom_text(aes(label = density), vjust = 0.6, color = "white", size = 4.5,
            fontface = "bold")+
  scale_x_continuous(limits = c(0,100))+
  xlab("% Density")+
  ylab("Chapter")+
  labs(title="March 2025 Chapter Densities: Universities",
       subtitle = "Fake Data Example")+
geom_vline(xintercept = 60, color = "red", linewidth = 1)+
  theme(
    plot.title = element_text(size = 20),
    plot.subtitle = element_text(size = 15),
    axis.title.x = element_text(size = 15),
    axis.title.y = element_text(size = 15),
    axis.text=element_text(size=12)
  )

my_ggp

Project #2

Example #2

To probe further, we could also look at new member recruitment statistics for the past 5 years; for example – within the (hypothetical) chapters that did not reach 60% density.

• While the previous plot represented chapter density, the following plot shows the count of new members recruited by year and by chapter.

chptime <- read.csv("~/GGPLOT/chptime.csv")
View(chptime)

ggplot()+
  geom_line(chptime, mapping = aes(x=year,y=newmembers, color = chapter), linewidth = 2)+
  geom_point(chptime, mapping = aes(x=year,y=newmembers,colors = chapter), size = 4)+
  
  labs(x="Year",y="New UFF Members")+
  labs(title="New Member Recruitment",
       subtitle = "2019-2024; UFF chapters below 60% density")+
  theme_dark(base_size = 14)+
  theme(
    plot.title = element_text(size = 25),
    plot.subtitle = element_text(size = 15),
    axis.title.x = element_text(size = 15),
    axis.title.y = element_text(size = 15),
    axis.text=element_text(size=14)
  )+
  
scale_color_discrete()+
    scale_y_continuous(limits=c(0,250))

Project #3: Bargaining Survey

        • Survey bargaining units – compare trends in what they want in bargaining

        • Survey leadership/bargaining teams – compare issues they run into

        • Display national metrics - create geo-maps to display data

Like the Climate Survey (Project #1), a Bargaining Survey can also be used as an organizing tool to: | | • sign up non-members | | • increase involvement among existing members | | • initiate follow-up contact | | • inform bargaining unit of importance of keeping their union

Project #3

Example #1

In a bargaining survey, we can present to respondents a list of commonly-reported bargaining issues, and have them rate the importance of each issue.

We can also include a write-in option for issues they believe should be prioritized, that are not already on the list we provide them.

• The following is an example of visualizing data if we aggregated respondent’s most popular choice of bargaining issue by state. This data could be presented in a more localized way, as well.

library(usmapdata)
library(maps)
library(dplyr)


set.seed(2522)
grouped_data <- us_map() |> 
  mutate(
    group = sample(
      c('income increases', 'paid parental leave', 'tenure protections', 'academic freedom'), 
      size = 51, 
      replace = TRUE
    )
  )
grouped_data |> 
  ggplot() +
  geom_sf(aes(fill = group)) +
  theme_minimal(base_size = 18)+
  labs(title="Bargaining Priorities",
       subtitle = "Faculty unions' top votes, by state")

Project #3

Example #2

We could also survey bargaining teams regarding recent issues they face during bargaining.

• The following is an example of what a plot would look like of the top-rated issues reported by bargaining teams.

bargaining <- read.csv("~/GGPLOT/bargaining.csv")
View(bargaining)

my_ggp <- bargaining%>%
  ggplot(aes(x = issue, y = count))+
  geom_col(aes(fill = issue))+
  scale_fill_gradient2(low = "white", 
                       high = "purple") + 
  xlab("Responses")+
  ylab("Count")+
  labs(title="Top-rated issues during bargaining",
       subtitle = "UFF faculty bargaining teams")+
  theme_set(theme_minimal())+
  theme_dark(base_size = 14)+
  theme(
    plot.title = element_text(size = 18),
    plot.subtitle = element_text(size = 15),
    axis.title.x = element_text(size = 15),
    axis.title.y = element_text(size = 15),
    axis.text=element_text(size=12)
  )+
  
  scale_x_discrete(limits = c("1", "2", "3", 
                              "4", "5"),
                   labels = c("Issue 1", 
                              "Issue 2", 
                              "Issue 3", "Issue 4", "Issue 5"))

my_ggp

Summary

Data-driven solutions can:

        • Increase membership & member involvement

        • Improve our interaction with our chapters

        • Identify common problems & share solutions

        • Help leadership at the national, state, and chapter-levels act based on feedback from bargaining units, chapter leadership, membership trends, etc.

        • Be highly customizable based on staff feedback

        • Be replicated nationwide

Data Solutions

Survey Design & Data Analysis Projects

Dr. Adela Ghadimi & Samantha Shepard, M.A.

Data-driven solutions:

Scope: UFF project or joint partnership with AFT and/or NEA

https://rpubs.com/samanthashep/datasolutions

Three proposed projects & their phases:

Project #1: Climate Survey

Project #2: Membership Data Analyses

Project #3: Bargaining Survey

Software

What is R?

What is R Markdown?

What is Reproducible Research?

Nationwide Replicability

Data Analysis & Visualization examples

Project #1: Climate Survey

Click link for example survey: https://bit.ly/UFFclimate

Project #1

Example #1

Project #1

Example #2

Project #1

Example #3

Project # 2: Membership Data

Project #2

Example #1

Project #2

Example #2

Project #3: Bargaining Survey

Project #3

Example #1

Project #3

Example #2

Summary

Data-driven solutions can: