An R data package full of questionnaires

The goal is to provide a catalogue of questionnaire details (e.g item labels, response categories, references, and more) to speed up the analysis and reporting of results.

Questionnaire data
- Item labels (short/long)
- Information about measurement, e.g. internal consistentcy, test-retest reliability
- Information about interpretation, e.g. clinical cutoffs, reliable change index
- Information about availability
- Citations
Helper functions for working with questionnaire data
- Check item/sum range
- Reproducible data correction
- Calculate prorated sums/means

Installation

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("milanwiedemann/questionnaires")

Available Questionnaires

Patient Health Questionnaire 9-item scale (PHQ-9; Kroenke, Spitzer, & Williams, 2001)
Generalized Anxiety Disorder 7-item scale (GAD-7; Spitzer, Kroenke, Williams, & Löwe, 2006)

Examples

Creating figures

This example illustrates how item labels from this package can be used in ggplot2 figures.

# Load packages
library(ggplot2)
library(tibble)
library(questionnaires)
set.seed(1234)

# Create some data
df <- tibble(item = 1:7,
             score = sample(1:3, 7, replace = T))

# Create figure using short labels from gad7
ggplot(df, aes(x = factor(item), y = score, fill = factor(item))) +
  geom_bar(stat="identity") +
  ggplot2::scale_fill_viridis_d(labels = gad7$item_label_short) +
  labs(x = "Item", y = "Score", fill = gad7$measure_short)

Creating Tables

This is a basic example which shows how to create a table with long label descriptions of the PHQ-9:

# Load packages ----
library(questionnaires) # For questionnaire data
library(dplyr) # For data manipulation
library(knitr) # For creating tables

phq9 %>%
    select(item_number, item_label_long) %>% 
    kable(col.names = c("Item", "Description"))

Item	Description
1	Little interest or pleasure in doing things
2	Feeling down, depressed, or hopeless
3	Trouble falling or staying asleep, or sleeping too much
4	Feeling tired or having little energy
5	Poor appetite or overeating
6	Feeling bad about yourself — or that you are a failure or have let yourself or your family down
7	Trouble concentrating on things, such as reading the newspaper or watching television
8	Moving or speaking so slowly that other people could have noticed? Or the opposite — being so fidgety or restless that you have been moving around a lot more than usual
9	Thoughts that you would be better off dead or of hurting yourself in some way

Reference items in text

When writing in RMardown it’s easy to reference item labels in text using the following:

Most clients showed most improvement on item 1 of the GAD-7 ('`r questionnaires::gad7[[1, "item_label_long"]]`')

Most clients showed most improvement on item 1 of the GAD-7 (‘Feeling nervous, anxious or on edge’)

How to contribute / add more questionnaires

A new questionnaire can be added to this library by creating a dataset with all information in the create-measures.R in the data-raw folder of this repositoty. Details for a new dataset needs to be added to data.R file in the data folder of this repositoty.

TODO / TOTHINK

Which licence would be appropriate for a data package like this?
maybe there should be a function to call the questionnaire e.g. get_questions("gad7", labels = "short", name = "short") and get_reponses("gad7", values = TRUE, wordings = TRUE). the functions would take the info from the underlying data sets
What are the legal challanges with copyright, license fees … and does it even matter for a project like this?
Would it be possible to just have short item labels for questionnaires that are not free to use?
Which information about a measure should be available in the help files, e.g. internal consistentcy, test-retest reliability, cutoffs, … ?
How to best name data sets so it doesn’t get confusing, e.g. gad7 vs general_health_anxiety_questionnaire_7item?
What’s a good package name, e.g. measureme, measureR, measuRes vs questionnaires?
What’s the best data structure to store this information, especially when questionnaires are not straightforward, e.g. different repsonse categories for each item?
How could different languages of the same questionnaire be integrated, e.g. suffix _en for English, _de for German?
This could also be integrated in a shiny app or similar, e.g. for data entry from paper/pencil or data collection?
add my functions to check the item & sum/mean range?
add my functions to calculate prorated mean?