Is there any other way to create dependent survey questionnaire using R Shiny?

Question

I would like to create a survey questionnaire where some of the questions are dependent on other question. I tried to use shinysurveys package. Below is the data for the questionnaire in table form. The questionnaire is for illustration purposes only.

The algorithm behind this app would be Question 2 will pop up if user answer for question 1 is 1. Question 3 will pop up if user answer for question 1 is 2. Both question 2 and question 3 will pop up if user answer for question 1 is 1 and 2.

Questions:-

On a scale from 1-3. how much you love sushi?
Answer selection: 1,2,3 (tick box multiple choice - can be more than 1 answer )
Really?
Answer selection: Yes, No (radio button multiple choice - only 1 answer)
Test
Answer selection: Yes, No (radio button multiple choice - only 1 answer)

Method 1 - I have tried the below code where I need to create my own input type, which is checkboxGroupInput using extendInputType but it doesn't work.

It will give this error if i tick box "1" and box "2" for question 1. And only show question that dependent on tick box "1".

This warning will be show Warning messages: 1: In if (input[[df$dependence1]] == df$dependence_value1) { : the condition has length > 1 and only the first element will be used.

extendInputType("checkbox", {
  shiny::checkboxGroupInput(
    inputId = surveyID(),
    label = surveyLabel(),
    choices = surveyOptions(),
  )
})
# Define a question as normal with the `input_type` set to the custom slider type defined above.
df_question <- data.frame(question = c("On a scale from 1-10, how much do you love sushi?", "On a scale from 1-10, how much do you love sushi?","On a scale from 1-10, how much do you love sushi?","Really", "Really", "Test"),
                              option = c("1", "2", "3", "Yes", "No", NA),
                              input_type = c("checkbox", "checkbox","checkbox","mc", "mc", "text"),
                              input_id = c("sushi_scale", "sushi_scale","sushi_scale","test", "test", "test2"),
                              dependence = c(NA, NA, NA,"sushi_scale", "sushi_scale", "sushi_scale"),
                              dependence_value = c(NA, NA,NA,"1", "1", "2"),
                              required = c(TRUE,TRUE,TRUE, TRUE, TRUE, TRUE))
# Watch it in action

  ui <- fluidPage(
    surveyOutput(df = df_question, "Sushi Scale Example")
  )
  server <- function(input, output, session) {
    renderSurvey()
  }
  shinyApp(ui, server)

Method 2 - Another way is to create a conditional panel in the UI and filter related data there. But, it is not convenient as I have to come up with several permutations.

library(dplyr)
library(shinysurveys)
ui <- shiny::fluidPage(
    
    uiOutput('ui')
    
  )
  
server <- function(input, output, session) {
    
    df_question <- data.frame(question = c("Really", "Really", "Test"),
                              option = c("Yes", "No", NA),
                              input_type = c("mc", "mc", "text"),
                              input_id = c("test", "test", "test2"),
                              dependence = c(NA, NA, NA),
                              dependence_value = c("1", "1", "2"),
                              required = c(TRUE, TRUE, TRUE))
    
    
    output$ui <- renderUI({
      fluidPage(
        checkboxGroupInput(inputId = "Sushi_scale", label = "On a scale from 1-3, how much do you love sushi?",
                           choices = c("1","2", "3")),
        conditionalPanel(condition = "input.Sushi_scale == '1'", 
                         shinysurveys::surveyOutput(df_question %>% 
                                                      filter( `question` == "Really")) ,
                                                    theme = "#ffffff"),
        conditionalPanel(condition = "input.Sushi_scale == '2'", 
                         shinysurveys::surveyOutput(df_question %>% 
                                                      filter( `question` == "Test")) ,
                         theme = "#ffffff"),
        
        )
        
    })
    
    
  }
  
  shiny::shinyApp(ui = ui, server = server)

Method 3 - This method doesn't work as well. I tried to pass reactive value to the UI.

library(dplyr)
library(shinysurveys)
ui <- shiny::fluidPage(
    
   uiOutput('ui')
    
)
  
server <- function(input, output, session) {
    
  df_question <- data.frame(question = c("Really", "Really", "Test"),
                              option = c("Yes", "No", NA),
                              input_type = c("mc", "mc", "text"),
                              input_id = c("test", "test", "test2"),
                              dependence = c(NA, NA, NA),
                              dependence_value = c("1", "1", "2"),
                              required = c(TRUE, TRUE, TRUE))
    
    vals <-reactiveValues()
    eventReactive(input$go,{
      
      if (input$Sushi_scale == "1"){
        vals$df <- df_question %>% filter(question == "Really")
      }
      
    })
    
    
    output$ui <- renderUI({
      fluidPage(
        checkboxGroupInput(inputId = "Sushi_scale", label = "On a scale from 1-3, how much do you love sushi?",
                           choices = c("1","2", "3")),
        shinysurveys::surveyOutput(vals$df) ,
                        theme = "#ffffff")
    })
    
}
  
shiny::shinyApp(ui = ui, server = server)

I discovered this website: https://github.com/EconometricsBySimulation/Shiny-Demos/tree/master/Survey where the writer creates a questionnaire via Shiny R. The demo is here: https://econometricsbysimulation.shinyapps.io/Survey/. The only thing is that I want the questionnaire to be on one page without having to click next to answer every question.

Are there any other recommendation to built this survey using Shiny R?

Is there any other way to create dependent survey questionnaire using R Shiny?

Answers (1)

Related Questions