Introduction to unga"
In unga: Makes Multiple Contingency Tables of a Data Frame

The main purpose with this package is to analyze data from a questionnaire. You will be able to:

1) make contingency tables of every variable in a data frame. 2) cross tabulate all variables against an explanatory variable. 3) summarize some of the response alternatives of questions in a questionnaire. 4) overview the data and the dimensions of all questions, i.e. explore how many response alternatives every question has in a data frame. 5) find the item non response levels.

Let's begin!

knitr::opts_chunk$set(collapse = T, comment = "#>")
library(unga)
set.seed(1014)

First use library 'unga' to load functions.

library(unga)

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

Overview data

Assume that you want to get a quick overview of your data frame with respect to how many response alternatives every question have in a survey. Let's use the function table_alt(). This is a good way to start when you want to use table_unga() and extract questions with a specific number of response alternatives. Let's generate some test data and apply the function table_alt() to it.

set.seed(123456)
gender<-round(runif(50,1,2))
variable_1<-round(runif(50,1,2))
variable_2<-round(runif(50,1,5))
variable_3<-round(runif(50,1,5))
df<-data.frame(gender,variable_1,variable_2,variable_3)

table_alt(df)

The output here tells you that you have two questions with two response alternatives, and two questions with 5 response alternatives.

Specific cross tabulation

Let's now apply the function table_unga() on questions with 5 response alternatives on it and use gender as our explanatory variable.

table_unga(5,1,gender,df)

What we have done is to apply the function table_unga() to all variables in the data frame. We have selected all (hypothetical) questions with 5 response alternatives (the first argument in the function) in the data frame and used "gender" as our explanatory variable (the third argument in the function). The output here is proportion tables of those questions. The fourth argument is the data frame.

As described above, the first argument in table_unga() let you extract questions with a specified number of response alternatives. The third argument is the explanatory variable, and the fourth argument is the data frame. But what about the second argument? You could easily ignore this if you want. But it has to be specified to 1 or 2. This will not affect the output in Output_1, i.e. what is printed by table_unga() function. The second argument let you summarize response alternatives, but only for questions with 5 or 6 response alternatives. The output is found in Output_2. Let us again look at our example but now look on Output_2:

Output_2

This means that you have summarized response alternatives for questions with 5 response alternatives so now you have only 6 rows instead of 10 rows. The second argument 1 in the function means that you summarize alternatives according to: (A+B,C+D,E). If you use 2 as input in the second argument you will summarize alternatives according to: (A+B,C,D+E).

You can also summarize questions with 6 response alternatives. The summation will be done according to: (A+B,C,D+E,F). Make sure you specify your secound argument to 1. The summarized output will again be found in Output_2.

General cross tabulation

If you want to cross tabulate on all variables in your data frame without extracting questions with a specific response alternative, use table_unga_general(). This will return a list of contingency tables.

Item non response

The function table_non_resp() let you overview item non response within a data frame.

set.seed(123456)
gender<-round(runif(50,1,2))
variable_1<-sample(c(1,2,3,4,5,6,NA),50,replace = TRUE)
variable_2<-sample(c(1,2,3,4,5,6,NA),50,replace = TRUE)
variable_3<-sample(c(1,2,3,4,5,6,NA),50,replace = TRUE)
df<-data.frame(gender,variable_1,variable_2,variable_3)

table_non_resp(df)