Using console script

Introduction

KoboloadeR is designed to facilitate the analysis workflow of a survey collected through any xlsform compatible platform.

The package is designed to support 2 user flow:

Graphical User Interface - GUI: A dedicated shinyApp kobo_shiny("app_main_koboloadeR.R") includes all instructions and options to run your project.
Console mode gives you way to run your analysis without GUI

In this tutorial, we will focus on the using the package using the console mode.

Start the project

First you will need to install all the required software described in the package readme file (JAVA, R, Rtools-if on windows - and Rstudio).

You can now create a project, enable the library with

Loads KoboloadeR package

library (koboloadeR)

Creates folders necessary and transfer files needed. This function implement good practise for analysis reproducibility with dedicated sub-folder for the code, the data and the output from the analysis.

kobo_projectinit()

This will also push on your data folder a demo project with a form template that you can adapt according to your needs. Note that in the development of your form, in addition of this template, you can use the Integrated Framework for Household survey knowledge base

Extend your xlsform

The very first step is to extend your exisitng xlsform to include new column for your project configuration and analysis plan.

koboloader concept is that all the configuration shall exclusively be done in the orginial excel file. This is done for various reasons:

No need to create any R scripts which allow user with minimum R knowldege to use the package
Reproducibility * Documentation: this approach ensure that the analysis workflow is de facto fully documented within that excel file

Define the path to your form file (note that we use .xls file only)

form <- “form.xls”

Now we will automatically add the required additional structure for the analysis plan and project configuration with that same file

kobo_prepare_form(form)

Generate dummy data

You can skip this step in case you already have your own questionnaire and dataset.

In order to run this tutorial, we will first generate a dummy dataset using the structure of the form.

This can be performed with

Define the path to your form file (note that we use .xls file only)

form <- “form.xls”

Now use the function to generate your dummy file.

kobo_dummy_data(form)

This function will generate multiple csv files as the template form includes begin_group elements corresponding to a hierarchical data structure. A hierarchical structure is created when a series of questions is repeated say for each member in the household, and combined with a household questionnaire.

All csv files will be saved per default in the data folder.

Set up your project configuration

Set up your analysis plan

Generate report

kobo_check_analysis_plan(form)

kobo_load_data(form)

kobo_rename_xlsform_dataframes(form)

kobo_crunching_report(form)

Iterate

You have now set up all your project and you got your initial reports.

Analysis is an iterative process. you will need to

run-analysis.R

Edouard Legoupil

2021-02-04