ITEM METADATA RECORD
Title: One by One: A Shiny Web-app for the Design and Analysis of Single-Case Experiments
Authors: De, Tamal
Michiels, Bart
Vlaeyen, Johan
Onghena, Patrick
Issue Date: Jul-2017
Conference: useR! annual conference edition:2017 location:Brussels, Belgium date:04 - 07 July 2017
Abstract: The increasing popularity of single-case experiments in the fields of educational and behavioral research necessitates the development of easy to use computer programs for the design and analysis of these experiments. SCRT (Single-Case Randomization Tests), SCVA (Single-Case Visual Analysis) and SCMA (Single-Case Meta-Analysis) packages in R already developed by researchers of KU Leuven - University of Leuven, Belgium, serve that purpose brilliantly, except that researchers should possess extensive knowledge of R to explore the full potential of these libraries (Bulté & Onghena, 2013).

A solution to this hindrance is to develop a web-based application which exposes functionality already coded in these packages in a user-friendly GUI. The platform chosen for that purpose is Shiny, an open source web application framework developed by RStudio. We first implemented the functions available in the above-mentioned packages in a carefully planned GUI. The GUI was split into tabs with a logical progression. In the first tab, the user can design the experiment by setting design parameters using the design functionality in the SCRT package. After the experiment has been conducted, the user can upload the observed data in the next tab. The third tab contains tools for visually analyzing the data using the SCVA package. In the fourth tab the user can perform randomization tests on the observed data using the SCRT package (Edgington & Onghena, 2007). Finally, in the fifth tab there are further tools for meta-analysis and calculation of effect size measures from the SCMA package.

The biggest challenge in the process of development was making sure the GUI was easy enough for the average user to be able to use without any external help while also ensuring the GUI to be reactive enough to adapt and change options for parameters with every user input. Furthermore, we added numerous tips and error messages to alert and explain the correct format of inputs required and outputs saved to the user. We also added more functionality to the basic SCRT, SCVA and SCMA packages and adapted the webapp accordingly. A final challenge will be to host the web-app on RStudio or on the university servers and allowing external researchers access to use the web-app and provide valuable feedback which can then be incorporated as required.
Publication status: published
KU Leuven publication type: IMa
Appears in Collections:Health Psychology
Methodology of Educational Sciences

Files in This Item:
File Description Status SizeFormat
Tamal-Kumar-De_170623_103005 (1).pdfConference poster Published 1183KbAdobe PDFView/Open

These files are only available to some KU Leuven Association staff members

 


All items in Lirias are protected by copyright, with all rights reserved.