With the emergence of the Internet and easy availability of vast quantities of data of all types and the computational power to manage, share, manipulate and mine all that data, the need for solutions to database problems is significant. The Databases Group explores solutions to these problems.
Project: Easy Survey Data Collection
Purpose: Develop a technique for easily entering survey data into a spreadsheet database.
Researchers: Tom Way, Mike Mason (Education & Human Services)
Data collection is a crucial step in gathering accurate, complete and correct data in a survey. Making data collection easy and incorporating entry constraints and real-time data validation are critical to producing high quality data. Processing of data is often done using statistical analysis packages (e.g., SPSS) or spreadsheet applications (e.g., Excel), so gathering and storing data in a compatible format is important, as is making the importing and exporting of data user-friendly.
This project is developing an approach for designing custom data entry forms that correspond to paper surveys, and connecting the form to an underlying database. The current platform being explored is Front Page for form design and Excel to serve as the backing store, as these are commonly available, well-known, and easy-to-use applications.
An initial proof-of-concept was completed in 2008, involving primarily manual design of the form and configuration of the communication between the form and the database. Future plans are to develop an application that will automate the process, or at least make the process less tedious.
One desired outcome is a measurement of the impact of using such a tool on survey efficiency, including aspects such as survey design time, data integrity and results analysis.
An important part of this project is the design of a form validation tool that will ascertain whether the form is complete, with no missing or duplicate fields or values, to the extent that such analysis is possible. An initial version of this tool was completed in 2008, with an extended version a desired result of continuing development.
Online and other forms of digital surveys are becoming more popular, and a significant proportion of them are designed directly in HTML. For medium and large survey instruments, the job of validating the form is complex and tedious, due to large numbers of form elements. This research explores the issues involved in creating such surveys and in validating them, enumerates those issues, and implements a software tool to solve them. Steps in this research include: