CSCIE259 Project Proposal

Brian S. Hughes

Title

Alternative Medicine Survey

Author

Brian S. Hughes

Abstract

The Alternative Medicine Survey will survey web respondants on their experiences with alternative medicines. Survey results will be recorded in a database that is browsable by survey administrators, and possibly summarized to all web browsers.

What

Develop an on-line survey of respondants' experiences with alternative medicines. Survey to be served to anonymous respondents' browsers over the web. Questions will be dynamically generated; that is, later questions and pages displays will be dependent upon earlier responses.

Why

The field of alternative medicine has received little rigorous attention, but has accumulated many anecdoctal reports of success from laypeople along with much scorn from health professionals. Collecting the self-reported results from as many respondents as possible, in a non-threatening manner, may help illuminate areas that show promise for further in-depth study for either positive or negative results.

These results are not projected to constitute a rigorous study of alternative medicine's results; rather, they may help other workers to focus on areas that may produce fruitful results.

How

The survey will run on Tomcat based servlets. The survey's static form definitions will be stored in XML; any dynamically generated forms will also be created as XML by the servlets. The servelts will convert these XML based forms into HTML for display on browsers by translating the XML using XSLT, and also by use of FO, and JSP. Survey results will be stored in either a JDBC accessible SQL database or in an XML database. Access to Web services may be included if there are available web service(s) that will help support the survey or the later analysis/result displays. The servlets will drop cookie(s) onto the respondants' browsers to a)store respondant's survey state and b)attempt to prevent some spamming respondants from flooding the results with multiple duplicate responses.

Questions

  1. XML versus SQL DB? Any short term advantage (short term == project delivery date) to using one or the other? XML feels more accessible. If so, what might I think about to aid conversion to SQL later. Do I ever need to actually convert to SQL if I'm orders of magnitude smaller than, e.g., Amazon?
  2. Do you know of any Web services out there whose benefits for this application warrant the time needed to add Web service access to the project (doing WSDL parsing, SOAP, etc)?