[Using Sakai] Advanced survey tool (Grossman,John E)

[Regan, Alan]

Hi, John.

If the point value is consistently 4 points, I wonder if the partial credit feature in Tests & Quizzes could help?  It would require the assessment creator to enable partial credit on each question, but I could see something like...

Total point value for each question = 4

Set up the partial credit alignment based on their scale needs, e.g. --
A. 100%
B. 75%
C. 50%
D. 25%

(So, the partial credit amounts would calculate to 4*100%=4, 4*75%=3, 4*50%=2, 4*25%=1.)

You'd have to enable the "anonymous" feature for student identities, to try to mimic an anonymous survey.

I've only used Tests & Quizzes for assessments, not surveys.  I'm not sure if this would accomplish your goals, but might be worth experimenting with?

Sincerely,

Alan Regan, MFA
Manager, Technology and Learning
Information Technology
Pepperdine University
(310) 506-6756


------------------------------

Message: 2
Date: Tue, 29 Nov 2011 12:00:57 -0600
From: "Grossman,John E" <john.grossman at mdanderson.org>
Subject: [Using Sakai] Advanced survey tool
To: "sakai-user at collab.sakaiproject.org"
        <sakai-user at collab.sakaiproject.org>
Message-ID: <CAFA79F9.1D604%john.grossman at mdanderson.org>
Content-Type: text/plain; charset="us-ascii"

Question: Does anyone have experience using a survey or assessment tool with Sakai that allows for survey questions to be associated with numeric codes and for database storage of the codes as part of the response data?

Background: Our Behavioral Sciences department is interested in delivering educational content via Sakai with surveys that would measure behavioral changes in the learners. They normally code survey responses using numeric values to simplify statistical analysis. For example, if a multiple choice question has 4 responses, they may prefer to analyze the data using coded values 1, 2, 3, and 4 rather than the textual content of the response.

We realize we could map responses to codes using XSLT, database formulas or Excel formulas, but ideally the survey tool itself would allow for definition of the codes at the time of survey creation and storage of the codes in the results database. This would avoid the need to write custom mapping routines for each survey/evaluation.

--
John Grossman
Director, Academic Technology Services
The University of Texas MD Anderson Cancer Center
713-745-0305