[Catalyst] KD, survey app, zymurgological incentives

Kieren Diment diment at gmail.com
Sat Apr 22 00:21:30 CEST 2006


On 22/04/06, Len Jaffe <lenjaffe at jaffesystems.com> wrote:
>
> I worte a survey app in Cold Fusion a few years ago. I
> could do it again in cat.
>
> Do you have any specs?
> Can you define "encourages good survey design" beyond
> putting "Don't forget to design your survey well!" on
> every page?

My survey app had these kind of question/answer
> possibilites (this list is not meant to be 100%
> inclusive, I may have forgotten a few):
> * short text: text, date
> * text field (HTML text area)
> * Multiple choice(pick one of x): T/F, Y/N,
> traditional (formatted as an ordered list), agree
> disagree (with default formats for 6 choices [strong
> agree, agree, neutral, disagree, strong disagree, N/A]
> and 10 choices [like 6 but with more degrees of
> nuance].
>    - the survey creator was free to add or delete
> options, and change the formatting (radio/dropdown),
> and labelling.
> * pick all that apply (multiple check boxes)
>
> Data was stored to an RDBMS and was availabel for CSV
> download for the folks who had to ananlyze the data in
> SAS of access or whatever.
>
> Sound like what you're looking for?



Broadly speaking that's there.   Also needs to deal with optional and
compulsory questions, as well as different pathways (if you answered yes to
question 10, go to question 23...)   I'll sit down and write a more detailed
spec seeing as there's some interest.

With the Likert scale (strongly agree to strongly disagree) there's some
debate as to whether there should be a neutral option, so the survey
designer needs to specify some characteristics of the likert scale.

One feature the app needs to have is the ability to dump the questionsi into
$data_struture (like XML) so that it can be output to latex or something
similar.

Export formats should range from CSV to various forms of SQL to SAS/SPSS/S+
formats - whatever minimises the burden for the poor survey administrator.

My motivation behind this, is that apart from wanting such a thing, there's
a lot of shite sold that does the job adequately or worse, but for an
inflated price.  The survey software market needs a bit of open source
disruption to shake out the cowboys ;)

For the size of the beer reward, I guess I'm looking at a basic
implementation which is good enough, with a single example of how to extend
one aspect - like export survey to XML, or export results to SQL or
something like that.

Anyway, I'll draw up a more detailed spec and post it up somewhere  in the
next couple of days.


Kieren
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.rawmode.org/pipermail/catalyst/attachments/20060422/8452d5b5/attachment.htm 


More information about the Catalyst mailing list