RE: [Bodington-developers] Questionnaire Functionality

[Brian P. C. <bm...@bm...>]

> > From: Brian Peter Clark
> [...]
> > Ultimately, in this case, the specs are reasonably clear, but not 
> > very good. The 50% rule is too wishy washy. 
> 
> What would you prefer?
> 
> 		- Peter

I said before that my preferred approach, if one has to have a 
compulsory questionnaire, would be a completely automated 
badger with questionnaire results revealed to the module leader 
after the final deadline has passed.  There is no arbitrariness, all 
students are treated equally, the students are not hassled 
personally, anonymity is preserved.  The price is one or two extra 
emails.  Staff time is not used up on the badgers; and the lack of 
reliability from the less than 100% response might well be 
balanced by the unreliability introduced by petulance induced by 
the badgering.

However, some remarks on the specs given. 

First of all, to tighten things up, specify whether or not any interim 
response summaries are accessible.  

The simplest case is where summary results are provided only 
after the final deadline (as opposed to any intermediate badgering 
deadline).

When does revealing the list of non-respondents compromise 
anonymity?  

When the number of initial respondents is small.  The danger is 
that it remains small and the anonymity of the initial responders is 
weakened.

So all decisions should be based on numbers, not percentages.  

If you have a less than 100% response and you want to badger this 
number up to 100%, I can't see the significance of 50%, at some 
deadline or otherwise.

For very small classes, I wouldn't allow badgering at all - or 
anonymity goes.   So set a numerical limit for initial respondent 
group size - 10? 15?  There is still an arbitrariness here, but it is 
better than one based on percentages.  (There might be stipulated 
minimum sizes for modules, I don't know.)

After the given badger deadline, if the number of respondents is 
greater than one's chosen lower limit, say 20, then release the 
names of the non-respondents.

It's hard to be definitive without more context, but hide the interim 
responses and nothing much can go wrong.

The cat is set amongst the canneries if any interim responses are 
made available to the module leader, or, horrors, he or she has 
continuous access to the responses.

If the interim results are made available at the badger deadline, this 
compromises the anonymity of a small number of those yet to 
submit (whose crime is being tardy).

If the module leader has continuous access to interim data, then 
there should be a first release of names if the number of 
respondents is more than the stipulated minimum.  Also, there 
should be no release of names if there are less than the same 
number left to respond.

There might be step-releases after this in blocks of the basic 
minimum unit.

In this, I've assumed that the badger deadline is separate from a 
line-in-the-sand deadline.  After the badger deadline, the student is 
late.  After the final deadline the student has sinned.  In this case 
one might demand that the penalty is loss of anonymity (making 
an honest response less likely).

It was the combination of these considerations that made me use 
the term wish washy: percentages instead of numbers, small 
modules, imprecise specification of the nature of the deadline, and 
lack of information about access to interim data.

Regards,

Widow Twanky


> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting language
> that extends applications into web and mobile media. Attend the live webcast
> and join the prime developer group breaking into this new coding territory!
> http://sel.as-us.falkag.net/sel?cmd
> _______________________________________________
> Bodington-developers mailing list
> Bod...@li...
> https://lists.sourceforge.net/lists/listinfo/bodington-developers
>