[QTI] Rounding outcome variables

[R.Bacon at surrey.ac.uk]

There is no neater way of rounding to 2dp that I know of.
 
Incidentally, the SToMPII system does implement this correctly, and is due to become
available (with a restricted set of question types to match the needs of the Physical
Sciences Question Bank) some time in the autumn.
 
It would be nice if the <round> operation allowed the number of significant figures
or decimal places to be specified (like <equalRounded>).
 
On the subject of equalRounded, if you wish to check whether a student has obtained
the correct value with the wrong precision, you need to supply equalRounded with the
number of figures they actually used. However, equalRounded can accept only literal
values or a template variable value for the number of figures, and you cannot set
template variable values within response processing. My question is, where is it best
to effect a non-standard change that allows the actual number of figures to be supplied
to equalRounded? Since template variables should not, I guess, be allowed to be processed
within outcomeProcessing, it looks as if equalRounded needs to be able to accept an
outcome variable value as a further alternative. Is this the best way to go, do you think?
 
Or have I missed an existing way of achieving this?
 
Best wishes,
Dick.

________________________________

From: ims-qti-bounces at lists.ucles.org.uk on behalf of Niall Barr
Sent: Fri 22/08/2008 14:27
To: IMS Question &amp;Test Interoperability
Subject: RE: [QTI] Rounding outcome variables



It seems to be done when the question response is displayed or moved to
the grade book, to a correct multiple response item result for an item
with three correct options is recorded as 0.99999 in the database.

That simplifies matching the Moodle way of doing things...

Niall

-----Original Message-----
From: ims-qti-bounces at lists.ucles.org.uk
[mailto:ims-qti-bounces at lists.ucles.org.uk] On Behalf Of Steve Lay
Sent: 22 August 2008 13:44
To: IMS Question &amp; Test Interoperability
Subject: Re: [QTI] Rounding outcome variables

There is a critical issue in this discussion which is sometimes lost
when people implement systems that use decimal representations of
numbers internally (or at least, seem to).

There is a big difference between displaying numbers to 2 decimal places
when you display your answer to the user and storing it in a rounded
form.

Dick has pointed out that there are times when you need to use the
decimal representation for checking answers (especially if you want to
know about the decimal representation used by the student, to
distinguish 3.14 and 3.140 for example) but this seems to be more
related to displaying the output of a scoring process?

When does moodle round to 2 d.p?

Steve

Niall Barr wrote:
> Hi all,
> 
> I'm trying to fix the (very broken) Moodle QTIv2 item export, so that
> it generates valid QTI items that behave exactly the same as the
> Moodle items. One thing Moodle does is round scores to 2 decimal
> places, so that a score made by adding three parts each with a value
> of 0.33333 becomes 1.
> 
> Is there a neat way of doinf this, or do I need to do something like
> this at the end of responseProcessing?
> 
>
> <setOutcomeValue identifier="SCORE">
>
>   <divide>
>
>     <round>
>
>       <product>
>
>         <variable identifier="SCORE"/>
>
>         <baseValue baseType="float">100</baseValue>
>
>        </product>
>
>       </round>
>
>     <baseValue baseType="float">100</baseValue>
>
>   </divide>
>
> </setOutcomeValue>
>
> 
>
> (Unfortunatly this doesn't work in either of the item players I'm
> using
> - bits are unimplemented in APIS (though it should only take an hour
> or so to fix that) and R2Q2 doesn't seem to calculate <product>
> correctly.)
>
> 
>
> Niall
>
>
> ----------------------------------------------------------------------
> --
>
> _______________________________________________
> IMS-QTI mailing list
> IMS-QTI at lists.ucles.org.uk
> http://lists.ucles.org.uk/lists/listinfo/ims-qti

_______________________________________________
IMS-QTI mailing list
IMS-QTI at lists.ucles.org.uk
http://lists.ucles.org.uk/lists/listinfo/ims-qti

_______________________________________________
IMS-QTI mailing list
IMS-QTI at lists.ucles.org.uk
http://lists.ucles.org.uk/lists/listinfo/ims-qti


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/ms-tnef
Size: 8533 bytes
Desc: not available
Url : /public/ims-qti/attachments/20080823/7addc1f3/attachment-0001.bin