Correct answer marked as correct and incorrect at different times

[Clodagh Carroll]

Hi all,

I have a question where the correct answer is sometimes being marked as correct and sometimes as incorrect, depending on how/when it is entered. Using the product rule on the below for example, sometimes gives credit for the correct answer and sometimes not (for the same thing).

In relation to the structure of the question, the power on the 2^nd bracket is a variable between 10 and 20 and the constant term in the 2^nd bracket is a variable between 1 and 5. The first bracket is always equal to the 2^nd bracket squared so that the product rule can be avoided if the student spots that.

The link to the question is: https://numbas.mathcentre.ac.uk/question/153055/avoidable-product-rule/embed/?token=5cc2f9aa-c58e-454c-b6bc-ae740bfd149b and screenshots of credit being awarded or not for the same answer are attached as Incorrect 1, Incorrect 2 and Correct. I spotted similar queries in the blog before but none that were resolved to help in this instance. Any suggestions welcome.

 Many thanks,

Clodagh

 

 

[Ben Brawn]

How often is it marking it incorrectly, any idea?

Have you changed the default settings on the accuracy tab in the mathematical expression part?

--
You received this message because you are subscribed to the Google Groups "Numbas Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to numbas-users...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/numbas-users/b830f13e-ad55-4ee6-8641-47d1be816071n%40googlegroups.com.

[Clodagh Carroll]

Hi Ben,

It marks it incorrectly maybe every 4 or 5 attempts, but sometimes if you keep entering the same answer in a given attempt it will eventually mark it incorrect. I did try changing the accuracy to 0.000001 but it didn't seem to make a difference, so I re-set it to 0.001. For some reason it seems more likely to mark the answer as incorrect when the second bracket is (x+5) to some power.

Regards,

Clodagh

You received this message because you are subscribed to a topic in the Google Groups "Numbas Users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/numbas-users/E9bHPVCHiSM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to numbas-users...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/numbas-users/CABW86oKWxNjK9MFaBqmu5NQG9EjF16TVD0YkBMjD%2B%3Dnt9ei4qQ%40mail.gmail.com.

[Ben Brawn]

I assume the issue is that with the large powers the numbers get so big that the small decimals don't stay happy. 

I would consider the constants that you are adding to x, let's say you wrote the question to use (x+a), and then instead of the accuracy tab starting at 0 and going to 1 (the default), I'd shift the interval so that the bracket you are raising to the power is closer to 0 (rather than 'a+0.5'). Here is an example where I have let 'a' go from 4 to 8 and the accuracy checking is from -8 to -4 and it works well from what I can see. 

https://numbas.mathcentre.ac.uk/question/share/view/e6651161-7831-45a9-8d08-6cabd844ee30

[Don Shearman]

I seem to recall having a similar problem a year or so ago. I think Christian suggested changing the range used for checking the answer. As Ben suggested, high powers of even small numbers can give problems. I think the trick was to keep the checks to numbers around 1. 

Don Shearman

Sent from my iPad

On 27 Feb 2024, at 23:51, Ben Brawn <ben....@gmail.com> wrote:

I assume the issue is that with the large powers the numbers get so big that the small decimals don't stay happy. 

To view this discussion on the web, visit https://groups.google.com/d/msgid/numbas-users/ef6ffc6c-d737-4439-9f83-339b7048016dn%40googlegroups.com.

[Clodagh Carroll]

Hi Ben,

Thanks for this - I have made similar changes to my question and it seems to have done the trick. I had a couple of alternative answers included for partial credit and it seems that the same changes need to be made in the Accuracy tab for each of these in order for it to be successful, but it looks like it's working as it should now.

Thanks very much for your time and help with this - much appreciated.

Best wishes,

Clodagh

To view this discussion on the web, visit https://groups.google.com/d/msgid/numbas-users/ef6ffc6c-d737-4439-9f83-339b7048016dn%40googlegroups.com.

[Clodagh Carroll]

Thanks Don - it does look like that was the issue. Good to know for future questions too.

Regards,

Clodagh

To view this discussion on the web, visit https://groups.google.com/d/msgid/numbas-users/9BF8551C-DB9F-48D5-BDB4-ECFF30685B88%40gmail.com.