[LON-CAPA-users] correct answer changing to incorrect

[Stuart Raeburn]

Hi,

The timestamp of Tue Nov 30 05:10:17 am is significant here.

5.10 am is the time LON-CAPA servers check their connections with  
other servers in the network, and resend any critical transactions to  
library servers which could not be completed when submission  
originally occurred.

One possibility is that the student was logged in to a server other  
than the library server, when the student's submission on part 11 was  
made, and the server was unable to complete the transaction to store  
it on the pitt.edu library server.  Given that part 11 was graded  
correct on 11/21, if this submission occurred after that date, but  
before the due date on 11/29, then correctness and tries information  
must also have not been transferred from library server to session  
server initially; hence a try was allowed and this was "Trial 1" as  
far as the access server was concerned.

When pending transactions were re-tried the next morning the store on  
the library server was successful. I recommend looking at the  
submission history to see which LON-CAPA host was hosting the  
student's session for this submission:

Tue Nov 30 05:10:17 am 2010 (EST)Part: 11 (ID 12) Trial 1  0.0490 m
Part: 11  incorrect

Stuart Raeburn
MSU LON-CAPA group

Quoting James Mueller <mueller@pitt.edu>:

> More details on this problem.
>
> If I check grades, he does have credit for getting this one correct.  
>   It is just that when he looks at the problem, it is displayed as   
> incorrect answer.  The other thing I don't understand is the due   
> date for this problem was Nov. 29, so where does this incorrect   
> submission on Nov 30 come from?
>
> -J. Mueller
>
> On Dec 8, 2010, at 3:51 PM, James Mueller wrote:
>
>> A student just complained that he solved a problem but later Lon   
>> Capa said he was incorrect.  I looked at his submissions and see   
>> the following.  From the submissions, it looks like he completed   
>> every part of the problem on Nov 21, but on Nov 30, his first   
>> incorrect answer for the first part appears to be entered again,   
>> and it is listed as the first trial.  I can give him credit for   
>> getting this correct, but does anyone have any ideas as to how this  
>>  might happen?
>>
>> -Jim Mueller
>>
>> Sun Nov 21 06:34:31 pm 2010 (EST)Part: 11 (ID 12) Trial not counted  0.0490
>> Part: 11  no_unit
>>  Sun Nov 21 06:34:33 pm 2010 (EST)Part: 11 (ID 12) Trial 1  0.0490 m
>> Part: 11  incorrect
>>  Sun Nov 21 06:35:54 pm 2010 (EST)Part: 11 (ID 12) Trial 2  2.0944 m
>> Part: 11  approx_ans
>>  Sun Nov 21 06:38:14 pm 2010 (EST)Part: 13 (ID 14) Trial not   
>> counted  284.8377 m/s
>> Part: 13  sig_fail
>>  Sun Nov 21 06:38:19 pm 2010 (EST)Part: 13 (ID 14) Trial 1  284.81 m/s
>> Part: 13  incorrect
>>  Sun Nov 21 06:38:23 pm 2010 (EST)Part: 13 (ID 14) Trial 2  284.84 m/s
>> Part: 13  incorrect
>>  Sun Nov 21 06:39:19 pm 2010 (EST)Part: 13 (ID 14) Trial 3  45.33 m/s
>> Part: 13  approx_ans
>>  Sun Nov 21 06:41:29 pm 2010 (EST)Part: 15 (ID 16) Trial 1  6.66 m/s
>> Part: 15  approx_ans
>>  Tue Nov 30 05:10:17 am 2010 (EST)Part: 11 (ID 12) Trial 1  0.0490 m
>> Part: 11  incorrect