Forum archive 2000-2006

Andrew L Vogel - Problem problem grader??

Andrew L Vogel - Problem problem grader??

by Arnold Pizer -
Number of replies: 0
inactiveTopicProblem problem grader?? topic started 4/1/2004; 2:45:01 PM
last post 4/8/2004; 11:15:50 AM
userAndrew L Vogel - Problem problem grader??  blueArrow
4/1/2004; 2:45:01 PM (reads: 983, responses: 3)
We have a problem we've written with value set at 4 points (there are four answer blanks) and some, not all, students are
getting full credit for the problem after submitting an incorrect answer in the first answer blank.  When I look at their problem as
an instructor and submit their answer the program even reports that their answer is incorrect(correctly), still they have 100% for the problem?

here is the problem (I cannot figure out what's going on, any
help is appreciated)  Andy


##DESCRIPTION
##KEYWORDS('probability','statistics','normal','normal distribution')
##
##ENDDESCRIPTION

DOCUMENT();        # This should be the first executable line in the problem.

loadMacros(
PG.pl,
PGbasicmacros.pl,
PGchoicemacros.pl,
PGanswermacros.pl,
PGgraphmacros.pl,
PGnumericalmacros.pl,
PGstatisticsmacros.pl
);

TEXT(beginproblem());
$showPartialCorrectAnswers = 1;

$pv = random(6000,12000,500);
$t = random(3,5,1);
$int=random(1,2,1);
$r = random(4,9,.25);
$i = ($r/100)/12;
$pmt = $pv*($i)/(1-(1+$i)**(-12*$t));

$ans1 = $pmt;
$ans2 = $pmt*$t*12;
$ans3= ($pmt*$t*12-$pv);
$ans4 = $pmt*(1-(1+$i)**(-12*($t-$int)))/($i);

BEGIN_TEXT
$BR
Tisha takes out a car loan for $pv dollars for $t years at $r % interest compounded monthly.  

$BR
Her monthly payment is  {ans_rule(10)}
$BR
The total amount she pays for the car is   {ans_rule(10)}
$BR
The total amount of interest she pays is   {ans_rule(10)}
$BR
The amount of money she owes after $int years is  {ans_rule(10)}
END_TEXT

ANS(num_cmp($ans1, tol=>0.005));
ANS(num_cmp($ans2, tol=>0.500));
ANS(num_cmp($ans3, tol=>0.500));
ANS(num_cmp($ans4, tol=>0.005));


ENDDOCUMENT();       # This should be the last executable line in the problem.

<| Post or View Comments |>


userMichael Gage - Re: Problem problem grader??  blueArrow
4/2/2004; 1:04:28 AM (reads: 1187, responses: 0)
I can't see anything wrong just looking at the problem.

One thing to keep in mind that there are two "scores" reported -- one is the score on the current submission of the problem. The other is the cumulative score, which is usually the highest score achieved. (If a student answers a problem correctly and then later redoes the problem incorrectly the current score will be zero but the cumulative score will remain 100%.) I've occasionally misread these messages.

The other trouble shooting tip I can offer is to use the debug command in the num_cmp facility

ANS(num_cmp($ans1, tol=>0.005, debug=>1 ));

This will result in a pink screen (to remind you not to leave the debug feature turned on) and print out of the steps that the num_cmp answer evaluator goes through in evaluating the answer. You can see where and when the "score" variable gets set to 1. It can be helpful to look at the code for NUM_CMP in PGanswermacros.pl at the same time to keep track of the order in which the filters are called within a num_cmp answer evaluators. This may help you locate where the problem is occuring.

Hope this helps.

-- Mike

<| Post or View Comments |>


userAndrew L Vogel - Re: Problem problem grader??  blueArrow
4/2/2004; 12:04:36 PM (reads: 1180, responses: 0)
Thanks for the tip on using debug.

After more research, it appears that the instructor created the set,
let students work for short time and then realized that they wanted
this problem to have 4 point value.  So they changed the value in
the modify problem set for all students area from 1 to 4. 
I'm not sure what the behavior of the program should be in such
a case, but my guess is that for those kids that had worked on
the problem before its value was changed form 1 to 4 the program
gave them 100% credit.

What do you think?

Andy

<| Post or View Comments |>


userArnold K. Pizer - Re: Problem problem grader??  blueArrow
4/8/2004; 11:15:50 AM (reads: 1203, responses: 0)

Hi Andy,

The "value" of a problem is separate from the "fraction correct" that is kept for each student in the database.  The "fraction correct" is a number bewtween 0 and 1 inclusive.  Whenever WeBWorK is asked to compute scores, it multiplies the "fraction correct" by the "value" to obtain the score.  So if a student did the problem correctly before the value was changed and then was scored after the value was changed, he or she should receive 4 points.

The per cent correct reported to students is 100*"fraction correct" rounded to 0 decimal points and is independent of the "value".  As Mike says, both the current and cumulative  per cent correct are reported.

I don't see how changing the value would have any effect on this (unless there is a really strange bug in WeBWorK associated with changing the value).

 

Arnie

 

<| Post or View Comments |>