### Formulas question: Marking of boxes with no entry

Number of replies: 1[b,c,d]

where nin Random Variables I have something like

a={1:5:0.1};

Then in the Global Variables I have something like

b=2*a;

c=5*b;

d=pow(c,2);

Grading variables have been set up so that incorrect answer for _1 does exclude the possibility of the user obtaining marks for _2 and _3 i.e. the error is carried forward.

cr0=_0/b >= 0.99 && _0/b <= 1.01;

cr1=_1/c >= 0.99 && _1/c <= 1.01;

cr2=_2/d >= 0.99 && _2/d <= 1.01;

Then the grading criterion is setup as

(cr1+cr2+cr3)/3

If a student attempts answering a, b and c, they obtain a mark depending upon how many of the answers are correct; however, if they only type an answer for one or two i.e. _0 and _1 but not _3, then they obtain zero marks.

Is it possible to solve this issue whereby the students need not input answers into all input boxes to obtain a mark?

### Re: Formulas question: Marking of boxes with no entry

Hello Jonathan,

If you only have one part, you can set the quiz behaviour to "Interactive with multiple tries". This will force students to answer all questions before checking their answers.

On the other hand, if you have several parts and you have set the quiz behaviour to "Adaptive mode", precisely so that the students do not have to answer all the questions at the same time, this is not the desired solution.

I thought of a workaround but haven't tried it yet. It would be a matter of dividing the part into three parts. Since we cannot transfer variables from one part to another, we could get around the problem by copying the answers given by the students from one part to the other, the copies being obviously hidden. This would allow a grading criterion to be applied in each part for a single answer but based on multiple answers.

As is often the case, this workaround wouldn't be the most elegant from a programming standpoint, but it would probably work reasonably well nonetheless. I will prepare an example a little later.

Being able to transfer variables from one part to another would be a nice improvement on the Formulas question. But from what Jean-Michel Védrine told me, it may be quite difficult to do. Anyway for now, the workaround would be the only option available.