Re: Hello and Comment

[William Haynes]

Good Morning Chris,

I’m glad to hear that you are finding the textbook and the exercises useful.  I’m interested in any feedback you have, so don’t hesitate to contact me.

The interactive problems are created using Numbas  https://numbas.mathcentre.ac.uk, which is open source, powerful, flexible and relatively easy learn.   Simple problems are very easy to write, complex problems are possible with more effort.   

The basic scoring algorithm I am using is this:   

 

A number is considered right if it is within 0.2% of the correct value, and close if it is within 1.0% of the correct value.  

The student gets full credit if both value and units are correct.

75% if the value is close with correct units

50% if the value is correct but has no units or wrong sign

25% if the value is close

        0% otherwise.

The point values for individual parts of a question are pretty arbitrary, and the total points for a question might add to any value.  I enter the percentage of possible points into my grade book.    

The system is smart enough to do unit conversions, so 1 kN and 1000 N are both correct, however I had to hack it to accept multiple representation of angles as correct.  It is possible that there are some bugs or edge cases here that I missed, so send me a screenshot if you ever see anything that is not working as expected.  Numbas itself expects all angles are measured CCW from the + x-axis, and that is what it shows as the "expected answer" box.

Here are some other tips which I pass out to my students

• You may attempt Numbas problems multiple times, or replace incorrect answers and try again.

• Be sure to refresh the problems a couple of times to see all the variations.

• Labels can be dragged to make labeling clearer.

• Sometimes the illustration doesn't load. Hit refresh in this case.

• Give your answers to engineering accuracy which is no more than 0.2% error. This means give three significant digits unless first digit is a 1, in which case give 4 significant digits.

• Most answers will need units.

• Units require proper capitalization. "m" not "M", "kN", not "kn", etc.

• For angles, use a degree symbol  "°"  or type   "deg".

• Greek letter π can be entered by typing "pi".

• Include a leading zero for numbers like 0.25 or -0.5.

• This is a new system and I am interested in hearing about any problems you have with it.

• If you feel that a problem is returning the wrong answer, take a screenshot of the problem/answer and let me know so that I can fix it. 

Numbas gives authors a lot control over feedback, but doing tricky things with feedback gets complicated so I haven’t done much with it beyond the basic.   For instance, you can build a question with multiple paths depending on the students answers to previous parts or based on choices that they make.     This refers to automatic feedback generated based on the student’s response.  I don’t think that there is any way to write individual feedback like a comment on a google doc.  

It’s possible to integrate Numbas questions with your LMS using the Numbas LTI Provider.  When you do, grades get entered automatically and it is possible to look over individual students work, see time spent, review history, change grades, give extra time, view class statistics, etc.   

The source code of all my mechanics questions are available for inspection or reuse here:    https://numbas.mathcentre.ac.uk/project/3501/browse/

The "basic" and "advanced" folders contain the individual problems, the "chapter exercises" and "homework sets" are bundles of problems.  If you create a Numbas account for yourself, you can clone them and experiment. If you want to write your own questions for the book, let me know and I will help you get started. 

Will

On Aug 12, 2021, at 9:14 AM, Christopher M Papadopoulos <christopher....@upr.edu> wrote:

Good Morning Dan,

Thanks for getting back, and I'll wait for Will to have a chance to look into my inquiry.  Let me just say that I loved the design of the problem, to allow multiple answers - that is exactly what is needed.

On further reflection, I think I see more of what happened in my exercise.  I missed the units, so that led to deductions.  I also believe in using units, but I think it would work better to have a separate box for units.  I never cared so much for entering numerical and text data in the same box.  But now I am aware and I will advise my students.  [I didn't check, but do these problems recognized unit conversions, e.g., rad, kips, etc.?]

Still, for my answer to Force A, my second answer for angle seemed to have been entirely ignored.  And for B and C, it seems that I was deducted 2 points in the angle - was that for the lack of 'deg' units or for format?

So maybe in the end the errors were mine and not yours, but a little more guidance in the remarks might help.  Also in the 'Expected Answer' for the angle, the reference axis there could also be specified.  Even better would be to show multiple solutions, although this could get out of hand.  Even if we agree on a standard +x-axis reference, we could debate if the preferred answer is, say, 225 deg or -135 deg.

Regarding Feedback, I hesitate to suggest too much, as I am not familiar with any of the coding complexities that might go along with it, but I envision something like the following:  So it does not get out of hand, maybe there is first an instructor registration, and then registration would allow two modes, 'standard' and 'feedback'.  Standard mode would be as is, and feedback mode would enable some sort of action or button to give feedback.

Feeback could possibly take the form of enabling a text highlight and comment option, just like commenting in Word or Google Docs.  For equations, maybe this would be slightly different to highlight or tag the entire equation, as it seems that the formatting does not allow highlighting by click and drag (but I love the ability to download the source coding!).  For the exercises, perhaps there would be discrete buttons to comment, but these could be tied specifically to the user's instance of the problem, with their particular data and responses.  When I wrote you my message, I scrolled through the entire exercise and used Snip to grab the segments to reconstruct the entire problem. To streamline this, maybe the system would generate a link for that entire instance, to avoid cutting and pasting.

If you'll allow me another quick thought, I think using "Fhat" for unit vector and "F" for vector is a little awkward, although I'm sure you have a reason for selecting that.  As you know, the standard books use "lambda" or "u", often with the "hat" accent.

Just a few thoughts.

Thanks

Chris

---

From: Dan Baker <dan....@colostate.edu>
Sent: Thursday, August 12, 2021 8:31 AM
To: Christopher M Papadopoulos <christopher....@upr.edu>
Cc: William Haynes <wha...@maritime.edu>
Subject: Re: Hello and Comment

 

Hi Chris,

I'm excited that you are trying out our OER book. I've cc'd my co-author Will Haynes as he has coded in all the Numbas exercises and should be able to provide more details.

We have not yet created a feedback system for the book. Its something we've talked about and not landed on a solid solution. Any suggestions?

Dan

____________________________________________________

Daniel W Baker Ph.D. PE, Teaching Associate Professor
  Office:  ENGR B206                       Phone: 970-491-0261
  Email:  dan....@colostate.edu  Twitter: @drdanteaches

  YouTube: DrDanTeaches                Pronouns: he, him

  Working to be a kind & supportive ally to all!

On Wed, Aug 11, 2021 at 2:01 PM Christopher M Papadopoulos <christopher....@upr.edu> wrote:

Hi Dan,

It was great to see you at ASEE a couple of weeks ago.  I'm going to try out your OER book with my class!

So, I am preparing my first unit on vectors and I went through Chapter 2 of your book.  I like Exercise 2.9, but I'm already having some questions about how the problems are scored.  A minor rounding error was deducted heavily, and in two other cases, the equivalent form of the answer (different reference for angle) were marked incorrect.  One was a rounding error that I think was minor and the Could you please check out my attached files for details?

Are these the kinds of things that you can address, and is there a common message board for people to post questions and comments of this sort?

Thanks

Chris

-- 

Christopher Papadopoulos, PhD
Professor
Department of Engineering Sciences & Materials
University of Puerto Rico, Mayagüez
Call Box 9000
Mayagüez, PR  00680-9000
787-832-4040 x3336
christopher....@upr.edu

greatidea.uprm.edu