[Bf-education] how to catch cheaters in a class?

Fri May 26 14:27:37 CEST 2017

Yes, it seems that the motivations and staying away of cheating is a
personal, individual manner. In fact, cheating is a significant signal sent
loudly: classes and assignments are boring and seem to be unnecessary.
However, this is education, the daily struggle with laziness and
unmotivated ppl. Indeed, more psychology than technical stuff.

You all helped me understand better what the real goal should be, and if I
start to dig and detect cheats it could mean that I failed as a teacher.
Unfortunately, students choose to study computer graphics (and my favorite
Blender) because they do not want other classes, even harder, like
calculus, etc. So, if they choose by elimination, not by selection, they at
start are not motivated enough to take the effort.

Thank you! :) I understand something, but more thinking and analyzing is
necessary to avoid cheating.

regards
pio

Piotr Arłukowicz, PhD, BFCT
University of Gdańsk, Faculty of Mathematics, Physics and Informatics,
Dept. of AI,
Wit Stwosz 57, 80-952 Gdańsk, room 121, tel.: +48585232151,
https://inf.ug.edu.pl/~piotao
Polish Blender Course: http://polskikursblendera.pl/ [PL]

2017-05-26 11:00 GMT+02:00 Ernesto Del Valle <cecilff at gmail.com>:

> The first thing I thought about when I read this message is the point that
> Ton refereed to. If Open Source and Creative Commons has shown us something
> is that copying is not bad. Saying that you copied something and not
> acknowledging it and the work of someone else, is. Basically it's lying.
> But using someones work and building upon it is not bad, and is
> definitively and important skill set in Blender specifically. Is just
> impossible to tell, how many times I have reused others work as well as my
> own. Just think of libraries.
> Of course it is important to develop the skills of building something from
> scratch, but in order to do that, one must be really motivated to do it, so
> that thing that is going to be build from zero, must be something one
> really relates to on a personal level, so it must have one's signature.
> Even in that situation, is way faster to just take some parts or solutions
> already done. Taking a solution made by some else, also makes me think on
> how they solve the problem and I can learn from that.
> The students must be encourage to make their own projects with their
> distinctive signature marked by their own ideas and tastes. If the projects
> they are to deliver are way too similar, and it's easier to just copy the
> file and put my name on it, I think the project is not making them learn
> much anyway, but just to repeat a set of instructions that deliver an
> expected result. In art in general and in Blender in specific, you have to
> be able to solve the unforeseen problems that arise always.
> So, to grade that kind of work is harder, but the learning is completely
> guarantee to be valuable and enjoyable.
>
> I love this subject, but I guess is more suited for a pedagogy specific
> panel.
>
> Happy blending!
>
> 2017-05-25 15:37 GMT-04:00 Knapp <magick.crow at gmail.com>:
>
>> I was also thinking about giving out team projects. This would mean, if
>> there were cheating, it would at least be a team effort.
>>
>> On Thu, May 25, 2017 at 7:37 PM, Antonio Carvalho <antoniorcn at hotmail.com
>> > wrote:
>>
>>> Hi Piotr,
>>>
>>>     I'm teacher in Brazil, working in many classes of Java Programming,
>>> C Programming and cheating is a problem faced all days.
>>>
>>>     Actually I'm giving individual works for each student (they can
>>> choose what they want to do), and according to Ton Roosendaal sugestion, I
>>> applied a competion, each student evaluate the job of other 3 (or more
>>> students), based in some criterias, distributing  a number of specific
>>> points for each criteria, for sample 18 points in case of 3 jobs, and to
>>> avoiding have them distributing 6 points for each, without evaluate
>>> correctly, I penalize the students that evaluate the jobs in certain
>>> criteria, in diferent way than the other peers.
>>>
>>>     It looks like complicated, but there are some tools which allow it,
>>> in my case I'm using Moodle (www.moodle.org) whoose have a kind of
>>> exercise type named workshop allowing this propose, in this kind of
>>> exercise the teacher can specify the criterias, the student can submit his
>>> own work, after that the tool will random who will evaluate whom, and at
>>> the end you can calculate the average of each student.
>>>
>>> Regards,
>>>
>>> *Antonio Rodrigues Carvalho Neto*
>>> Faculdade de Tecnologia do Estado de São Paulo (FATEC)
>>> campus Zona Leste e Carapicuíba
>>> Analise e Desenvolvimento de Sistemas e Desenvolvimento de Jogos Digitais
>>> antonio.rcarvalho at fatec.sp.gov.br
>>> antoniorcn at hotmail.com
>>>
>>> Em 25/05/2017 13:24, Ton Roosendaal escreveu:
>>>
>>> Hi,
>>>
>>> When a teacher starts putting tricks in place to avoid cheating he's
>>> losing it. I wouldn't solve the symptom (cheating) but the cause (students
>>> don't like homework or assignments).
>>>
>>> Give them something that relates to them (build your own bedroom) or
>>> makes it personal (give each a different letter of alphabet to do something
>>> with). Think of a challenge involving competition. Or teamwork. And they
>>> should actually learn skills from it. It's the process what counts then,
>>> not the result.
>>>
>>> -Ton-
>>>
>>> --------------------------------------------------------
>>> Ton Roosendaal  -  ton at blender.org   -   www.blender.org
>>> Chairman Blender Foundation, Director Blender Institute
>>> Entrepotdok 57A, 1018 AD, Amsterdam, the Netherlands
>>>
>>>
>>>
>>> On 25 May 2017, at 07:52, Piotr Arłukowicz <piotao at inf.ug.edu.pl> wrote:
>>>
>>> Thanks Mike,
>>> in fact, I already was forced to do such 'heuristics', because lots of
>>> files were 'too' similar.
>>>
>>> However, in case of animation this is rather hard to tell, especially
>>> when students are opening the same file with assignment. I could
>>> potentially solve that telling them to import rather than open, and then
>>> such lovely random string or just something (creation timestamp?) will be a
>>> nice addition to the normal Blender.
>>> How many teachers are still here? Don't have any of you cheating
>>> problems?
>>>
>>> Maybe it's a good idea to create such a plugin, where student can
>>> 'submit' the work right from Blender...
>>>
>>> pio
>>>
>>>
>>> Piotr Arłukowicz, PhD, BFCT
>>> University of Gdańsk, Faculty of Mathematics, Physics and Informatics,
>>> Dept. of AI,
>>> Wit Stwosz 57, 80-952 Gdańsk, room 121, tel.: +48585232151
>>> <+48%2058%20523%2021%2051>, https://inf.ug.edu.pl/~piotao
>>> Polish Blender Course: http://polskikursblendera.pl/ [PL]
>>>
>>> 2017-05-25 0:12 GMT+02:00 Mike Pan <mike.c.pan at gmail.com>:
>>>
>>>> Very interesting question...
>>>>
>>>> You can certainly build an addon that saves a random string of some
>>>> sort into the Blend file. Or even something that tracks the originating
>>>> computer name and total time spent editing the file (to prevent copy+paste
>>>> from blendswap/turbosquid) but the students can always 'forget' to use the
>>>> correct blend version or the addon.
>>>>
>>>> I think a more fool-proof approach might be to analyze the students'
>>>> files based on a bunch of heuristics. This way, the blender file doesn't
>>>> have to be special, but you can still catch copycats.  Here are some things
>>>> you can look at:
>>>> - Datablock names. Especially mesh, material and image names, which is
>>>> something many people don't bother changing.
>>>> - Node positions. Even if the material is identical, chances are the
>>>> nodes are arranged differently. (unless they are using the material panel
>>>> to generate all the nodes)
>>>> - Look at exact value of properties? (eg. If both students are using a
>>>> particle system, unlikely they are both emitting exactly 2740 particles
>>>> from frame 77-333)
>>>> - if an image texture has been packed and not "made relative" yet, it
>>>> might contain the full path of the image, which is telling if it originated
>>>> from another user/computer.
>>>>
>>>> That's all i can think of for now. Hope that helps,
>>>>
>>>> Mike
>>>>
>>>> On Wed, 24 May 2017 at 12:14 Piotr Arłukowicz <piotao at gmail.com> wrote:
>>>>
>>>>> Hi all,
>>>>> is there any way for a teacher to tell whether or not some files in
>>>>> the class, which were collected from an assignment, are copied from the
>>>>> same person or were created on the same computer?
>>>>>
>>>>> I have classes and I have collected quite a few blend files to check.
>>>>> They are similar in few areas, so I'm unsure if they were created by
>>>>> different persons.
>>>>>
>>>>> To solve this problem Julian wrote a small patch few years ago which
>>>>> stored a random number inside blend file (so I could at least tell if
>>>>> somebody copied somebody's else work and modified it slightly), but
>>>>> unfortunately I've got a bunch of files made in just an ordinary, brand
>>>>> blender release 2.78.
>>>>> Files from students are suspiciously similar (for example a
>>>>> manipulator is often set to rotate, not translate).
>>>>>
>>>>> So, is there ANY way to tell?
>>>>> If not, it could be a good idea to introduce just a random number or
>>>>> microsecond stored when file is created and then never changed. This small
>>>>> thing could make life easier and could also detects nasty cheating, which,
>>>>> unfortunately happens too often in some countries :(
>>>>>
>>>>> anybody?
>>>>>
>>>>> regards
>>>>> pio
>>>>>
>>>>> pz
>>>>> piotr
>>>>> --
>>>>> Piotr Arlukowicz
>>>>> -----BEGIN GEEK CODE BLOCK-----
>>>>> Version: 3.1
>>>>> GCS/ED/IT/S d++(-)>--pu s(+):(+)> a C++(+++)$@>++++$  ULAVISC*()$>+++$
>>>>> P++(+++)$>++++ L++(+++)$@>++++$ !E---(---)>++ W++(+++)$@>+++ N(+)>++ o--?
>>>>> !K-(-)>-$ w++(+)>-- !O-(-)>- !M-(-)>-- !V-(-)>- PS(+)>++ !PE()>+  Y PGP>+
>>>>> t(-) !5? !X R()>* tv- b++ DI++ D+(++)>+++ G++@ e++++>+++++ h---()>++
>>>>> r+++ y+++
>>>>> ------END GEEK CODE BLOCK------
>>>>> _______________________________________________
>>>>> Bf-education mailing list
>>>>> Bf-education at blender.org
>>>>> https://lists.blender.org/mailman/listinfo/bf-education
>>>>>
>>>>
>>>> _______________________________________________
>>>> Bf-education mailing list
>>>> Bf-education at blender.org
>>>> https://lists.blender.org/mailman/listinfo/bf-education
>>>>
>>>>
>>> _______________________________________________
>>> Bf-education mailing list
>>> Bf-education at blender.org
>>> https://lists.blender.org/mailman/listinfo/bf-education
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Bf-education mailing listBf-education at blender.orghttps://lists.blender.org/mailman/listinfo/bf-education
>>>
>>>
>>>
>>> _______________________________________________
>>> Bf-education mailing list
>>> Bf-education at blender.org
>>> https://lists.blender.org/mailman/listinfo/bf-education
>>>
>>>
>>
>>
>> --
>> Douglas E Knapp, MSAOM, LAc.
>>
>> _______________________________________________
>> Bf-education mailing list
>> Bf-education at blender.org
>> https://lists.blender.org/mailman/listinfo/bf-education
>>
>>
>
>
> --
> Linux User #412877 - http://counter.li.org/
> Macuare Producciones - http://www.macuare.org
>
> _______________________________________________
> Bf-education mailing list
> Bf-education at blender.org
> https://lists.blender.org/mailman/listinfo/bf-education
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.blender.org/pipermail/bf-education/attachments/20170526/702d4a6b/attachment.html>