Hello,

this is an interesting question. I think we can never avoid cheating, but we can minimize it.

In general, team work has always problems that it is difficult to really assess the work of the individual students correctly. You might include a scheme for peer evaluation, but this has to be handled with care.

Based on my experience it is sometimes easy to find plagiarism between projects when carefully checking  them in detail. We simply see that there are too many similarities. Sometimes it is also easy to find code that students took from the internet, especially if it is "too" elegant or uses elements that were not really introduced during the lectures. In those cases it is important to set clear examples and punish them.... I think this will reduce the number of trials in the future as students will know that they take a risk. But perhaps I am also too optimistic. They have to learn that we can find plagiarism.

Another important element is to ask students to present in detail their solution to you and then you have the possibility to ask detailed questions. Even if they took the solution from somebody else, you can at least ensure that students study it carefully, try to understand it and learn from it.

Just my two cents.

Regards

Your concern for deep-learning is laudable. It takes a lot of work for the professor to structure meaningful rather than surface level education. Have you considered incorporating cooperative learning into the module? Using various assessment devices ranging from peer review to presentation/discussion of solutions in detail as suggested by Christina you could decrease plagiarism while enhancing employer-desired skills.

I am listing a few articles with details that might be of interest.

Porter, L., Guzdial, M., McDowell, C., & Simon, B. (2013). Success in Introductory Programming: What Works?. Communications Of The ACM, 56(8), 34-36. doi:10.1145/2492007.2492020

Ismail, M. (2010). The Effects of Mind Mapping with Cooperative Learning on Programming Performance, Problem Solving Skill and Metacognitive Knowledge among Computer Science Students. Journal Of Educational Computing Research, 42(1), 35-61.

Also interesting is the game based approach from a Research Gate author:

Triantafyllakos, G., Palaigeorgiou, G., & Tsoukalas, I. A. (2011). Designing Educational Software with Students through Collaborative Design Games: The We!Design&Play Framework. Computers & Education, 56(1), 227-242.

Hi,

this is a great question.

In my experience, I have found only two ways to significantly reduce plagiarism:

1. do coursework in labs under examination conditions and tell the students that you will run plagiarism detection software on their submissions. Here the item of coursework must be small enough that it can be done in a 1 or 2 hour lab. Here students cannot really confer to try and solve a problem together. Also, even if the students could in principle use email or chat software to help each other, in my experience time is sufficiently tight that students who are asked for solutions/help by others are reluctant to pass information around as it eats in their own solution time and they run the risk due to lack of time the recipient will not try to understand the solution and redo it but he/she will effectively clone it, with minor changes, that will then be easily detected by the plagiarism software, which would obviously put both the originator and the recipient in trouble. Also, because of the exam conditions,  both originator and recipient will be aware that there is always the risk of being noticed messing around with email/chat, which would of course trigger an investigation.

2. use revision control software to track the progress of the student's coursework, instruct the students that they are expected to work on their assignment over a set period (e.g., 2-3 weeks) and that they need to do regular commits as this will count for a certain percentage of the final mark. In this way the teacher does not only have access to the final product (the submission), but can also see the steps taken by the students in the development of their solutions. Here students can confer and help each other, but they cannot commit the same partial solutions as this would be very noticeable. I've used this strategy for several years for group coursework (in a 50% coursework, 50% exam module) in conjunction with peer evaluation forms, but a colleague here at Essex has tested it in individual assignments in a 100%-coursework module, with success.

Hope this helps.

Riccardo

With reference to plagiarism detection software I would like to issue the warning that we must be very careful when using those. They do not detect all plagiarism and especially there might equally be false positives, so we should not rely on such software alone to detect plagiarism but can use them as one tool to help.

Regards

I think a big question is also related to the learning outcomes. Written examinations can only be used to assess a specific set of learning outcomes while project  based evaluations can assess different learning outcomes. Therefore I think that we need different types of assessments for different learning outcomes and should not focus only one one type like written examination, oral examination, project based work, etc.

Regards

I guess some universities were using turnitin plagarisim detection software ., which works well..

Many thanks for all your feedback. Here are some more thoughts on the topic.

Below is a summary of my teaching philosophy related to student assessment:

Students need to understand boundaries.

• It is important that students learn to respect boundaries.This includes amongst others understanding that cheating and plagiarism are unethical and that deadlines need to be met. Classes should be provided to prevent these things from happening.

The assessment needs to be meaningful, fair, and transparent.

• It is vital to consider the necessity of each assessment carefully to avoid that students and lecturers get overloaded with work. Care needs to be taken to prevent unethical behaviour. Sufficient feedback needs to be provided so that students can see that they have been fairly treated.

And here are the actions I am planning to apply in the next semester:

I have to clarify what is seen as plagiarism (copying other people's work) and what the consequences are. The dilemma I have is that my general teaching philosophy is to promote teamwork while the assessment in this module requires individual work. The boundaries between allowed collaboration and cheating (job sharing) need to be clear and students need to be aware of these boundaries. I picked up some ideas about how to avoid plagiarism and cheating from a German TV show that seem to provide a good starting point (http://www.zdf.de/ZDFmediathek#/beitrag/video/2241408).

• Just before the coursework is released provide a session focussing on cheating and plagiarism: defining the boundaries and illustrating the consequences.
• At the end of this session ask students to sign a contract that they will not cheat or plagiarise. Doing this in the class room directly after the training will increase the impact.
• On the form and during the session on plagiarism the consequences for students if they plagiarise will clearly pointed out. Students need to know what the consequences of their actions might be so that they can avoid these.

The module runs in Spring 2015 - so still some time before I can find out if this has a positive effect on avoiding cheating and plagiarism. Perhaps you have a chance to try it earlier. It would be great if you could then share your experience :-).

I have now written a working paper related to the topics discussed above: "Changing the Assessment Strategy for the Computer Science Module Simulation for Decision Support: A Longitudinal Study".

http://www.cs.nott.ac.uk/~pos/docs/pos-PGCHE-WP-v1.0.pdf

Very apt question in an era... Where we have to either publish or perish..

Have a look at article.. 

Hope it helps

http://www.tandfonline.com/doi/pdf/10.1080/08989621.2015.1084472