views:

96

answers:

2

My job is to spread the word about this free SCJA (Sun Certified Java Associate) course at my school. I'm also receiving all subscribers information such as e-mail, name, an phone no. I already have almost 100 suscribers, and there'll be possibly more in these days. The problem is that the course has limited capacity for just 60 of them.

Consider the following:

  • The course objective is to certificate at least 45 people.
  • Due to the course duration and 'intensive-course' approach, candidates must know the basics of OOP concepts and be familiar with the Java programming language.

So, I have to select the best candidates among these suscribrers somehow. I was thinking about a single 20 question quiz, but that'll be kinda trivial since they'll have access to the Internet and complete correctly this quiz to 100%. On the other hand I could give them a list of requirements to build a little system in Java with some tricky methods (to add a little challenge). I prefer the second (although checking all the candidates entries will be such a pain).

What kind of system will fit for this pre-selection test? What challenging requirements will you add?

Any new ideas to pre-select candidates are very welcome!

+2  A: 

Pick the first 60 to apply, or else a random drawing. Make sure the requirements and expectations are clear. Don't use any criteria where someone can argue about their score.

le dorfier
Pick first is better IMO - this way you're likely to select more organized people that don't delay application and coursework submissions until the last moment.
sharptooth
If I do it this way, a lot of good candidates to potentially acquire the certification will be out of the selection. I know several friends and classmates that knew about this course first and registered almost instantly. But honestly some of them just did it because it was 'free'. Some of them are not that good at Java or even at programming at all, so they probably won't get the certification.At first the selection would be made via first-in-register. However, as I mention, the course objective is to certificate 45 people at least. Therefore the probabilities of success will be very low.
Rigo Vides
I don't think it's realistic to expect to start with any 60 students and have 45 qualify at the end. Maybe you should allow everyone to come to the first lecture (or televise it or videotape it). Then have a quiz at the start of the second lecture. I bet all the unqualified students will drop out before the second lecture. If not, use the quiz.
le dorfier
I think that if the course starts with 60 good candidates that are ready to certificate (and not intended to learn from scratch or go to a programming course) then is realistic to have a 75% successful qualified candidates at the end of the course. That's why a selection prior to the course it's crucial to achieve it's objective. Videotape? That's a really good idea! I'll probably record the course, not for selection, but to share the video on the web (if there are no legal terms or copyright issues from Sun Microsystems) in a future. Thanks for your feedback le dorfier.
Rigo Vides
A: 

If it is possible to get the info, you could use grades or courses taken as a filter.

If not, I definately agree with le dorfier and sharptooth: Pick a simple system that isn't subjective. This saves time both for you and the applicants.

Anders Lindahl
Thanks for your answer Anders. Grades are not a good metric to take IMO, since they don't reflect the real knowledge of the candidates. Anyway, there's no easy way to get that information, since we lack a system where we could check for all the students' academic info. I think is worth to make the suscribers spend some time at a selection process since the course is totally free, and that even includes the certification exam. So we're talking about $300 that the students are saving with this.
Rigo Vides