------------------------------------------------ On Thu, 6 Mar 2003 10:25:22 -0500, Ajit Deshpande <[EMAIL PROTECTED]> wrote:
> On Thu, Mar 06, 2003 at 05:30:06AM -0800, Paris Sinclair wrote: > > [..] > > easily come to the conclusion that it is quite impossible to achieve a test > > that is going to be considered ideal by very many of the involved parties. > > I dont agree with the above.. > > At our company we use a simple perl test as an elimination > tool to weed out the hundreds of resumes we receive. > > It is a one hour test that examines the following key things: > > - A good understanding of perl data-strucutres and references > in particular > > - Simple/basic perl OO skills > > - Basic programming suave (i.e parse a simple file, look > throuhg a hash, print out some values) > > - Awareness level of perl resources like CPAN/perldoc/mailing > lists etc. > > I see a lot of benefit in administering such a test as a > screening device. But you are comparing apples and oranges. Administering a home grown test that you know fits well with the qualifications needed to perform the work in teh specific position you are offering is certainly a valid way to weed out non-acceptable candidates. But on the other hand requiring a 3rd party exam that may or may not adequately test the requirements, is a completely different matter. For instance on the brain bench exam for CGI developers there is a question about the constructor of Net::FTP. While the commonly used constructor of 'new' is correct, there is no reason why a Perl constructor *must* be named new, and someone that is working with CGI may never need to use the Net::FTP module so would not know its particular API. There are sufficiently general questions on the exam as well, but this is an instance where the test questions may not fit the position offered, and in general most test administrators couldn't pass the test themselves, where in the case of your home ! grown test, somewhere there that wrote it, can, and is therefore more likely to be able to gauge the adequacy of prospectives. http://danconia.org
