[Year 12 SofDev] SD Exam - Question 4

[Andrew Shortell]

Null  - if there is nothing then what is the point in proceeding further
Type ­ if there is other than numbers then no point in doing any other
checks
Length ­ no point in checking the database if there is incorrect number of
digits
Existence ­ checking the database is a significant step which goes outside
of the current program to an external source

I have not yet seen the exam.. It is not offered here but as a teacher of
IS/SoD for a few years and lectured to 3rd year programmers this is how I
would approach it.

Effectiveness in terms of why go to the next step when it has failed a
step...
It would be pretty silly to put it in the reverse order would n¹t it...
Check against the database before you even know if anything has been
entered? Etc

Andrew


-- 
Andrew Shortell 
(soon to leave Wallan SC)



On 17/11/10 1:10 PM, "Mark KELLY" <kel at mckinnonsc.vic.edu.au> wrote:

> Hi all.  I've started getting a bee in my bonnet (or a terrier in my trousers)
> about Section C, Q4.
> 
> Percy is writing the program specification. For the section dealing with entry
> of the Medicare number he writes
> 
> If the user makes a mistake entering the Medicare number, the program has to
> provide as much information as possible about what is wrong. Therefore include
> these validation tests.
> Length - to make sure that the correct number of characters has been entered
> Existence - to make sure that the number entered exists on Medicare's database
> Null - to make sure something has been entered
> Type - to make sure that only numeric characters have been entered
> Note: These tests must be performed in the most effective order.
> 
> a. Place these validation tests in the most effective order. 1 mark
> 
> Then students have to give 3 reasons why it's the most effective way of doing
> those tests. (3 marks)
> 
> 
> Hmm. What is "effective" about the order? It must relate to how well it does
> the job of finding invalid numbers.
> 
> But if all 4 tests are going to be conducted, any order of the 4 tests will
> yield exactly the same results.
> 
> It might be different if we wanted to find invalid numbers soonest, or with
> the minimum number of tests, but they are efficiency criteria, not
> effectiveness.
> 
> I'm seriously starting to believe that this is a Bull Mastiff of a question...
> All orders must be equally effective.
> 
> I can usually work out what the examiners were probably trying to get at, but
> this one has me completely baffled (like multichoice Q3 on the ITA paper). 
> I have no idea whatsoever of what knowledge the examiners are hoping I will
> show.
> 
> So, can anyone please tell me:
> 
> 1. how one order can be more effective than another.
> 2. what order is most effective, and why the reverse order would be less
> effective!
> 
> Growl.
> 
> Mark

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.edulists.com.au/pipermail/sofdev/attachments/20101117/fc9e44d4/attachment.html