[dbrowning]

01/02/2018 at 2:08 pm

Are both terms “float” and “real” acceptable to mean a number with a decimal place in the N5 exam or should pupils only use “real” but not “float” or the other way around?

[Scott Leiper]

01/02/2018 at 2:22 pm

I would say both are acceptable.  In the current version of the BCS glossary, page 322 it says,

“Real-type data are numbers that include a fractional part. Internally the computer will usually store real-type data using floating-point representation (see page 319). It is sometimes called float-type.”

The SQA draw their definitions from this text, so I can’t see how they could argue.

[Enrico Vanni]

23/02/2018 at 11:35 am

Bit late to the party to respond to this, but from the available documentation it is fair to say that the SQA’s approach to this is ‘inconsistent’.

 

Marking schemes for previous exams have indicated sometimes that all application specific terms for data types should not be accepted, rejecting ‘single’ as a data type for a real number for example, and at other times application specific terms have been specifically included as acceptable answers – for example permitting ‘object’ and ‘container’ as alternatives for graphic data types.  This second example is now however moribund as graphic data types were culled from the N5 course spec (a kneejerk solution to clear up this anomaly perhaps?), but the inconsistent approach has never been formally addressed.

 

So assume nothing and expect common sense not to prevail.  The safe option is to tell pupils that only terms used in the course spec will be accepted, but that of course penalises those who read round the narrow defines of the course content and give answers that are real-world correct but make (external) marking awkward, less automated and would require the marking skills of and payment of suitable remuneration to someone with broad professional knowledge, given the freedom to exercise their powers of discretion without being scrutinised at every turn, so get no credit.