For this unit, a system called SAFE is used for you to submit coursework, and for me to produce marks and feedback, instead of the University's standard Blackboard system
SAFE stands for System for Admin in the Faculty of Engineering (I was its originator, but I no longer have any responsibility for it)
If you look at the top of the unit page, you will find the link to the section of SAFE for this unit
The next slide shows what that looks like
Follow the link W1
(week 1) circled in red
Follow the circled link, which is a form you have to read and sign once, before submitting for the first time
Sign at the bottom (tick and press button)
Follow the link just under the sign button which says "coursework submission" (or use your browser's back button)
Press "Choose file" then press "Sign and Submit"
You can download files after submitting them to check
You can re-submit with the same filename whenever you have made a better version
Don't submit compiled programs, only source
In schools, there is a common marking scale used where "if you get everything right, you get full marks"
The scale used in universities isn't like that at all
At university, each mark is usually presented out of 100
It is called a percentage, but it isn't a percentage of anything in particular
In the USA and other countries, they use a GPA score
Experience shows that this is exactly the same scale, but it is out of 5 instead of being out of 100
For example, "first class" means a percentage of 70, or a GPA score of 3.5
The idea is that marks from 0 to 100 represent the entire range of human ability and achievement, thought of as a statistical "normal distribution"
A mark of 50 represents an average university student's mark
As Bristol is an above average university, the average mark is expected to be above 50
But since 100 represents the level of the greatest genius who could ever live, marks of 100 are expected to be vanishingly rare
The main thing about the scale is that it is nonlinear
That means the first few marks are easy to get, and higher marks are harder and harder to get
University staff are very familiar with the scale, through subject conventions, experience, and a system of external examiners
A rough rule of thumb is:
At the end of your degree, you will get a grade, based on an average percentage mark - the grades are
There are various lesser awards or credit transfers
Masters degrees use the same scale, except with a 50% pass mark, so getting <50% in an MEng results in a BSc
With IQs there is an increasing emphasis on broad spectrum tests instead of a single linear measure
So too with university degrees there is an increasing emphasis on your transcript (range of marks across subjects) rather than just your grade
A degree is a test of ability, knowledge, skill, creativity, and effort
It is possible that a particular assignment may be easy enough for you to get 100%, but over many assignments in a unit, testing various different skills, the weighted averaging process ensures that the overall average for a unit is usually a fair measure of success
The average of unit marks over a year or a whole degree also balances out unit mark differences
It is acknowledged that there is a small amount of uncertainty left, so there are various borderline handling rules
In a programming unit like this one, the university scale gives us a problem
Programming is, initially anyway, about precision
A program needs to be completely correct, and that is what you should be aiming to achieve
In other words, it would be good if you all got 100%, but that doesn't match the university scale
There are various techniques for handling this issue, including assignments with fuzzy goals, multiple stages, open ended extensions and so on
One technique being tried in this unit, at least with some assignments, is to deliberately split an assignment into two halves
The first half is a totally closed task, and the second half is a totally open-ended task
The closed part is about precision; every detail is provided, so you know exactly what to try to achieve
You are also given in advance all the tests which will be used in marking to measure success
That means you have every chance to get full marks and, failing that, you will know exactly what mark you are going to get when you submit
That means (a) we want everybody to get full marks and (b) we can automate the marking
For the open part, you will be given almost no guidance, you will be free to go in any direction you want
You are expected to use the same skills as in the first half, but creating your own task with your own tests
Your program should pass all your own tests
The marking will be by hand, based on our overall impression of how much advanced ability, knowledge, skill, creativity and effort you demonstrate
We will aim to make your total mark match the university scale
For the open part, you can choose a fairly small and self-contained task, and aim for perfection
Or, you can tackle something bigger or more speculative, but only produce a proof-of-concept prototype
It is perfectly acceptable to stop when you run out of time, and submit what you have done with a description of your aims and how far you got
The nominal amount of time allowed for assignments for this unit is 5 hours per week, based on a notional 40-hour week including lectures, tutorials and so on
You should keep track of whether or not you are exceeding this, and only do more if you are able and willing
Everyone should aim to do the closed parts of the assignments
If you can't get that much done during the allowed 5 hours, you should concentrate on making your programming skills more efficient, getting help where necessary
Don't tackle the open part at all, unless you are confident that you have the closed part sorted, and that you have enough time and energy