CS135 Marking Schemes

For every assignment you will have to create/modify three files. These three files should completed before the marking meeting and .

Many versions of these files are archived, so take advantage of this opportunity to modify the previous marking scheme files to suit your assignment. This is an especially good idea if your assignment builds off a previous term's assignment or utilizes questions from previous assignments.

Marking Scheme

The marking scheme refers to the document that you will distribute and cover with the graders during the marking meeting. This file should contain more detailed criteria explantions and marking intructions for graders than the marking rubric that will uploaded to MarkUs. This is usually the first document you write during the pre-marking process. Most of the previous marking schemes found in the archive are written using Microsoft Word and converted to pdf for distribution to markers.

Style criteria are often maintained from assignment to assignment and require very few changes unless the style guide for the course gets major changes between assignments.

Contracts

Especially in later assignments, there may be many acceptable contracts. Students may define their own types, or may use literal definitions for compound types (Eg. (listof (list Nat Str)) instead of AL). You must be careful to cover as many of these as possible when writing the marking scheme. The Fall 2014 marking schemes have good examples of the most current format used for this.

It may be useful to look through student's submissions in the handin folder, or egrep for incorrect contracts that you may suspect students will use. This will give you a more complete idea of what needs to be mentioned in your marking scheme (no pun intended).

To find (mostly) all contracts for a single function: (run from handin/a0x directory)

egrep -r --include="FILENAME.rkt" -m 1 -i "^.*;; *FUNCTION-NAME: *" *

Replace FILENAME and FUNCTION-NAME appropriately. All options after -r are optional but greatly increase the speed of the search.

Further, if you are interested in a more concise output, this variation will be more helpful:

egrep -r --include="FILENAME.rkt" -m 1 -i "^.*;; *FUNCTION-NAME: *" * | tr A-Z a-z | perl -pi -e 's/:/: /g' | perl -pi -e 's/->/ -> /g' | tr -s " -" | sed 's/ $//' | sort | uniq

It attempts to remove all duplicates by standardizing spacing and capitalization. Keep in mind that this converts all characters to lowercase, so types such as "Nat" and "Int" will appear to be incorrect.

To find and sort all contracts for all functions: (run from handin/a0x directory)

egrep -r --include="*.rkt" -i -h "^.*;; *(\w|-)+: *\w*.*->" * | sort

The "| sort" can be removed to view the unsorted output in real time, but this is generally not useful.

For concise output, as above:

egrep -r --include="*.rkt" -i -h "^.*;; *(\w|-)+: *\w*.*->" * | tr A-Z a-z | perl -pi -e 's/:/: /g' | perl -pi -e 's/->/ -> /g' | tr -s " -" | sed 's/ $//' | sort | uniq

Marking Rubric

There are several MarkUsScripts useful for this course. For generating the marking rubric to upload to MarkUs, the buildrubric script can be utilized to do the bulk of the work for making a new rubric.

This rubric is a csv file containing the criteria outlined in the marking scheme as well as the weighting and marking levels. A general criteria in this file has the format

Criteria Name,Weight,Very Poor,Weak,Passable,Good,Excellent,Very Poor level description.,Weak level description.,Passable level description.,Good level description.,Excellent level description.

and will have multiple lines such as the one above separated by a newline.

Annotations

-- AdamLupton - 2014-12-17