Advice on matters related to the administration of Victorian Certificate of Education (VCE) assessment is published annually in the
VCE and VCAL Administrative Handbook. Updates to matters related to the administration of VCE assessment are published in the
VCAA Bulletin.
Teachers must refer to these publications for current advice.
Excepting third-party elements, schools may use this resource in accordance with the
VCAA's Educational Allowance (VCAA Copyright and Intellectual Property Policy).
When designing learning activities for the School-assessed Task, teachers will refer to the areas of study and outcomes, including key knowledge and key skills, as listed in the VCE Algorithmics (HESS) Study Design.
The Algorithmics (HESS) School-assessed Task is in three parts involving Unit 3 Outcome 3, Unit 4 Outcome 1 and Unit 4 Outcome 2.
The School-assessed Task contributes 20 per cent to the study score for Algorithmics (HESS). Details of the assessment task can be found on page 18 of the
VCE Algorithmics (HESS) Study Design.
Teachers must be aware of the requirements for the authentication of the
VCE Algorithmics (HESS) School-assessed Task as specified in the Assessment: School-based Assessment section of the
VCE and VCAL Administrative Handbook.
The VCAA conducts professional learning for teachers of Algorithmics (HESS) in February / March of each year through webinars and on-demand videos. These are known as SAT Webinars and SAT On-demand videos. Details of these professional learning sessions are advertised in the November edition of the
VCAA Bulletin each year.
There should be a single, coherent problem. It is a good idea to choose a problem with a response that constitutes a 'good' response that is not immediately obvious. Canonical problems with widely published ‘best’ solutions are not appropriate.
There should be an obvious place to start for the data model, such as a simple graph-based solution, and there should also be a range of sophisticated data model solutions which require the combination of a graph with other abstract data types.
There should be an obvious place to start for the algorithm and there should also be a range of sophisticated algorithmic solutions that span several of the studied algorithms. There must be scope to apply the advanced algorithm design techniques that are studied in Unit 4 to the problem, such as divide and conquer, dynamic programming, backtracking or heuristic methods.
No elements of the solution, such as a choice of data structures, should be specified in the problem description.
The first part of the School-assessed Task requires students to create an algorithmic solution for a real-world problem. Students use the studied ADTs in combination to specify how the information in the problem will be abstracted and they design algorithms incorporating iterative and recursive functions in structured pseudocode to solve the problem. Students may use, adapt and combine the studied graph algorithms in their solution to the problem as appropriate. The prompt for the School-assessed Task should describe a real-world computational problem. The chosen problem should be one that has scope for both simple and complex solutions. Students are expected to consider several possible solutions and from these to select the most fit-for-purpose design. Students evaluate the suitability of their solutions and justify their final design drawing on the requirements of the original problem.
Students are not expected to consider the efficiency of their design in the Unit 3 Outcome 3 part of the School-assessed Task. In Unit 4, students will be required to develop more efficient solutions and hence it is helpful if students have not tried to address this earlier.
Detailed example – SAT Unit 3 Outcome 3
Design an algorithm that recommends a new friend to someone in a social network.
Students need to decide on the data structures used to store the information in their social network and then create their own algorithms to create recommendations for new friends. Many of the studied graph algorithms, such as PageRank and shortest path algorithms can be applied to this problem.
Sample data models
A friends of friends model with likes
Figure 8
Image description
A proximity model
Figure 9
Image description
- Students design the ADTs to support the social network information with justification given of the usefulness of their approach for the purpose.
- Students design algorithms in pseudocode to use the information in the model with justification of logic and of the inclusion of any of the studied graph algorithms.
- Students explain the meaning of the solution in context of the original problem.
When selecting a prompt for the Unit 3 Outcome 3 part of the School-assessed Task, the teacher must take into consideration that the students will be required to further develop the efficiency of their solution in Unit 4.
Teachers should have at least one checkpoint for students to submit their work on the School-assessed Task for this section. This might, for example, include just the first two elements of this part of the School-assessed Task. Checkpoints support students to complete the task within the allocated time and supports the authentication of student work.
Issues identified after marking Unit 3 Outcome 3
If the design of the data model and algorithm combination to solve a real-world / applied problem in Unit 3 Outcome 3 is incomplete or contains significant errors, students may not have a functioning algorithm to analyse for the Unit 4 Outcome 1 component. In this situation, the teacher may need to provide the student with a minimal functional solution to the Unit 3 Outcome 1 task for the student to analyse; for example, a brute force search solution. The modified design is not reassessed, however, this prevents negative consequential effects for the second component of the SAT in Unit 4 Outcome 1.
Unit 4: Area of Study 1 – Formal algorithm analysis
The second part of the School-assessed Task requires students to analyse the time complexity of the simple algorithm design they created in the Unit 3 Outcome 3 School-assessed Task component. This will require students to establish the efficiency of their algorithm and to explain the real-world consequences of this on the feasibility of using their algorithm within the real-world context. Students may use call graphs, recurrence relations, Master Theorem (where / if appropriate) and Big-O notation. Students can evaluate the suitability of algorithms / modules based on their time complexity and recognise common features in their simple algorithms that have the same time complexity.
Detailed example – SAT Unit 4 Outcome 1
Analysis of a
FriendshipRecommenderAlgorithm from pseudocode and/or coded and created in Snap/Edgy from Unit 3 Outcome 3.
The call graph of Friendship Recommender Algorithm
Figure 10
Image description
Modules defined in pseudocode and modules/blocks coded
FriendshipRecommenderAlgorithm in
Edgy:
-
BuildFriendshipModel
-
InitialiseProximityOnEdges
-
AdjustProximityByHobbies
-
AdjustProximityByDistanceApart
-
FloydWarshallShortestPaths
-
FriendRecommender
Detailed documentation on the use of
Edgy.
Algorithm Friendship
The pseudocode shows the block actions for the
FriendRecommenderAlgorithm that can be coded in Edgy.
When green flag clicked
BuildFriendshipModel
InitialiseProximityOnEdges
For each edge of all the edges do
AdjustProximitybyHobbies(item 1 of edge, item 2 of edge)
AdjustProximitybyDistanceApart(item 1 of edge, item 2 of edge)
FloydWarshallShortestPaths
FriendRecommender(FWDist,70)
Builds a complete graph for a network of people where the nodes are the people and the edges represent proximity
BuildFriendshipModel.
Initialises the proximity on all the edges of the complete graph
InitialiseProximityOfEdges.
For each edge in the person network:
- compare to all other people and reduce the proximity when people have hobbies in common as they are likely to be friends
AdjustProximityByHobbies
- compare to all other people and add the straight line distance to the current proximity as friendship for people further apart may be harder to maintain
AdjustProximityByDistanceApart.
Use FloydWarshall to work out the shortest proximity between all pairs of people
FloydWarshallShortestPaths.
Based on a shortest distance proximity threshold value between pairs of people, make friend recommendations if they are not already friends
FriendRecommender.
BuildFriendshipModel – Module
The pseudocode shows the blocks which define a function called
BuildFriendshipModel, which has no input parameters, that can be coded in Edgy.
BuildFriendshipModel
script variables p1, p2
set FriendsList to list {list {Ann, Cal}, list {Cal, Ben}, list {Ann, Tom}}
set PersonInterestsDictionary to dictionary {{key=Ann, value=list{Art,Cooking,Gaming}, {key=Cal,value=list{Gaming, Running,Soccer}}, {key=Ben,value=list{Cooking,Gaming}}, {key=Sue, value=list{Art,Soccer}},{key=Tom,value=list{Running,Art}}
Set PersonXYLocationDictionary to dictionary {{key=Ann,value=”5,10”}, {key=Cal,value=”10,2”}, {key=Ben,value=”7,9”}, {key=Tom,value=”3,20”}, key=Sue,value=”7,2”}}
new graph
For each pair of FriendsList do
set p1 to item 1 of pair
set p2 to item 2 of pair
add edge p1,p2
set relationship attribute of edge p1,p2 to friends
For each node1 of all the nodes do
For each node2 of all the node do
If not (node1 equals node2) then
If not (edge node1,node2 exists) then
add edge node1,node2
Set relationship attribute of edge node1,node2 to unknown
Builds a complete graph for a network of people.
Sets up existing friendships.
Sets up peoples’ hobbies.
Sets up peoples’ locations.
Time complexity
Creates various structures to hold the information for the algorithm. Each person is represented by a node in a complete graph.
O(|V|) + O(|V|) + O(|V|)
O(|V|2) to create a complete graph with proximity on edges.
InitialiseProximityOnEdges – Module
The pseudocode shows the blocks which define a function called
InitialiseProximityOnEdges, which has no parameters, that can be coded in Edgy.
InitialiseProximityOnEdges
For each edge of all the edges do
If the relationship attribute of the edge equals friends then
set color of the edge to red
set label-color of the edge to red
set label of edge edge to 50
Else
set label-color of the edge to green
set label of the edge to 100
Uses the complete graph of connections between all pairs of people and initial values for friends and unknowns.
Time complexity
Iterates through all edges of the graph once.
O(|E|) On a complete graph this is equivalent to O(|V|2).
AdjustProximityByHobbies – Module
The pseudocode shows the blocks which define a function called
AdjustProximityByHobbies, which has two parameters person1 and person2, that can be coded in Edgy.
AdjustProximitybyHobbies(person1,person2)
script variables reduceProximityLinear
Set reduceProximityLinear to 0
If the dictionary PersonInterestsDictionary contains both person1 and person2 then
set p1Hobbies to PersonInterestsDictionary values where key=person1
set p2Hobbies to PersonInterestsDictionary values where key=person2
For each hobby of p1Hobbies do
If p2Hobbies contains the hobby then
Add 20 to reduceProximityLinear
If (reduceProximityLinear > 0) then
set label of edge person1,person2 to label minus reduceProximityLinear
set color of edge person1,person2 to cyan
Examines hobbies in common between two persons as input and decreases proximity where common hobbies exist.
Time complexity
Checks the hobbies of person1 and compares these to person2’s hobbies to adjust proximity of persons.
Assume each person has a small number, say k hobbies O(k).
AdjustProximitybyDistanceApart – Module
The pseudocode shows the blocks which define a function called
AdjustProximitybyDistanceApart, which has two parameters person1 and person2, that can be coded in Edgy.
AdjustProximitybyDistanceApart(person1, person2)
script variables x1, y1, x2, y2, cords
set EuclidianDistanceApart to 0
set coords to PersonXYLocationDictionary values where key=person1
set x1 to item 1 of split coords by comma
set y1 to item 2 of split coords by comma
set coords to PersonXYLocationDictionary values where key=person2
set x2 to item 1 of split coords by comma
set y2 to item 2 of split coords by comma
set EuclidianDistanceApart to
if (EuclidianDistanceApart > 5) then
set label of edge person1,person2 to label + EuclidianDistanceApart
set color of edge person1,person2 to purple
Works out distance apart based on (x,y) coordinations of a person’s home.
Time complexity
Checks the Euclidean distance of person1 to person2 to adjust the proximity of persons. A constant amount of work is done in this module. O(1).
FloydWarshallShortestPaths – Algorithm adapted for this model
The pseudocode shows the blocks which define a function called
FloydWarshallShortestPaths, which has no input parameters, that can be coded in Edgy.
FloydWarshallShortestPaths
script variables row, i, j, distanceViak
set i to 0
set FWNodeMap to empty list
For each item of keys in dictionary PersonXYLocationDictionary do
Add 1 to i
Add list {i, item} to FWNodeMap
End for
Set FWDist to empty list
Repeat length of dictionary PersonXYLocationDictionary times
Set row to empty list
Repeat length of dictionary PersonXYLocationDictionary times
Add infinity to row
End Repeat
Add row to FWDist
Set row to empty list
End Repeat
set i to 0
for each fromnode of keys in dictionary PersonXYLocationDictionary do
add 1 to i
set j to 0
for each tonode of keys in dictionary PersonXYLocationDictionsry do
add 1 to j
if (fromnode equals tonode) then
replace item j of item i of FWDist with 0
else
replace item j of item i with label of edge fromnode,tonode
End if
End for
End for
For k = 1 to length of FWNodeMap do
For i = 1 to length of FWNodeMap do
For j=1 to length of FWNodeMap do
Set distanceViak to (item k of item i of FWDist) plus (item j of item k of FWDist)
If item j of item i of FWDist > distanceViak
Replace item j of item i of FWDist with distanceViak
Time complexity
The Floyd Warshall Shortest path algorithm is a well known graph algorithm O(|V|3) that computes the shortest distance between all node pairs (people pairs).
FriendRecommender – Model
The pseudocode shows the blocks which define a function called
FriendRecommender, which has two input parameters FWDist and Dist, that can be coded in Edgy.
FriendRecommender(FWDist,Dist)
script variables closeness, person1, person2
set FriendshipsRecommendedList to empty list
For I = 1 to length of FWNodeMap do
For j = 1 to length of FWNodeMap do
set closeness to item j of item I of FWDist
If (not (i=j)) and (closeness < Dist) then
set person1 to item 2 of item I of FWNodeMap
set person2 to item 2 of item j of FWNodeMap
If (not relationship of edge person1,person2 equals friends) then
add list {person1,person2,”FW ShortestPath”} to FriendshipsRecommendedList
Based on shortest distance proximity between pairs of people who are unknown to each other recommends friendships.
Time complexity
The FriendRecommender module uses the FloydWarshall shortest path distance output with a threshold distance input to recommend new friends for people in the model that are represented by graph nodes.
Has a double nested loop on input size of |V| which is the number of nodes (people) in the friendship network, which is O(|V|2).
Time complexity analysis in summary for the whole Friendship Recommender algorithm
|
BuildFriendshipModel | 2 × O(|V|) + O(|V|2) to create a complete graph with proximity on edges |
InitialiseProximityOnEdges | Runs on a complete graph O(|E|)
Which is equivalent to O(|V|2) |
AdjustProximitybyHobbies | Runs O(k) for each edge on a complete graph, where k is the number of hobbies, assumed constant therefore |V|2 × O(k) => O(|V|2) |
AdjustProximitybyDistanceApart | Run for each edge on a complete graph
|V|2 × O(1) = O(|V|2) |
FloydWarshallShortestPaths | O(|V|3) |
FriendRecommender | O(|V|2) |
TALLY of modules | 2O(|V|) + O(|V|2) + O(|V|2) + O(|V|2) + O(|V|2) + O(|V|3) + O(|V|2)
|
OVERALL Time Complexity for the total solution Algorithm | O(|V|3) since the contribution by the FloydWarshallAlgorithm is the highest order polynomial. |
Conclusion | The amount of work done by this algorithm to compute Friend Recommendations is a cubic function of the number of people represented as nodes in a graph model. |
Issues identified after marking Unit 4 Outcome 1
If the formal time complexity analysis of the designed algorithm for the applied problem in Unit 4 Outcome 1 is incomplete or contains significant errors, students have the opportunity to make adjustments to their analysis. Teachers can provide feedback on the quality of the analysis but the adjustments must be student initiated, not teacher directed. The modified analysis is not reassessed. However, this opportunity prevents negative consequential effects for the third part of the School-assessed Task in Unit 4 Outcome 2.
The third part of the School-assessed Task requires students to use advanced algorithm designs to improve the efficiency of computation of their Unit 3 Outcome 3 School-assessed Task solution. Students will generally design solutions for the same problem as solved in Unit 3 Outcome 3. At times, adding an additional constraint to the problem may be appropriate so that students are required to apply the advanced algorithm design techniques that they have studied.
Detailed example – SAT Unit 4 Outcome 2
An example of a constraint added to a Unit 3 School-assessed Task Project could be a ‘friend recommender algorithm’ based in a social network model optimised for a given ‘distance’ between friends with added constraints.
Example Constraint added for Unit 4 Outcome 2
Recommended: friends must live at most 30 minutes away by car or public transport and have a friend in common or at least two likes in common.
Students can use algorithmic design patterns of divide and conquer, dynamic programming, backtracking and heuristics and randomised searches to improve the efficiency of the Unit 3 Outcome 3 School-assessed Task solution. Students can explain the limitations and benefits of selected advanced algorithm design and / or heuristic approaches. They can compare different approaches and the suitability or not of their School-assessed Task data model and algorithm.
| Outcome | Assessment task | Study score contribution |
|---|
| Unit 3 Outcome 1 |
School-assessed Coursework
In response to given stimulus material, create one or more designs of a data model using abstract data types to capture the salient aspects of a real-world information problem. | 6% |
| Unit 3 Outcome 2 |
School-assessed Coursework
In response to given stimulus material: - create one or more designs of algorithms that apply algorithm design patterns; or select appropriate graph algorithms to solve information problems
- implement an algorithm.
| 6% |
| Unit 3 Outcome 3 |
School-assessed Task – Part 1
The design of a data model and algorithm combination to solve a real-world / applied problem, including:
- a specification of the problem
- a consideration of multiple solution options
- the selection of a suitable, coherent, clear and fit-for-purpose solution.
| 8% |
| |
Unit 3 Total | 20% |
| Unit 4 Outcome 1 |
School-assessed Task – Part 2
A formal time complexity analysis of the designed algorithm for the applied problem, and an explanation of the consequences of these results on the algorithm’s real-world application. | 6% |
| Unit 4 Outcome 2 |
School-assessed Task – Part 3
A design of an improved data model and algorithm combination to solve the applied problem, including:
- the selection of an efficient, coherent and fit-for-purpose solution
- a time complexity analysis
- a comparison to the original solution.
| 6% |
| Unit 4 Outcome 3 |
School-assessed Coursework
At least one task from the following:
- a response to a case study or stimulus material
- a written report
- an annotated visual report
- an oral report
- structured questions.
| 8% |
| |
Unit 4 Total | 20% |
Unit 3 Sample approaches to developing an assessment task
Area of Study 1
On completion of this unit the student should be able to define and explain the representation of information using abstract data types, and devise formal representations for modelling various kinds of real-world information problems using appropriate abstract data types.
The
VCE Algorithmics (HESS) Study Design pages 9–10 provides details of the key knowledge and key skills related to Unit 3 Outcome 1 and the corresponding area of study, Data modelling with abstract data types.
Teachers should be familiar with the area of study and outcome statement, and the relevant key knowledge and key skills in order to plan for the assessment task. It should be noted that the assessment task does not have to identify every key knowledge and key skill dot point, nor should the task focus on too narrow a range of key knowledge and key skills.
When developing assessment tasks, teachers should refer to the VCAA policies and school assessment procedures as specified in the
VCE and VCAL Administrative Handbook section:
Scored assessment: School-based Assessment.
Teaching and learning activities should be selected to enable students to demonstrate their understanding of the key knowledge and key skills. These activities should include a range of theoretical and practical activities to develop and extend student knowledge. Sample
teaching and learning activities are included in this Support material for teachers.
Students should be advised of the timeline and conditions under which the task is to be completed. The assessment task must directly assess the student’s understanding of the key knowledge and key skills as well as their ability to apply these to the assessment task. Due dates and duration of assessment are school-based decisions.
Students should be given instructions regarding the requirements of the task, including time allocation, format of student responses and the marking scheme / assessment criteria. The marking scheme / assessment criteria used to assess the student’s level of performance should reflect the VCAA Performance descriptors for
Unit 3 Outcome 1.
The approach that teachers employ to develop their assessment task may follow the steps below:
- Determine the number of activities that students will be required to complete.
- Select the key knowledge and key skills to determine the content to be used in the assessment task. Consider the
performance descriptors.
- After developing the task check that the task is accessible and provides students with sufficient opportunities to demonstrate their ability to meet the requirements of the outcome.
- Teachers may develop their own marking schemes for this outcome, provided they are consistent with the VCAA Performance descriptors for
Unit 3 Outcome 1. To be consistent with the performance descriptors, teacher-generated schemes must take into account the following points:
- explanation of the roles of abstract data types for data modelling
- reading and writing abstract data types signature specifications
- using abstract data types in accordance with specifications
- identifying and describing the properties of graphs
- applying abstract data types to model real-world problems by selecting appropriate abstract data types and justifying their suitability
- modelling basic network and planning problems with graphs.
Teacher judgment should be used to determine the weighting of each criterion within the School-assessed Coursework task. While weightings are not explicit within the VCAA Performance descriptors, teachers must understand that the criteria are not intended to be equally weighted.
- Teachers should first complete the assessment task themselves by creating designs of a data model in response to the given stimulus material. This will assist them to understand how students may respond to the stimulus material. It will further ensure that marking schemes are appropriate for the task being provided. Refinements to the task may occur as a result of this process.
Schools may determine the conditions for assessment tasks. Assessment tasks should be a part of the regular teaching and learning program and should not add unduly to student workload. Students should be advised of the timeline and conditions under which the task is to be completed. It is recommended that assessment tasks be completed in class under supervision within a limited time frame.
The overall assessment program for the unit should include a variety of activities, include provision for authentication of student work and take into consideration the overall workload for students.
The teacher must decide the most appropriate time and conditions for conducting this assessment task and inform the students ahead of this date. This decision is a result of several considerations including:
- the estimated time it will take to teach the key knowledge and key skills for the outcome
- the likely length of time required for students to complete the task
- the classroom environment in which the assessment task will be completed
- whether the assessment task will be completed under open-book or closed-book conditions
- any additional resources required by students
- when tasks are being conducted in other subjects and the workload implications for students.
The teacher must consider the authentication strategies relevant for each assessment task. Information regarding VCAA authentication rules can be found in the
VCE and VCAL Administrative Handbook section:
Scored assessment: School-based Assessment.
The VCAA Performance descriptors can be used and adapted to the specifics of the task to assess a student’s level of performance. The assessment tools (performance descriptors, rubrics and / or marking guide) should reflect the outcome, key knowledge and key skills. The assessment task and assessment tools should be explained to students before they commence the task.
The VCAA
VCE Assessment principles underpin all VCE assessment practices.
The
performance descriptors in the Support materials give a clear indication of the characteristics and content that should be apparent in a student response at each level of achievement from very low to very high. The specified assessment task is listed on page 13 of the
study design. The assessment task is to be out of 50 marks.
Area of Study 2
On completion of this unit the student should be able to define and explain algorithmic design principles, design algorithms to solve information problems using basic algorithm design patterns, and implement the algorithms.
The
VCE Algorithmics (HESS) Study Design pages 10–11 provides details of the key knowledge and key skills related to Unit 3 Outcome 2 and the corresponding area of study, Algorithm design.
Teachers should be familiar with the area of study and outcome statement, and relevant key knowledge and key skills in order to plan for the assessment task. It should be noted that the assessment task does not have to identify every key knowledge and key skill dot point, nor should the task focus on too narrow a range of key knowledge and key skills.
When developing assessment tasks, teachers should refer to the VCAA policies and school assessment procedures as specified in the
VCE and VCAL Administrative Handbook section:
Scored assessment: School-based Assessment.
Teaching and learning activities should be selected to enable students to demonstrate their understanding of the key knowledge and key skills. These activities should include a range of theoretical and practical activities to develop and extend student knowledge. Sample
teaching and learning activities are included in this Support material for teachers.
Students should be advised of the timeline and conditions under which the task is to be completed. The assessment task must directly assess the student’s understanding of the key knowledge and key skills as well as their ability to apply these to the assessment task. Due dates and duration of assessment are school-based decisions.
Students should be given instructions regarding the requirements of the task, including time allocation, format of student responses and the marking scheme / assessment criteria. The marking scheme/assessment criteria used to assess the student’s level of performance should reflect the VCAA Performance descriptors for
Unit 3 Outcome 2.
The approach that teachers employ to develop their assessment task may follow the steps below:
- Determine the number of activities that students will be required to complete.
- Select the key knowledge and key skills to determine the content to be used in the assessment task. Consider theperformance descriptors.
- After developing the task check that the task is accessible and provides students with sufficient opportunities to demonstrate their ability to meet the requirements of the outcome.
- Teachers may develop their own marking schemes for this outcome, provided they are consistent with the VCAA Performance descriptors forUnit 3 Outcome 2. To be consistent with the performance descriptors, teacher-generated schemes must take into account the following points:
- interpreting pseudocode and executing it manually on given input
- writing pseudocode
- identifying and describing recursive, iterative, brute-force search and greedy patterns within algorithms
- designing recursive and iterative algorithms
- designing algorithms by applying the brute-force search or greedy algorithm design patterns
- writing modular algorithms using abstract data types and functional abstractions
- selecting appropriate graph algorithms and justifying the choice based on their properties and limitations
- explaining the correctness of the specified graph algorithms
- using search methods on decision trees and graphs to solve planning problems
- implementing algorithms, including graph algorithms, as computer programs in a very high-level programming language that directly supports a graph abstract data type
- demonstrating the correctness of simple iterative or recursive algorithms using structured arguments that apply the methods of induction or contradiction.
Teacher judgment should be used to determine the weighting of each criterion within the School-assessed Coursework task. While weightings are not explicit within the VCAA Performance descriptors, teachers must understand that the criteria are not intended to be equally weighted.
- Teachers should first complete the assessment task themselves by creating designs of algorithms in response to the given stimulus material. This will assist them to understand how students may respond to the stimulus material. It will further ensure that marking schemes are appropriate for the task being provided. Refinements to the task may occur as a result of this process.
Schools may determine the conditions for assessment tasks. Assessment tasks should be a part of the regular teaching and learning program and should not add unduly to student workload. Students should be advised of the timeline and conditions under which the task is to be completed. It is recommended that assessment tasks be completed in class under supervision within a limited time frame.
The overall assessment program for the unit should include a variety of activities, include provision for authentication of student work and take into consideration the overall workload for students.
The teacher must decide the most appropriate time and conditions for conducting this assessment task and inform the students ahead of this date. This decision is a result of several considerations including:
- the estimated time it will take to teach the key knowledge and key skills for the outcome
- the likely length of time required for students to complete the task
- the classroom environment in which the assessment task will be completed
- whether the assessment task will be completed under open-book or closed-book conditions
- any additional resources required by students
- when tasks are being conducted in other subjects and the workload implications for students.
The teacher must consider the authentication strategies relevant for each assessment task. Information regarding VCAA authentication rules can be found in the
VCE and VCAL Administrative Handbook section:
Scored assessment: School-based Assessment.
The VCAA Performance descriptors can be used and adapted to the specifics of the task to assess a student’s level of performance. The assessment tools (performance descriptors, rubrics and / or marking guide) should reflect the outcome, key knowledge and key skills. The assessment task and assessment tools should be explained to students before they commence the task.
The VCAA
VCE Assessment principles underpin all VCE assessment practices.
Theperformance descriptors in these Support materials for teachers give a clear indication of the characteristics and content that should be apparent in a student response at each level of achievement from very low to very high. The specified assessment task is listed on page 13 of the
study design. The assessment task is to be out of 50 marks.
Unit 4 Sample approaches to developing an assessment task
Area of Study 3
On completion of this unit the student should be able to explain the historical context for the emergence of computer science as a field and discuss modern machine learning techniques and the philosophical issues they raise.
The
VCE Algorithmics (HESS) Study Design pages 16–17 provides details of the key knowledge and key skills related to Unit 4 Outcome 3 and the corresponding area of study, Computer science: past and present.
Teachers should be familiar with the area of study and outcome statement, and relevant key knowledge and key skills in order to plan for the assessment task. It should be noted that the assessment task does not have to identify every key knowledge and key skill dot point, nor should the task focus on too narrow a range of key knowledge and key skills.
When developing assessment tasks, teachers should refer to the VCAA policies and school assessment procedures as specified in the
VCE and VCAL Administrative Handbook section:
Scored assessment: School-based Assessment.
Teaching and learning activities should be selected to enable students to demonstrate their understanding of the key knowledge and key skills. These activities should include a range of theoretical and practical activities to develop and extend student knowledge. Sample
teaching and learning activities are included in this Support material for teachers.
Students should be advised of the timeline and conditions under which the task is to be completed. The assessment task must directly assess the student’s understanding of the key knowledge and key skills as well as their ability to apply these to the assessment task. Due dates and duration of assessment are school-based decisions.
Students should be given instructions regarding the requirements of the task, including time allocation, format of student responses and the marking scheme / assessment criteria. The marking scheme / assessment criteria used to assess the student’s level of performance should reflect the VCAA
Performance descriptors for Unit 4 Outcome 3.
The approach that teachers employ to develop their assessment task may follow the steps below:
- Determine the number of activities that students will be required to complete.
- Select the key knowledge and key skills to determine the content to be used in the assessment task. Consider theperformance descriptors.
- After developing the task, check that the task is accessible and provides students with sufficient opportunities to demonstrate their ability to meet the requirements of the outcome.
- Teachers may develop their own marking schemes for this outcome, provided they are consistent with the VCAA Performance descriptors forUnit 4 Outcome 3. To be consistent with the performance descriptors, teacher-generated schemes must take into account the following points:
- explaining the historical context for the emergence of computer science as a field
- describing the general structure of a Turing machine
- demonstrating the existence of hard limits of computability using the Halting Problem
- describing and comparing the Turing Test, strong AI and weak AI as conceptions of artificial intelligence
- describing the Chinese Room Argument and mounting an argument for or against it
- explaining, at a high level, how data-driven algorithms can learn from data
- explaining the optimisation objectives for training SVMs and neural network binary classifiers
- explaining how higher dimensional data can be created to allow for linear classification
- describing the structure of a multi-layer perceptron neural network
- evaluating the output of a small multi-layer perceptron neural network using forward propagation
- explaining the consequences of model overfitting or underfitting
- explaining and discussing ethical issues related to artificial intelligence and data-driven algorithms.
Teacher judgment should be used to determine the weighting of each criterion within the School-assessed Coursework task. While weightings are not explicit within the VCAA Performance descriptors, teachers must understand that the criteria are not intended to be equally weighted.
- Teachers should first complete the assessment task themselves by selecting one of the following:
- a response to a case study or stimulus material
- a written report
- an annotated visual report
- an oral report
- structured questions.
This will assist them to understand how students may respond to the stimulus material. It will further ensure that marking schemes are appropriate for the task being provided. Refinements to the task may occur as a result of this process.
Schools may determine the conditions for assessment tasks. Assessment tasks should be a part of the regular teaching and learning program and should not add unduly to student workload. Students should be advised of the timeline and conditions under which the task is to be completed. It is recommended that assessment tasks be completed in class under supervision within a limited time frame.
The overall assessment program for the unit should include a variety of activities, include provision for authentication of student work and take into consideration the overall workload for students.
The teacher must decide the most appropriate time and conditions for conducting this assessment task and inform the students ahead of this date. This decision is a result of several considerations including:
- the estimated time it will take to teach the key knowledge and key skills for the outcome
- the likely length of time required for students to complete the task
- the classroom environment in which the assessment task will be completed
- whether the assessment task will be completed under open-book or closed-book conditions
- any additional resources required by students
- when tasks are being conducted in other subjects and the workload implications for students.
The teacher must consider the authentication strategies relevant for each assessment task. Information regarding VCAA authentication rules can be found in the
VCE and VCAL Administrative Handbook section:
Scored assessment: School-based Assessment.
The VCAA Performance descriptors can be used and adapted to the specifics of the task to assess a student’s level of performance. The assessment tools (performance descriptors, rubrics and / or marking guide) should reflect the outcome, key knowledge and key skills. The assessment task and assessment tools should be explained to students before they commence the task.
The
VCE Assessment principles underpin all VCE assessment practices.
Theperformance descriptors in this Support material for teachers give a clear indication of the characteristics and content that should be apparent in a student response at each level of achievement from very low to very high. The specified assessment task is listed on page 18 of thestudy design. The assessment task is to be out of 50 marks.
The performance descriptors are advisory and designed to support teacher judgments in making holistic assessments of students' demonstration of the key knowledge and key skills for each outcome. They provide a way for teachers to differentiate between levels of student achievement. Teachers may use their professional judgement to customise the performance descriptors in line with the
VCE and VCAL Administrative Handbook and the
VCE assessment principles.