Inspera question randomisation

• MCQ answer options are always randomised by Inspera and this setting is not changeable.
• Real-time corrections during the exam are possible.

E.g., In the same calculation, one version of the text gives students a price of $1.12 to work with but another version gives students a price of $2.43 to work with.

• The Assessment Services advises caution on the use of this type of randomisation, even though the technology allows for it to be created, because of the complexity and potential for errors to creep in when setting and/or marking the exam.
• Real-time corrections during the exam are possible.

E.g., Section 1 = Question 1
Students would be randomly assigned one of the following four questions – each would be set up as a separate question in Inspera.

• Describe the attributes of Audrey Hepburn
• Describe the attributes of Humphrey Bogart
• Describe the attributes of Katherine Hepburn
• Describe the attributes of James Dean

• This should be used with caution, as one question could be harder/easier to answer than the other and could cause inequities between students.
• This can be used with any question type only if each question has the same maximum marks and should only be used where the questions are similar in nature and difficulty.
• To achieve this, use the Inspera Assessment functionality of Sections in question sets and “pull questions in a section”.
• Although Inspera functionality does enable ‘pulling’ multiple questions into a section, this is not permitted as it would create a randomising of the order of questions.  It is recommended only one question should be pulled from a few options into a section. This will enable students to easily locate any corrections.
• It is also important to note that enabling ‘pull’ in sections means markers will not be able to quickly navigate through candidates focusing on one question option at a time; a marker will be presented the question option presented to the student for that section.

The sequence in which question 99, say, appears would be different for each student.

• Current limitations with the functionality for updating and advising corrections do not enable this to be a viable option.
• Altering the order of questions can lead to omission of questions and anomalies in the answer sheets.
• Randomisation of question order makes it problematic to issue correction notices during an exam. It creates uncertainty and increased anxiety for students who raise queries during the exam about a potential error that we are unable to resolve. When an error isn’t resolved quickly during an exam, we receive increasing numbers of student queries, and then complaints following the exam, with claims that the unresolved error caused students to become distracted throughout the rest of their exam, impacting their performance. We do not have robust processes for identifying those exams with randomisation nor functionality to easily communicate issues or make and communicate corrections.  We will be exploring opportunities to improve functionality and processes to determine if this could be an option going forward.

Each student receives a random set of y questions from a test bank of z questions.

• Current limitations with the functionality for updating and advising corrections do not enable this to be a viable option.
• This functionality would certainly reduce the possibility/ability for students to cheat. However, the strong concerns around it are around equity/fairness to all students. It is simply not possible to say that all questions within a test bank would have a similar level of difficulty and so some students might randomly be being given a “harder” section of questions than others (even if, say, they are grouped into easier, medium, and harder questions and each student received three questions from each category).
• Since students would randomly be receiving completely different questions, real-time corrections would not be possible.