• HOME >
  • Statistical Moderation of VCE School-based Assessment

Statistical Moderation of VCE Coursework

Assessment in the VCE

VCE studies have three graded assessments in Units 3 and 4. VCE VET studies have two graded assessments. In each study the graded assessments include a mix of school-based assessment and external examination, and the Victorian Curriculum and Assessment Authority (VCAA) combines each student’s scores for these assessments into a single Study Score.

The external examinations may be written, performance or oral examinations. There are two forms of school assessment – coursework assessment and School-assessed Tasks. Most studies, including VCE VET studies, have coursework as the form of school assessment. The studies that have School-assessed Tasks are Art, Design and Technology, Food and Technology, Media, Studio Arts, Systems Engineering and Visual Communication and Design.

Details of the graded assessments in each study are provided in VCE Study Designs, Assessment Guides and other VCAA publications.

Back to Top

What is moderation?

Moderation is a process of ensuring that the same assessment standards are applied to students from every school doing a particular study. Statistical moderation is a process for adjusting schools’ assessments to the same standard, while maintaining the students’ rank order given by the school. The VCAA uses statistical moderation to ensure that the coursework assessments given by different schools are comparable throughout the State.

Back to Top

Why is moderation needed?

School assessment is an important part of the VCE. In many studies it counts for 50% of students’ assessments. It is important that the assessments made by all schools throughout the State are comparable and fair to all students.

The VCE gives teachers some flexibility in deciding the teaching and learning activities and the coursework assessment tasks they will use to assess the learning outcomes specified in each Study Design. As a result, the coursework assessments from two different schools will sometimes be based on two different sets of assessment activities, although they will be assessing the same learning outcomes.

In some cases, the assessment tasks or topics set by one school may be easier than those set by another school. As well, the marking by the teachers in one school may be stricter or more generous than the marking in another school.

In a school where assessment tasks are easier and marking is generous, students would get higher marks for the same standard of work than they would in another school where the assessments and marking are harder. If we do not take account of this when using schools’ assessments to calculate the students’ Study Scores, some students would be treated unfairly.

Back to Top

How does statistical moderation work?

Statistical moderation is a process for adjusting the level and spread of each school’s assessments of its students in a particular study, to match the level and spread of the same students’ scores on a common external score. Because the external score is based on examinations done by all students across the State, it is a common standard against which schools’ assessments can be compared.

Each VCE study includes at least one external examination and the VCAA will use the examination scores in each study as the basis for statistical moderation of schools’ assessments. In studies with two examinations, scores from both examinations will be used.

The VCE assessment program also includes the General Achievement Test (GAT). Rather than using examination scores alone, statistical moderation will use students’ GAT and examination scores in studies where in doing so a better match with schools’ assessments throughout the state is achieved. In all such cases, the examination scores will always be the major influence.

The first step in moderating schools’ assessments in each study is to identify the moderation group for each study at each school.

For example, the moderation group for Legal Studies at a particular school is all the students doing Legal Studies at that school. If there are a number of Legal Studies classes at the school, the students in all the classes constitute the moderation group.

If a school has only a very small number of students doing a study, then it is advisable for the school to combine with another school for moderation purposes. When this happens, students from both schools doing that study constitute the moderation group.

Some students have results that need special treatment. The results for these students are initially left out of the moderation process, and the moderated scores for these students are calculated later, in line with the rest of the group. Results are removed for students who do not have complete coursework assessments, examination scores or GAT scores (in studies where the GAT scores are used in the moderation process), or who have an application for a Derived Examination Score.

The second step is to form an external score for each student doing the study, based on their examination scores for the study and, for a number of studies, using their GAT scores as well. These external scores are used as the common standard for all schools teaching that study.

The third step is to use the external scores of the moderation group to adjust the school coursework scores for the group. To do this, the moderation procedure ensures that:

  • the highest* moderated score is made equal to the highest external score
  • the median* and quartiles* of the moderated scores are made equal to the median and quartiles of the external scores.

The moderation procedure aims to make the mean (average) of the moderated scores as close as possible to the mean of the external scores. The procedure is then applied to the school’s coursework score for each student to obtain their moderated coursework score.

The moderation procedure is not influenced by students with anomalously low external performance, or by students who did very poorly on the school assessment but very well on the external assessment.

* The highest, median and quartiles are scores at the top, middle, 25% and 75% positions in large groups.

Back to Top

The outcomes of statistical moderation

These are the coursework scores given by each school. The scores from School A are higher than those from School B. For School A the median is about 74. For School B the median is about 58.

Outcomes of statistical moderation of Accounting assessments given by 2 schools.

These are the external scores for each school. The scores for School A are more spread out than those for School B. Compared with their external scores, School A's coursework scores were slightly high and School B's were low.

Outcomes of statistical moderation of Accounting assessments given by 2 schools.

These are the schools' moderated scores. School A's scores were adjusted downwards and School B's scores were adjusted upwards. In each school, the student's rank order, as determined by the school's coursework scores, is preserved.

Outcomes of statistical moderation of Accounting assessments given by 2 schools.

School A
In School A, student a has the highest coursework score (90) and the highest external score (89). After moderation, student a stays at the top of the group with a moderated score set equal to the school's highest external score (89). Students b and c have equal coursework scores (79), but student b has a higher external score (70) than student c (67). They receive identical moderated scores (73) and remain equal third in the school's rank order. Student d has the lowest coursework score (52) and the lowest external score (40). After moderation he/she stays at the lowest score.

School B
In School B, student x has the highest coursework score (71) and the second highest external score (78). After moderation he/she stays at the top with a moderated score equal to the school's highest external score (79).  Student y has the seventh highest coursework score (56) and the lowest external score (47). After moderation he/she stays in seventh position for the school, but his/her coursework score is adjusted upwards, along with the scores of all other students in the school. Student z has the lowest coursework score (37) and the second lowest external score (49). After moderation he/she stays on the lowest moderated score.

This example shows that the coursework scores of both schools have been adjusted to the same standard. Differences in the schools' marking have been removed so that the students in each school received final moderated scores that more fairly reflect their relative level of achievement.

Back to Top

Key points to remember

  • Statistical moderation is required so that school assessments can be used fairly as part of VCE assessment. It ensures that the assessment of all students, no matter what school they attend, is comparable and fair.
  • Statistical moderation adjusts schools’ assessments in accordance with students’ scores on common external examinations.
  • Any adjustment to a student’s score is determined by the external scores for the whole group, not by the student’s own external score.
  • Statistical moderation does not change the rank order of students, as determined by the school’s coursework scores. A student given the top score for coursework by his/her school will have the top score after statistical moderation, no matter how they perform on the examination(s).
  • Students who have applied for a Derived Examination Score or have an unusual performance on the examinations will not affect the moderation process.
  • Students who do not complete their examinations will still have their coursework scores statistically moderated, using information from the rest of the group.

Back to Top

Frequently Asked Questions from teachers

 

There are two very weak students in my class.  Will they drag down the marks of the others?

No.  Research has shown that this is not the case provided that school-assessed coursework marks are a valid reflection of their level of achievement.  It would be a mistake for you to give them ‘encouragement marks’ just to make them appear to catch up with others in the class.

 

One student in my class performed very badly in the VCE examination.  Will this drag down the marks of the others?

No.  Analysis of case studies has shown that this is not the case.  In a class of typical size there are some students who do a little better than expected and some who do a little worse; the effects on statistical moderation tend to cancel out.  Students who perform significantly worse in the exam are omitted when the scales are aligned in statistical moderation.

 

Should low achievers be discouraged from sitting their VCE exam so that they do not drag down the results of other students?

There is no evidence that low achievers have this effect.

 

A student in my class was sick and did not take the GAT.  Will this affect the results of the rest of the class?

No.  This student’s results will not be used to align the scales.

 

My school does not focus on the GAT and spends no time preparing students for it.  Given that the GAT is used in statistical moderation, will this disadvantage my students?

The GAT tests thinking skills, not academic content.  Knowledge equivalent to that of an average Year 10 student is assumed and if any further information is needed it is provided in the questions.  At least one person on the panel that reviews the GAT is an ESL specialist and words that might be difficult for a student whose linguistic background is not English are replaced or explained.  The GAT is not a general knowledge test.

 

Last year I taught a really weak class and had to set them easy coursework assignments to keep them interested.  At the end of the year, as a result of statistical moderation, their coursework marks were substantially decreased.  Was I wrong to set easy assignments?

Keeping students engaged is essential for learning.  However, you have to manage students’ expectations by explaining to them that moderation will align their marks with those of the rest of the State.  More importantly, if your students are only exposed to relatively easy coursework they will not be adequately prepared for the rigour of the VCE examination.

 

Is it true that statistical moderation is carried out separately for each class?

No.  Statistical moderation is applied to moderation groups, i.e. all students in a given study at a school plus any partner schools.  It is essential that all results for the moderation group are assessed on the same scale, i.e. they allow the students’ assessments to be ranked consistently across all classes.  This ‘internal moderation’ must be done by schools before marks are entered through VASS.

 

Is it true that teachers should always give the maximum possible score to the top student in each school-assessed coursework assignment in order to spread out the students?

No.  The spread of marks will happen naturally if you use suitable assignments and tasks and use the marking advice provided in the Assessment Handbooks.

 

Am I disadvantaging my top students if I cannot separate them at the top of the assessment scale?

This is quite possible.  If these students have been getting 100% regularly in school-assessed coursework they will tend to bunch at the top.  These results will make ‘good’ students appear equal to ‘excellent’ students, denying the latter the opportunity to show how good they really are.

 

I am normally perceived as a “hard” marker.  Will this disadvantage my students?

All things being equal, a hard marker will award lower marks than a generous marker.  However, your school’s marking scale will have no permanent effect on students because it is replaced by the scale of the external scores during statistical moderation.  It should be remembered that if there are two or more teachers of the same study at your school it is important that they are all equally hard markers.

Back to Top

Understanding Statistical Moderation in the VCE (Brochure)

Brochure (pdf - 1.76mb)