Scribd
Upload
17 views2 pages

Estimating Marks Guidance For Centres 06.03.19!8!1 1 Copy 1

Uploaded by

eleventaken
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
17 views2 pages

Estimating Marks Guidance For Centres 06.03.19!8!1 1 Copy 1

Uploaded by

eleventaken
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

AQA City & Guilds CCEA OCR Pearson WJEC

Information for centres


ESTIMATING THE MISSING MARK WHEN A CANDIDATE IS ABSENT FROM AN
EXAMINATION
This document explains one approach to estimating the missing mark when a candidate is absent (for an
acceptable reason) from a unit or a component in either a modular or a linear specification. The same
principles apply when a mark is missing because the script has been lost. Hereafter, any reference to
‘component(s)’ also applies to unit(s) in a modular specification.

Several other methods are available to awarding bodies and may be employed by them if necessary. Different
procedures result in very similar estimates. The conditions under which an estimated mark can be awarded
in the case of candidate absence are set out in Chapter 4 of the JCQ publication A guide to the special
consideration process, which is available on the JCQ website.

Under the procedure known as the z-score method, the difference between the candidate’s estimate and
the performance of candidates generally on the component in question is the same as the average difference
between the candidate’s performance and the performance of candidates generally on the other
components.

If the candidate performed on average slightly better than candidates generally on the other components,
then the estimate for the missing mark will be slightly above the general performance on that component.

The difference between the performance of the candidate in question and the performance of candidates
generally is measured in terms of standard deviations. For example, a candidate whose performance on a
component is slightly above the performance of candidates generally on that component might be half a
standard deviation above the mean mark for that component. On another component, however, where
his/her performance is slightly below the performance of candidates generally, he/she might be half a
standard deviation below the mean mark. The number of standard deviations above or below the mean is
called the z-score.

In linear GCEs in England, and in both linear and modular GCSEs throughout the UK, if a z-score approach is
employed, it is standard practice for the calculation to use the available marks from all components in the
specification it. Occasionally there will be exceptions to this practice for technical reasons.

In modular GCEs, the standard for AS and A2 units is different. Therefore, only AS marks are normally used
in the z-score calculations for AS units and similarly only A2 marks are normally used in the z-score
calculations for A2 units. This restriction on the use of AS and A2 units is waived for two-unit A Level
qualifications. For these, a missing mark for the AS unit is estimated using the result for the A2 unit; a missing
mark for the A2 unit is estimated using the result for the AS unit.

A similar method may be used in other situations where a mark needs to be estimated.

It should be recognised that the z-score approach may not prove suitable for some small entry options. In
such cases awarding bodies reserve the right to employ alternative approaches if necessary.

1
In the examples below the means and standard deviations are exact whole numbers. Of course this would
be unlikely to occur in practice but it makes the principles easier to understand. The examples refer to
linear qualifications and raw marks, but the same calculations would apply, in terms of uniform marks, for
modular qualifications.

Example 1

In a two component linear specification the mean raw mark for all candidates on component 1 is 53 and
the standard deviation is 5, while the mean on component 2 is 34 and the standard deviation is 3.

A candidate scores 43 on component 1 but is absent for component 2.

Mean Standard deviation Candidate’s mark

Component 1 53 5 43
Component 2 34 3 Absent

The candidate’s mark on component 1 is 10 marks, or 2 standard deviations, below the mean. Therefore
the estimate for component 2 is also 2 standard deviations below the mean. This is

34 – (2x3) = 28

Example 2

In a three component linear specification the mean raw marks for all candidates on components 1, 2 and 3
are 50, 80 and 38 respectively, with standard deviations of 8, 12 and 3. Component 1 accounts for 30% of
the assessment, component 2 for 50% and component 3 for 20%.

A candidate scores 58 on component 1 and 104 on component 2 but is absent for component 3.

Weighting Mean Standard Candidate’s mark


deviation
Component 1 30% 50 8 58
Component 2 50% 80 12 104
Component 3 20% 38 3 absent

The candidate’s mark on component 1 is 8 marks or 1 standard deviation above the mean and the candidate’s
mark on component 2 is 24 marks or 2 standard deviations above the mean. The average (taking account of
the weightings) is

(30 x 1) + (50 x 2)
(30 + 50)

= 1.625

Thus, the estimate for component 3 is

mean mark + (1.625 x standard deviation)

= 38 + (1.625 x 3)

= 42.875, which is rounded to 43.

You might also like