Moderation of centre assessed work and how marks are applied
You’ll be required to upload evidence for learners selected in the sample. Once the full sample of work is uploaded, it will be moderated in stages as outlined in Table 1 above.
Stage 1: When your marks are moderated, a tolerance level is applied to allow for differences between assessment judgements. The tolerance is a specified number of marks, specific to each component to determine the differences between centre marking and moderator marking that can be taken as a legitimate variation in judgement.
If the differences between the moderator’s mark and your mark are within tolerance for all learners in the sub-sample, no further work will be moderated, and your marks will be applied to all learners in the cohort.
This means that no change is made to your marks, resulting in centre marks being used as the final marks for that particular component.
If the difference in marks is outside the specified tolerance, moderation will continue to stage 2.
Example – in tolerance
The table provides an illustration of centre marks within tolerance for a centre-assessed component where the example tolerance is -/+ 6 marks of the total mark. In this case, no further moderation is needed and the centre mark is applied as the final result for this component.
|
Learner no.
|
Centre
|
Moderator mark
|
Difference
|
|
004
|
68
|
69
|
-1
|
|
008
(lowest mark)
|
22
|
21
|
1
|
|
014
|
47
|
46
|
1
|
|
016
(highest mark)
|
81
|
80
|
1
|
|
017
|
58
|
58
|
0
|
|
020
|
31
|
28
|
3
|
Stage 2: If the centre’s mark and moderator’s mark are out of tolerance for any learners in the sub-sample, an adjustment will be needed to align your standard of marking with our agreed standard. The additional learners for the stage 2 sample will be moderated and the pattern of differences between the moderator’s mark and centre’s mark will be reviewed. If there is a consistent pattern of marking, a fair adjustment can be made from the stage 2 sample and no further work will be moderated. Marks will generally be adjusted using the regression process, which is outlined below.
If your marking pattern is inconsistent, the moderator will progress to stage 3 moderation.
Example – Out of tolerance (with consistent marking pattern)
This illustrates the outcome of the stage 1 moderation as outside of tolerance, therefore further moderation needed, and so contains learners in the stage 2 sample.
|
Learner no.
|
Centre
|
Moderator mark
|
Difference
|
|
004
|
68
|
75
|
-7
|
|
008
(lowest mark)
|
22
|
28
|
-6
|
|
014
|
47
|
53
|
-6
|
|
016
(highest mark)
|
81
|
89
|
-8
|
|
017
|
58
|
62
|
-4
|
|
020
|
31
|
37
|
-6
|
|
009
|
30
|
36
|
-6
|
|
015
|
31
|
39
|
-8
|
|
011
|
38
|
45
|
-7
|
|
005
|
47
|
54
|
-7
|
Although the marks moderated were outside of the +/- 6 tolerance limit, a consistent pattern of marking is evident (see graph below). Therefore, a fair adjustment can be applied using the regression process.
Stage 3: If the centre’s marks and the moderator’s marks are out of tolerance and inconsistent, the remaining learners in the sample will be moderated. After moderating the full sample, the pattern of differences between the moderator’s marks and the centre’s marks will be reviewed. a fair adjustment will be made, and no further work will be moderated.
In most cases the stage 3 sample of work will be sufficient to determine the size and scale of the adjustment that need to be applied.
Example - out of tolerance (with inconsistent marking pattern)
|
Learner no.
|
Centre mark
|
Moderator mark
|
Difference
|
|
001
|
28
|
|
|
|
002
|
57
|
65
|
-8
|
|
003
|
72
|
|
|
|
004
|
68
|
60
|
8
|
|
005
|
47
|
40
|
7
|
|
006
|
48
|
|
|
|
007
|
81
|
79
|
2
|
|
008
|
22
|
23
|
-1
|
|
009
|
30
|
30
|
0
|
|
010
|
61
|
|
|
|
011
|
38
|
48
|
-10
|
|
012
|
65
|
58
|
7
|
|
013
|
66
|
59
|
7
|
|
014
|
47
|
56
|
-9
|
|
015
|
31
|
38
|
-7
|
|
016
|
81
|
85
|
-4
|
|
017
|
58
|
55
|
3
|
|
018
|
50
|
58
|
-8
|
|
019
|
37
|
|
|
|
020
|
31
|
25
|
6
|
Stage 4 (additional samples): In circumstances where centre’s marks are found to be inconsistent throughout the sample, and a fair adjustment can’t be made to the whole cohort, then additional samples will be required (typically the work of all learners in the cohort). We’ll contact you to request additional work to be submitted.