Academic managers online nowChecking the checkers, and publishing what we find.Submit assignment →
Quality assurance

Checking the checkers, and publishing what we find.

The editorial process describes what happens to every piece. This page is about whether that process is actually working — measured by people who did not do the work, against a scorecard published in full, with the results printed whether they flatter us or not.

Four mechanisms
01Blind re-marking of delivered worka sample is scored again by an auditor who did not write, review or manage it.
02A weighted scorecard, publishedsix criteria with fixed weights and a pass mark, so "quality" is a number someone can dispute.
03Monthly calibrationthe same piece scored by several auditors, to catch the standard drifting rather than the writing.
04Root cause, not blameevery failure is traced to the checkpoint that should have caught it, and that checklist is amended.

The scorecard every audited piece is marked against.

Six criteria, weighted. An auditor scores each from 1 to 5 without knowing who wrote the piece or which manager owned it. Below 4.00 overall, or any single criterion at 2, triggers the escalation path.

Brief answeredWeight 30%

The set question, in the terms the rubric uses — not the topic in general.

scored 5 of 5
Depth of analysisWeight 25%

Weighing and evaluating where asked, rather than describing at length.

scored 4 of 5
Sources and accuracyWeight 15%

Sources exist, are current, and say what they are cited as saying.

scored 4 of 5
Citation consistencyWeight 15%

One style throughout, with pinpoints and a matching reference list.

scored 5 of 5
Structure and signpostingWeight 10%

The argument is findable by a marker reading at speed.

scored 4 of 5
Presentation and handbookWeight 5%

Formatting, word count, cover sheet and declarations as specified.

scored 5 of 5
Weighted score
4.50
out of 5.00 · pass mark 4.00
Pass

At or above the 4.00 pass mark with no criterion at 2. The piece stands as delivered and the score joins the specialist’s rolling average.

Consequences attached to this score
Nothing is required. The score is logged against the piece and the specialist.
The specialist’s rolling average moves; above 4.5 sustained, their fee per piece rises.
You can request the scorecard and the second-read note for your own order at any time.

The scorecard is interactive so you can see how the weighting behaves — move any criterion and watch what it costs. Weights and thresholds are design placeholders.

The numbers, as at the last review.

Published whether they moved the right way or not. Where a figure got worse, the reason sits underneath it rather than in a footnote.

Delivered pieces audited
11.4%
↑ from 9.8%

Target is one in ten; the increase came from more triggered audits, not more random ones.

Audits scoring 4.0 or above
94.1%
↑ from 92.6%

The gain is entirely in citation consistency after the checklist was rewritten in April.

Delivered on or before deadline
98.7%
↓ from 99.1%

Four late deliveries in June, all in one subject where a specialist withdrew mid-order. All refunded automatically.

Pieces returned at second read
19.2%
↑ from 17.4%

Deliberate. Second readers were told to return borderline work rather than annotate it — a rising number here is the system working.

Free revisions requested
8.3%
↓ from 9.7%

Tracks the outline-approval rule introduced for pieces above 1,500 words.

Refunds paid, all causes
1.9%
↔ unchanged

Two thirds were automatic late-delivery refunds rather than disputes.

Every figure above is invented sample data from the design bundle. None of it has been measured — see the note at the foot of this page.

How pieces are picked for audit.

A sample that only ever contains suspicious work tells you nothing about the rest.

Random draw

50% of audits

every delivered piece is eligible, drawn weekly

Triggered by risk

25%

first piece from a new specialist, or a subject with a thin panel

Triggered by signal

20%

a revision request, a low rating, or a return at second read

Requested by you

5%

any client can ask for their own piece to be audited

What the auditor can and cannot see.

Blind marking is only blind if the identifying material is actually withheld.

The brief, the rubric, the citation style and the delivered file
The second-read note, but only after their own score is submitted
Who wrote it, and who reviewed it
Which manager owned the order, or what it was charged
Your name, institution or any client detail
Calibration exercises per year
12 †
Auditors per exercise
4–6 †
Acceptable spread in scores
≤ 1.0 point
Checklist amendments in 2026 so far
7 †

When something fails, this is the sequence.

Your order is put right before anyone internally discusses cause — you are not waiting on our process.

Step 01
Same day

Your order is put right first

A rewrite or a refund is decided before any internal conversation about cause. You are not waiting on our process.

Owner

Academic manager

Step 02
48 hours

Traced to a checkpoint

The failure is attributed to the checkpoint that should have caught it — not to a person. Almost always the checklist was silent on it.

Owner

Managing editor

Step 03
Same week

The checklist is amended

Wording is changed, and every reviewer is told what changed and why. The old wording stays on file.

Owner

Managing editor

Step 04
Next monthly review

Checked for recurrence

If the same cause appears twice after amendment, it becomes a training item and the panel member is re-scored on their next three pieces.

Owner

Quality review

Keeping specialists on the panel.

Entry is the easy part. Staying is what the scorecard is for.

Entry

a degree in the field, a sample assessed blind, and a paid trial piece scored against the same scorecard.

Rolling average

the last ten scored pieces. Below 3.8 pauses new offers until three supervised pieces clear 4.0.

Removal

any fabricated source, or two audits below 3.0, ends the arrangement immediately.

Uplift

a track record above 4.5 raises the fee per piece — quality is the only thing that pays more here.

Re-verification

credentials re-checked every two years, because a panel is a list that quietly goes stale.

Ask for the audit trail on your own order. Most people never do.

The scorecard and the second-read note for your own piece are yours on request, at any time, without a reason.