Letter-code questions look mysterious but they almost always follow a single rule: each letter is shifted forward or backward by the same number of places in the alphabet. Once your child sees a code as a set of alphabet hops rather than a secret language, these questions become some of the most reliable marks on the whole paper. In GL Assessment Verbal Reasoning papers, letter codes appear as one of the named question types (Bond labels the code families around Types L to O), and, crucially, GL never tells you in advance which types are in a given paper. That is exactly why a dependable, repeatable method matters more than memorising tricks.
What a letter-code question actually asks
There are two closely related jobs. In an encoding question you are given the rule through an example pair, then asked to code a new word. In a decoding question you are given the code and asked to work backwards to the real word. Both rest on the same idea: a fixed alphabet shift. Sometimes the shift is the same for every letter (a "constant" code); more often each letter has its own shift that follows a small pattern, such as +1, +2, +3. The examiner is testing whether your child can find that pattern quickly and apply it accurately under time pressure.
The two-step method
Step 1: Write the alphabet along the top of the margin, A to Z, once at the start of the section. This single strip serves every code question on the page, so you only draw it once. Step 2: For each letter in the example, count how many places it has moved and write the shift above it. If CAT becomes DBU, then C to D is +1, A to B is +1, T to U is +1, so the whole word is a simple +1 code. Apply that same +1 shift to the target word and you are done.
Constant codes are the friendly ones. The pattern to watch for is a changing shift. Suppose DOG is coded as EQJ. Counting the hops: D to E is +1, O to Q is +2, G to J is +3. The shift itself is growing by one each letter (+1, +2, +3). To code a new three-letter word you would apply exactly that ladder. Always count every letter's hop separately before you assume the code is constant; that habit alone prevents most careless errors.
A worked wrap-around example
The one thing that trips children is wrapping round the end of the alphabet. Treat the alphabet as a loop: after Z you return to A. If a question shifts Y by +3, you count Y to Z (that is 1), Z to A (that is 2), A to B (that is 3), landing on B. So Y +3 = B. The same works backwards: A minus 2 wraps to Y. Drawing the strip helps, but teach your child to say "carry on past Z, back to A" out loud so the loop becomes automatic. Backward shifts near the start of the alphabet (A, B, C) wrap in the same way to the end (X, Y, Z).
Reference table: common code types
| Type | Example given | Rule | Apply to |
|---|---|---|---|
| Constant shift | CAT → DBU | Every letter +1 | DOG → EPH |
| Growing shift | DOG → EQJ | +1, +2, +3 | CAT → DCW |
| Backward shift | SUN → RTM | Every letter −1 | MAP → LZO |
| Wrap-around | ZOO → CRR | Every letter +3 (Z→C) | YAK → BDN |
| Word code (letter→letter map) | Given a coded list | Match by position | Decode the odd one |
Encourage your child to check their answer by decoding it back. If applying the reverse shift returns the original word, the code is right. That thirty-second check catches nearly every slip.
Speed tips that save real seconds
GL papers are tightly timed at roughly a minute per question, and many children lose marks not because they cannot do codes but because they recount the alphabet from A every single time. Fix that with a few anchors and a clear routine.
- Memorise letter positions: A=1, E=5, J=10, O=15, T=20, Z=26. Counting from the nearest anchor is far faster than starting at A.
- Write the alphabet strip once, lightly, and reuse it for the whole section.
- Mark the hop above each letter before touching the target word — never do both at once.
- If a code looks constant but one letter disagrees, you have probably miscounted a wrap; recount that letter only.
- For word-code questions, line the coded words up under each other so repeated letters reveal the mapping.
Consistent practice is what turns these steps into instinct. Our free worksheets include dedicated code drills, and a full timed mock paper shows your child what codes feel like at real GL pace. For pupils sitting a specific consortium, the centre-specific mocks mirror the format their school actually uses.
How codes feed into the Standardised Age Score
Every correct code answer lifts the raw score, which is then converted into a Standardised Age Score (SAS) that adjusts for your child's exact age in months. That age adjustment is why a younger Year 6 child is not disadvantaged against an older classmate. SAS typically runs from about 69 to 141, with 100 as the average, and many grammar schools look for around 111 or higher — though exact pass marks vary by area and year and are never published in advance. Because codes are so learnable, they are one of the most efficient places to bank marks and nudge that score upward.
Codes rarely travel alone. They sit beside letter sequences, letter and number analogies, and hidden words in the same paper, so it pays to see the whole VR landscape. Our overview of the exam guides and the revision hub map how each type connects, and you can browse the full spread of question families on the blog index.
Frequently asked questions
Is every letter code just a Caesar shift?
Most are a fixed or patterned alphabet shift, but watch for growing shifts (+1, +2, +3) and wrap-arounds past Z. Always count each letter's hop separately rather than assuming the code is constant.
Does GL tell you which question types will appear?
No. GL does not disclose in advance which of its Verbal Reasoning types are in a given paper, so your child should be comfortable with all of them, codes included.
How long should a code question take?
Aim for well under a minute. With the alphabet strip drawn once and letter positions memorised, a well-drilled child can complete a code in around 30 to 40 seconds.