Table of Contents
Use practice quizzes to create errors, not to collect scores. The value sits in the questions you get wrong and in what you do with them afterwards. A quiz taken, marked and forgotten teaches almost nothing. A quiz whose misses get logged, re-tested and closed out will move your real exam result.
That distinction matters more in trade subjects than in most. Delivery terms, origin, classification and payment mechanisms all contain pairs of concepts that look alike until a question forces you to separate them. You find those confusions by being tested, not by reading.
What follows is a working method: how to set up the log before you start, how to space the sessions, how to write your own questions when the bank runs dry, and the point at which practice quizzes start to make you worse.
What practice quizzes are actually for
Two different jobs get confused here. A practice quiz can be a diagnostic, telling you what you do not know. It can also be a study method in its own right, because the act of retrieving an answer strengthens the memory more than reading the same fact again.
Most people only use the first job. They take a quiz, look at the percentage, feel encouraged or discouraged, and go back to reading. The second job is where the gain is.
Re-reading fails for a specific reason. Reading a page a second time makes it feel familiar, and your brain reads familiarity as knowledge. It is not. Recognition and recall are separate abilities, and an exam only tests the second one. You can recognise every sentence in a chapter on origin rules and still be unable to state, cold, what makes a good wholly obtained in one country rather than another.
Closed-book retrieval closes that gap. Answer first, check second. Always in that order.
Set up the error log before your first quiz
Open a spreadsheet before you answer a single question. If you build the log afterwards you will not go back and populate it, and the quiz will have been wasted.
The five columns
- Topic. Which domain the question sits in, using the same names the exam uses.
- What the question actually asked. Rewrite it in your own words, in one line.
- Why my answer was wrong. Not the correct answer. The reason mine failed.
- Why each other option was wrong. This is the column people skip and it is the one that pays.
- Re-test dates. Three boxes, ticked as you re-attempt the item.
Column four does the heavy lifting. Distractors in a well-built question are not random. They are the near-miss concepts, the ones a half-prepared candidate confuses with the right answer. Explaining why each one fails forces you to articulate the distinguishing test, and the distinguishing test is what the exam is really checking.
Tag your confidence on every answer
Before you submit each question, mark it: sure, fairly sure, or guess. This takes a second and it changes what the quiz tells you.
A correct answer tagged “guess” is a miss. Log it exactly as if you had got it wrong. On the real exam that guess will land the other way often enough to cost you the pass, and it is invisible in a raw score.
The opposite case is more useful still. A wrong answer tagged “sure” is the highest priority item in the whole log. It means you are carrying a belief that is confidently incorrect, which no amount of further reading will dislodge unless you go and check the underlying rule. Those items go to the top of the queue.
Space the sessions instead of stacking them
Same hours, different spacing, different result. A single long Sunday produces a good feeling and poor retention. Short sessions across more days retain better, and the retest of a missed item is worth most when it happens just as you are starting to forget the answer.
One workable pattern, for someone with a few weeks before an exam:
| Session | What you do | Conditions | What it tells you |
|---|---|---|---|
| Baseline, day one | A full-length mixed set before any revision | Timed, closed book | Which domains to prioritise, and your natural pace |
| Topic sets, days two to ten | One weak domain per session, small batches | Untimed, closed book | Whether the rule has landed at all |
| First re-test | Only the items you missed, two days after logging them | Closed book, no hints | Whether the fix stuck overnight |
| Second re-test | The same items again, about a week later | Mixed in with new questions | Whether you can retrieve it without the topic cue |
| Mock, final week | Full length, full mix, no notes | Exact exam length and format | Readiness, pacing, and stamina |
| Day before | Read your own error log only | Light, short | Nothing new. This is consolidation |
The second re-test is the one people drop, and it is the one that converts a temporary fix into something you still have three weeks later. Mixing those items in with unrelated questions matters too. When a whole session is about one topic, the topic itself becomes the hint.
Write your own questions from the source material
Every question bank runs out. When it does, stop hunting for another one and start writing questions yourself. Turning a passage of rule text into a question that has one defensible answer and three plausible wrong ones requires you to understand exactly where the boundaries of the rule sit.
Write three per study session. Put them in the log with a blank answer, then attempt them a week later when you have forgotten your own wording.
A worked example
Take the point at which risk for the goods passes from seller to buyer under a delivery term. A weak self-written question asks what the term stands for. A useful one gives a fact pattern and forces a judgement: goods are damaged after loading but before the vessel sails, and you must say who carries the loss under the agreed term.
Build the wrong options from the confusions you know exist. One option should reflect the belief that risk and cost always move together. Another should reflect the belief that risk moves when title moves. A third should reflect the belief that insurance cover determines who bears the loss. Each of those is a real misunderstanding, and writing them out forces you to name why each is wrong. Our guide to exactly when responsibility passes from seller to buyer is a good source to build this kind of question from.
The Incoterms® rules used in those questions are published by the International Chamber of Commerce. GTCS is an independent Globalior assessment and is not affiliated with, endorsed by or accredited by the International Chamber of Commerce.
The same technique works elsewhere. For rules of origin, build fact patterns where processing happens in two countries. For classification, write questions where a product plausibly fits two headings. Ambiguity is the point.
Simulate the real conditions, but not on every session
Full simulation is expensive in time and it is only worth it two or three times. Use it for the baseline and the final mock. Everything in between can be untimed, because the goal in those sessions is depth of review rather than speed.
When you do simulate, match four things: the total number of questions, the total time, the question types, and the resources allowed. If the real exam is closed book, close the book. Practising with your notes open trains a skill you will not be able to use.
Format matters here. The GTCS certification introduction sets out an exam of 60 questions in 60 minutes, using multiple choice and true or false, drawn at random from a larger bank. A simulation that matches that gives you the thing you actually need, which is a felt sense of one minute per question. True or false items in particular reward a slow read, because a single qualifier such as “always” or “only” changes the answer.
Read the topic breakdown properly
Any quiz that reports by topic is worth more than one that reports a single number. But a topic score built on four or five questions is a fragile measurement, and people over-react to it constantly.
Use the score as a ranking, not as a grade. The question to ask is which domains sit at the bottom relative to the others, not whether a given domain hit some absolute figure. Then confirm the ranking against your error log. If a domain scores low but your log shows only misreads in it, the problem is your reading speed, not that topic.
GTCS reports a score plus a topic-level breakdown across eight domains, issued immediately after the attempt. Learn those domain names before you start practising, so your own log uses the same labels.
When practice quizzes start to mislead you
There are three failure modes, and all three feel like progress while they are happening.
- Bank memorisation. After enough passes you start recognising questions rather than answering them. Your score climbs and your knowledge does not. The tell is that your score on a fresh set is far lower than on the familiar one.
- Distractor familiarity. You learn which option “looks like” the answer in a particular bank’s house style. That skill does not transfer to a different bank or a different exam.
- Score chasing. Repeatedly taking easy sets because the number feels good. If you are getting nearly everything right, the session is not teaching you anything. Go find harder material.
The defence against all three is simple. Reserve at least one untouched set as a clean final check, and never review it beforehand. If your score on fresh material is close to your score on practised material, the learning is real.
Mapping this onto a trade contracts exam
Trade exams cluster their difficulty in predictable places. Questions about definitions are easy. Questions about boundaries between two adjacent concepts are hard, and that is where practice testing earns its keep.
Build deliberate question pairs on the boundaries you know are slippery. Risk transfer against title transfer. Cost allocation against risk allocation. The customs value of goods against the invoice price. The role of a document as evidence against its role as a contract. Payment obligations under a documentary credit against the underlying sale contract, where the way a letter of credit operates catches people out because banks deal in documents and not in goods.
Classification deserves its own treatment, because it is procedural rather than conceptual. Practising how to work through an HS classification as a sequence of steps is more useful than memorising code numbers, and a self-written question that asks for the next correct step tests that sequence directly.
If you are preparing for a contracts assessment specifically, check the published domain list first and structure your log around it. What the GTCS exam covers is set out in eight domains, there are no prerequisites, and one retake is included. Knowing the shape of the thing before you start practising saves a good deal of misdirected effort.
Key takeaways
- Practice quizzes work as a study method, not only as a measurement, because retrieving an answer strengthens memory more than re-reading the same material.
- Build the error log before your first quiz, with a column explaining why each wrong option was wrong.
- Tag confidence on every answer and treat a lucky correct guess as a miss.
- Re-test missed items twice, first after a couple of days and again about a week later, mixed in with unrelated questions.
- Write your own questions once the bank runs out, building the wrong options from misunderstandings you know people hold.
- Full simulation is worth doing two or three times, matching question count, time limit, question types and permitted resources.
- Keep one untouched question set as a clean final check, because a rising score on a familiar bank can be recognition rather than knowledge.
Frequently asked questions
How many practice questions should I do before a trade exam?
Count closed-out log entries rather than questions attempted. A candidate who works through a small set and properly resolves every miss is better prepared than one who runs through a large bank without review. Volume without review is the most common waste of study time. Stop when your error log has no open items and your score on fresh material holds up.
Should I take a practice quiz before I have studied anything?
Yes. A cold baseline tells you which domains to prioritise and gives you an honest sense of your pace. Expect a poor score and do not treat it as a verdict. Its only job is to direct the next two weeks of work. Take it timed, so the pacing data is real.
Is it better to review questions I got right as well as wrong?
Review the correct ones you were not sure about, and skip the ones you were confident on. That is what the confidence tag is for. Reviewing every correct answer burns time you could spend on the misses. Reviewing none of them hides the lucky guesses that will not repeat on exam day.
How close to the exam should I stop taking practice quizzes?
Do your last full-length run several days out, not the night before. The day before is for reading your own error log and nothing else. A bad score in the final 24 hours damages your confidence without leaving time to act on it, and there is no evidence that cramming new material that late helps.
What if the same question types keep catching me out?
That is a pattern, and it is good news because patterns are fixable. Group your log entries by question shape rather than topic. If every miss involves a negative in the stem, the fix is a reading protocol. If every miss involves two adjacent concepts, the fix is learning the distinguishing test between them.
Can practice quizzes replace reading the underlying rules?
No. Questions tell you where the gaps are and force retrieval, but they rarely give you the structure of a rule. Use quizzes to find the gap, then go to the source to close it, then return to questions to confirm it stayed closed. Doing only one half of that loop is why people plateau.
If you want a view on whether a contracts assessment fits what you are trying to prove, contact the Globalior team and ask before you book.