A-Level Economics · Method · DRAFT
Why you should test yourself before you revise, not after
I call it measure-then-solve. Before I teach a student anything, I make them attempt real exam questions cold — no notes, no warm-up, no revision — because the attempt itself is both the diagnosis and the most effective piece of studying they will do that week. The research is unusually clear on this: across the leading experiments, students who tested themselves recalled between about 1.5 and 2.4 times as much a week later as students who spent the same time, or more, re-reading the material — and, crucially, the re-readers were the ones who felt more confident.
That last clause is why this method needs explaining rather than just doing. Every instinct a hardworking student has — read the chapter again, make the notes prettier, watch the video once more — produces a genuine feeling of mastery and a poor exam grade. The evidence says the feeling and the grade come apart, and it says so with numbers.
What the research actually found
Three experiments do most of the work. I will give you the real figures rather than a slogan, because the slogans in circulation are overstated.
Finding 1 — testing beats re-reading, but only on the exam that matters
Roediger and Karpicke gave university students short science passages. One group read a passage then took a free-recall test on it, with no feedback; the other simply read it again. Tested five minutes later, the re-readers won: 81% against 75%. Tested a week later — the condition that resembles an actual exam — the result flipped: the tested group recalled 56%, the re-readers 42%. In a second experiment, students who studied once then took three tests recalled 61% after a week; students who read the passage four times and never tested recalled 40% — despite having read it, by the researchers' count, 14.2 times against 3.4.
Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255.
Note the trap in that five-minute result: re-read, then check yourself straight away, and you will feel brilliant. The method is optimised for the twenty minutes after you close the book, not for June.
Finding 2 — once you know it, re-reading it adds nothing
Karpicke and Roediger had students learn 40 Swahili–English word pairs. Once a pair had been recalled correctly, it was either kept in the testing rounds or dropped from them. A week later, students who kept testing recalled about 80%; students who dropped items from testing but carried on studying them recalled 36% and 33%. Repeated studying after the first successful recall produced, in the authors' words, no measurable learning a week later. The effect size was d = 4.03, and the two groups' score distributions did not overlap at all.
Karpicke, J. D., & Roediger, H. L., III (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968.
This is the study behind the bolder multiples you see quoted online, and it is also the narrowest: foreign-language vocabulary, one week. I use it as evidence for a principle — keep testing what you already know — not as a promised number.
Finding 3 — it works for understanding, not just facts
The obvious objection is that economics is not vocabulary. Karpicke and Blunt tested exactly that. Students learned science texts either by building concept maps while looking at the material, or by practising retrieval; learning time was matched exactly. A week later, on short-answer questions including inference questions requiring students to connect concepts, retrieval practice scored 0.67 against concept mapping's 0.45 — described in the paper as about a 50% improvement, d = 1.50. In a second experiment of 120 students, 101 (84%) did better after retrieval practice. Ninety of those same 120 (75%) had predicted the opposite.
Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. Science, 331(6018), 772–775.
That 84%-against-75% split is the finding I care about most as a tutor. It is not that students are lazy. It is that the strategy which works feels worse while you are doing it, and students are reasonable people acting on how it feels.
Why measuring is only half of it
A test on its own changes nothing if nobody acts on the result. That is the point Paul Black and Dylan Wiliam made in Inside the Black Box, still the most-cited work on classroom assessment in Britain. Reviewing more than twenty controlled studies, they found that strengthening formative assessment — assessment actually used to change what happens next — produces "typical effect sizes of between 0.4 and 0.7", which they note are "larger than most of those found for educational interventions". Their own illustration: a gain of effect size 0.4 "would improve performances of pupils in GCSE by between one and two grades".
Two of their conclusions are built into how I work. First: "for formative purposes a test at the end of a block or module of teaching is pointless in that it is too late to work with the results." Second: "if pupils are given only marks or grades, they do not benefit from the feedback on their work." A number at the top of an essay is information about the student, delivered to the student, in a form they cannot use. So I never lead with the mark. I lead with which of my three diagnostic categories the loss came from — a knowledge gap, an application failure, or exam technique — because those three demand completely different lessons, and treating all three with "revise it again" is why a stuck grade stays stuck.
They also found that better formative assessment helps lower attainers more than the rest. That matches what I see: the student who is quietly guessing gains most, because for the first time somebody has located the actual fault instead of prescribing more effort.
Where the measurement should point
Once you know what a student can and cannot do unaided, the question is what to teach. Here I lean on a framing idea rather than an experimental result, and the difference matters. Lev Vygotsky defined the zone of proximal development as "the distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance" (Mind in Society, 1978, p. 86). That is developmental psychology about children learning, not a study of exam revision, and I am not going to pretend otherwise. But it names the thing precisely. A cold attempt at a past paper gives me the first number: what this student does alone. The lesson is worth having only if it operates in the gap above that — not on material already secured, and not so far ahead that I would simply be performing economics at them. The measurement exists to locate the edge.
What this looks like in one question
A worked micro-example
I give a new Year 13 student a cold 12-marker: "Assess the likely impact of a rise in interest rates on the UK housing market." No notes. Her definitions come back immaculate — interest rates, the transmission mechanism, demand and supply of housing, all correct, all textbook. Then the analysis reads, in full: "Therefore higher interest rates will reduce demand for housing, so house prices will fall."
That is around 5 out of 12, and a re-read of the chapter would not move it, because the knowledge is already there. This is not a knowledge gap; it is an application failure. What is missing is the chain — the point–cause–consequence links that turn a claim into economics. Higher base rate, therefore higher mortgage rates, therefore higher monthly repayments, therefore lower affordability for a first-time buyer at a given income, therefore fewer buyers able to bid, therefore excess supply at the prevailing price, therefore prices fall. Six links where she wrote one.
That diagnosis takes twelve minutes and changes the entire course of the tuition. Without it, I would have spent three lessons teaching her monetary policy she already knew — and she would have thanked me for it, and her grade would not have moved.
How to use this yourself, this week
You do not need me for the first step, and I would rather you did it than didn't. Take a past paper question on a topic you believe you know. Close everything. Write the answer cold, to time. Then mark it and ask, of every mark you lost: did I not know it, know it but fail to apply it, or know and apply it but fail to present it in the form the paper wanted? Those three answers point at three different remedies, and only one is "learn it again".
Then — and this contradicts everything you have been told — put the questions you got right back into the pile and test yourself on them again next week. That is the 2008 finding. Dropping a topic because you have proved you know it is the most expensive habit in A-Level revision.
Diagnosis before treatment is not a line I invented; it is how every serious field operates. A doctor who prescribes before examining is not being efficient, they are guessing. The evidence above is simply why I am willing to spend a first lesson finding out what is actually wrong, rather than a first lesson looking impressive.
References
- Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255.
- Karpicke, J. D., & Roediger, H. L., III (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968.
- Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. Science, 331(6018), 772–775.
- Black, P., & Wiliam, D. (1998). Inside the Black Box: Raising Standards Through Classroom Assessment. London: King's College London School of Education. (The full research review: Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education, 5(1), 7–71.)
- Vygotsky, L. S. (1978). Mind in Society: The Development of Higher Psychological Processes. Cambridge, MA: Harvard University Press, p. 86.