Step 2 of 10
Most marks are won or lost before you write a sentence. A GP question is built around a few words doing real work and at least one assumption you can challenge. Learn to see them, and the essay half-writes itself.
Weak essays usually fail at diagnosis, not execution. The student reads a question about, say, the environment, and writes everything they know about the environment. The marker reads it as a drift to the general topic, which is the difference between a Band 2 response and a Band 4 one. The cure is to spend the first few minutes working on the question itself, not reaching for content.
Three reads do the work. First, find the word doing the real work. Second, find the assumption the question takes for granted. Third, decide what scale it is asking you to measure on. Once those three are on your rough paper, you have a question you can actually answer rather than a topic you can only talk about.
Every GP question hides its difficulty in a small word: 'too much', 'only', 'always', 'necessary', 'good', 'successful', 'real', 'true'. These are the operative terms, and engaging them is one of the rarest and highest-value moves there is. A question is not 'is there concern about the environment'. It is whether there is 'too much' concern, which makes it a question about proportion, not about whether the environment matters.
The mistake is to define the term once in the introduction and then forget it. Markers notice, and they ask candidates to address the key word 'throughout the essay'. So plant the operative word in every topic sentence. If the word is 'too much', every paragraph should be measuring proportion. If it is 'only', every paragraph should be testing whether the single cause really is the only one. The term is the thread; do not drop it after the first paragraph.
The weak version restates the topic and promises a list of pros and cons. The strong version isolates 'too much', redefines the question as one of proportion, and tells you what every paragraph will measure.
Most GP questions smuggle in an assumption you are free to accept or challenge, and noticing it is a fast route to conceptual engagement. 'Should governments protect citizens from unhealthy lifestyles' assumes that government has a role here at all, that 'unhealthy' is well defined, and that 'protection' is the right frame rather than education or personal autonomy. You do not have to reject the assumption. You have to see it and decide what to do with it.
Often the sharpest move is to test a trade-off the question takes for granted. When a stem says 'X at the cost of Y' or 'only X', the strongest answers ask first whether the trade-off is even real before agreeing or disagreeing. A question that assumes efficiency and democracy are opposed can be answered best by showing where they actually reinforce each other. That is the difference between answering the question on its own terms and examining the terms themselves.
The stem tells you what kind of thinking is wanted, and each type demands something specific. 'How far' and 'to what extent' want calibration: a position on a scale, not a binary. 'Should' is normative and needs a value framework, whose interests and what purpose. 'Discuss' takes no position and asks you to build one. 'Consider the view that' wants you to examine a claim before accepting or rejecting it. 'Is it true that' usually wants 'yes in these respects, no in those', with a clear principle of distinction. 'Assess' or 'evaluate' wants criteria named first, then a judgement against them.
Misreading the type is a quiet way to lose marks. A 'discuss' answered as a one-sided argument reads as imbalanced. A 'should' answered without any sense of whose good is at stake reads as description. Name the type before you plan, because it decides the shape of the whole essay.
For 'extent' questions especially, decide which scale you are locating the claim on: scope (how broadly it applies), degree (how strongly it holds), conditions (when it holds and when not), time (was it true once, is it now), or domain (in work, in family, in politics). The axis is not decoration. It usually becomes the structure: a domain axis gives you a paragraph per domain, a time axis gives you a before and after.
This is the hinge between reading the question and writing the answer, and it leads straight into taking a stance and into evaluation, the next two steps. Crack the question well and the thesis is nearly written, because you already know the word you are measuring, the assumption you are testing, and the scale you are measuring on.
Book a free trial lesson and see how we teach GP from the exam backwards, with marked practice against the band descriptors.