IB Geography HL common mistakes
IB Geography HL · Understand the traps and how to avoid them.
In Paper 2 and Paper 3 answers, the candidate describes the case study (e.g. a named river basin or megacity) narratively without ever naming or applying the geographic concept the question targets (e.g. 'the water balance', 'the demographic transition model', 'cumulative causation'). The response reads as a travelogue rather than a geographic explanation.
Candidates memorise case-study content as stories and revise facts about the place rather than the concept. Under time pressure they write everything they know about the location, hoping marks will land, instead of selecting the concept and using the case study as evidence for it.
What to do instead
For every case study, build a two-column revision card: left column = the named concept/process/model; right column = 3–4 specific facts from the place that illustrate it, each with a number or date. In the exam, open the paragraph by naming the concept, then deploy the facts as evidence. Practise converting a 'describe the place' sentence into a 'the concept explains the place because…' sentence.
The candidate treats command words as interchangeable, writing a long descriptive account when the question says 'evaluate', 'to what extent', or 'discuss', and never reaching a judgement or conclusion.
Descriptive writing feels safer and is easier to produce under exam pressure, and many candidates do not have a mental checklist of what each command word demands. They also run out of time before reaching the evaluative section.
What to do instead
Drill a command-word table (describe / explain / compare / assess / evaluate / to what extent / discuss) with the required structural move for each. For evaluative questions, budget the last 25% of the time for a conclusion that states a clear position, weighs at least two factors, and acknowledges a counter-argument. Write the conclusion first in planning notes so it cannot be dropped.
In the geographic skills and fieldwork sections, the candidate reads values off a graph, choropleth, or isoline map without checking the legend, units, scale, or the class intervals used, and then quotes figures that are wrong or mis-scaled (e.g. reading a logarithmic axis as linear, or misclassifying a choropleth shade).
Candidates jump straight to the visual pattern and skip the metadata. Many have practised interpreting graphs in maths or science contexts where axes were linear and legends obvious, so they do not habitually interrogate map class intervals or axis transformations.
What to do instead
Adopt a fixed reading routine before quoting any number: title → legend/keys → units → axis type and scale → class intervals → direction of any anomaly. Practise with deliberately awkward stimuli (log axes, unequal class intervals, discontinuous isolines) and state the value with its unit and the interval it falls in.
When asked to describe or explain a pattern on a map or graph, the candidate lists isolated data points ('Country A is 45, Country B is 23') instead of identifying the spatial pattern, gradient, or anomaly, and never uses locational language (north–south, coastal, urban–rural, core–periphery).
Candidates believe quoting numbers equals using data, and they lack a vocabulary of spatial description. They also fear making a generalisation that could be marked wrong, so they stay at the level of individual values.
What to do instead
Practise the pattern–evidence–anomaly structure: state the overall pattern first with a locational descriptor, then cite two or three figures as evidence, then name one anomaly and suggest why it departs. Build a bank of spatial descriptors (gradient, clustering, concentration, discontinuity, core–periphery) and use them explicitly.
The candidate answers a physical geography question (e.g. on the water cycle, ocean circulation, or climate) purely in physical terms and never connects it to human activity, or answers a human geography question with no reference to physical processes or environmental consequences.
Candidates mentally file topics as 'physical' or 'human' and revise them in separate silos. The exam's synthesis questions reward integrated thinking, but the candidate's revision never modelled that integration.
What to do instead
For each topic, write one 'linkage sentence' that joins the physical and human: e.g. abstraction rates altering a drainage basin water balance; ocean acidification affecting fisheries and coastal economies; urban heat islands linking urban morphology to microclimate. Rehearse these so integration becomes automatic rather than an afterthought.
In the internal assessment, the candidate states a research question that is too broad or not geographic (e.g. 'Is pollution bad?'), then collects data without a clear hypothesis, sampling strategy, or justification, and the fieldwork method cannot actually answer the question posed.
Candidates choose a topic they find interesting but do not narrow it to a measurable geographic relationship. They also treat fieldwork as data collection for its own sake and skip the methodological reasoning that the mark scheme rewards.
What to do instead
Frame the research question around a measurable spatial relationship (e.g. 'How does distance from the city centre affect traffic volume along a transect?'). Justify sampling method, sample size, and site selection explicitly. Before collecting data, write one sentence explaining how each data set will test the hypothesis; discard any data that cannot.
The candidate presents fieldwork data in the internal assessment with no appropriate technique - raw tables, no mapping or statistical treatment - and offers no evaluation of data reliability, limitations, or sources of error.
Candidates equate 'presenting data' with 'putting it in a table' and lack confidence with simple statistical or cartographic techniques. They also see evaluation as criticism of their own work rather than as evidence of geographic judgement.
What to do instead
For each data set, select a representation matched to the data type (choropleth for density, isoline for continuous surfaces, scatter with trend line for correlation, kite diagram for transect change). Add at least one simple statistical measure (e.g. Spearman's rank, mean and range, index of dispersion) and interpret it. Write a limitations paragraph that names specific sources of error and states how each affected the results.
In global interactions (HL extension) questions, the candidate describes globalisation or international trade in general terms without using HL-specific concepts such as core–periphery, cumulative causation, the multiplier effect, or the role of transnational corporations and global production networks, and without reference to a specific named example.
HL extension content is often revised last, and candidates fall back on general knowledge from SL-level topics. They also confuse 'global interactions' with 'globalisation is increasing' and so never reach the analytical apparatus the HL paper requires.
What to do instead
Build a concept map for global interactions linking each HL concept to at least one named example (e.g. a TNC and its production network, a newly industrialised country, a trade bloc). In answers, explicitly use the concept vocabulary and trace a causal chain (investment → employment → multiplier → migration → core–periphery shift) rather than asserting that globalisation happens.
The candidate uses a model or theory (demographic transition, Rostow, Malthus/Boserup, the water balance, the urban land-use models) as if it were a description of reality, asserting that a country 'is at stage 3' without acknowledging the model's assumptions, limitations, or exceptions.
Models are taught as summaries and are easy to memorise, so candidates deploy them as labels. They have rarely been asked to critique a model, so they do not think to question its applicability to a specific place or time.
What to do instead
For every model, prepare a critique card: underlying assumptions, what it omits, a counter-example, and a sentence on when it is and is not useful. In the exam, use the model as a tool ('the DTM would predict X, but country Y departs because…'), explicitly flagging the limitation rather than treating the model as fact.
The candidate answers optional-theme questions (freshwater, oceans, extreme environments, food, urban environments) with generic content that could apply to any theme, ignoring the distinctive processes and terminology of the chosen option, and often misreads which option the question belongs to.
Candidates revise optional themes unevenly, often relying on one well-prepared option, and then try to stretch that knowledge to fit a question from a less-prepared theme. They also skim the question stem and miss the theme label.
What to do instead
Before answering, identify the optional theme and the specific process named. Prepare a one-page 'signature processes and terms' sheet per optional theme (e.g. freshwater: drainage basin stores and flows, hydrograph lag time, eutrophication; extreme environments: permafrost, glacial mass balance, adaptation). If a question falls outside your prepared option, choose a different question rather than forcing generic content.