Test yourself on Graph, Chart and Data Skills with exam style questions, marked instantly.
Flashcards that come back just before you would forget them.
A revision plan that remembers what you know and shows what to do next.
See which topics you have nailed and which need another look.
Quiz yourself on Graph, Chart and Data SkillsGet your flashcardsBuild my revision planTrack my progress📊 You've Got Data. Now What?
Every geography paper — AQA, OCR B, Edexcel — includes graphs, charts, tables and maps. These are not decoration. They are the evidence you are supposed to analyse. The ability to extract meaning from data, communicate it precisely, and connect it to geographical explanation is one of the highest-value skills in the whole GCSE. A student who reads a climate graph and writes "rainfall is high in summer" scores Level 1. A student who reads the same graph and writes "total annual precipitation is approximately 1,600 mm, concentrated in the June–September wet season, likely driven by the northward migration of the ITCZ, with a pronounced dry season in December–February where monthly rainfall falls below 20 mm" scores Level 3. The difference is not knowledge. It is precision.
This topic gives you the complete toolkit: how to choose the right graph, how to describe any graph using the TACT framework, how to read climate graphs and population pyramids in detail, how to interpret scatter graphs and correlations, how to use statistical measures appropriately, and — crucially — how to avoid the mistakes that drop marks at every level.
📈 Graph Types: What They Are and When You Use Them
Choosing the right graph type is itself a geographical skill. Different data types have different structures — continuous, discrete, proportional, spatial — and each requires a different visual representation. Getting the graph type wrong makes it impossible to see the pattern in the data. Examiners also ask you to evaluate graphs: "How useful is this graph for showing...?" or "What type of graph would be more appropriate?" You need to know not just what each type does, but what it cannot do.
Line Graph
A line graph shows continuous data changing over time — data where values in between measured points are meaningful. The x-axis is always time (or a continuous variable) and the y-axis shows the measured value. The connecting line implies that the value changes smoothly between measured points.
Use for: temperature change over a year; river discharge over days; global CO₂ concentration 1960–2024; UK population growth 1800–2020.
Do NOT use for: comparing discrete categories (e.g., GDP of five different countries — these are separate, unconnected values, not a continuous sequence).
Strength: Shows trends and rates of change clearly. Easy to identify where values rise, fall or plateau.
Weakness: Can imply a smooth change between data points when the real data may be more erratic. Requires continuous, ordered data.
Bar Chart
A bar chart shows discrete (separate) categories. Each bar represents a distinct category with no implied connection to adjacent bars. Bars should not touch (unlike a histogram — see below).
Use for: comparing GDP between different countries; rainfall totals by month; number of earthquakes by magnitude band; population of different cities.
Strength: Easy to compare values between categories at a glance. Clear visual hierarchy — tallest bar is immediately obvious.
Weakness: Does not show change over time as clearly as a line graph. Cannot show the relationship between two variables.
Histogram
A histogram looks like a bar chart but works very differently. It shows the frequency distribution of continuous data that has been grouped into classes. The bars touch each other (no gaps) because the data is continuous — each class leads directly into the next. The area of each bar represents frequency, not just height.
Use for: showing how data is distributed across ranges — e.g., number of days with rainfall between 0–10 mm, 10–20 mm, 20–30 mm etc.; age distribution in grouped form.
Critical distinction: In a bar chart, the x-axis shows category names. In a histogram, the x-axis shows a continuous numerical scale with class boundaries. If the x-axis has labels like "Under 5", "5–9", "10–14" etc., you are looking at a histogram structure even if the bars appear similar.
Common mistake: Calling a histogram a "bar chart" — examiners will deduct marks for this in graph-choice questions.
Pie Chart
A pie chart shows how a whole is divided into parts. The full circle represents 100% of a total, and each segment shows the proportional share of each category.
Use for: land use breakdown (residential 35%, commercial 20%, green space 15%...); energy mix by source; employment by sector.
Strength: Immediately shows which category dominates. Clear visual proportions for 2–5 categories.
Weakness: Very difficult to compare two pie charts side by side (segments that look similar in size are hard to differentiate). Useless for showing change over time. Breaks down with more than 5–6 categories (too many thin slices to read). Cannot show absolute values — only proportions.
Scatter Graph
A scatter graph shows the relationship (correlation) between two variables. Each data point represents one observation with both an x-value and a y-value plotted. A line of best fit (trend line) can be added to show the overall pattern.
Use for: showing the relationship between GDP per capita and life expectancy; birth rate and HDI; altitude and temperature; distance from city centre and house prices.
Strength: The only graph type that directly visualises correlation. Can identify anomalies (outliers that do not fit the trend) clearly.
Critical rule: Correlation does NOT prove causation. Just because two variables move together does not mean one causes the other. There may be a third variable driving both.
Climate Graph (Climograph)
A climate graph is a combined graph showing average monthly temperature (as a line graph) and average monthly precipitation (as a bar chart) on the same diagram. It always has two y-axes: temperature (°C) on the right, precipitation (mm) on the left. Time runs along the x-axis, showing all 12 months.
Use for: showing the climate of a specific location — standard format used in geography globally. You will see these in every geography exam.
What to read from it: hottest and coldest months; annual temperature range; wettest and driest months; total annual precipitation; climate type based on the pattern.
Population Pyramid
A population pyramid is a double-sided horizontal bar chart showing the age-sex structure of a population. Age groups (cohorts, typically in 5-year bands) run vertically up the y-axis. Males are shown on the left, females on the right. Each bar shows the number or percentage of the population in that age group and sex category.
Use for: showing the demographic structure of a country at a specific point in time; comparing LIC and HIC population structures; showing the effects of population policy, migration, or demographic transition.
Strength: Shows age structure, sex ratio, life expectancy, birth rate and death rate patterns all in one diagram.
Weakness: Shows only one point in time — cannot show change over time directly.
Choropleth Map
A choropleth map uses shading or colour to show how a variable varies across geographic areas. Darker shading typically represents higher values (though always check the key).
Use for: HDI by country; unemployment rate by region; average rainfall by county; population density by area.
Strength: Immediately shows geographic patterns and regional variation.
Weakness: Shows only one value per area — hides variation within areas. A region with high average income may still contain areas of severe poverty. Class boundaries can create misleading "sharp" changes at borders that don't really exist in the data.
Flow Line Map
A flow line map uses arrows of varying width to show the movement of people, goods, or other quantities between locations. The width of the arrow is proportional to the volume of movement.
Use for: showing migration patterns; trade flows; traffic volumes; global shipping routes.
Strength: Shows both direction and magnitude of movement simultaneously.
Weakness: Complex to draw accurately; arrows can overlap and become confusing; difficult to read precise values.
📋 Graph Types at a Glance
| Graph Type | Best Used For | Key Limitation |
|---|---|---|
| Line graph | Change over time; continuous data | Not for discrete categories; implies smooth change between points |
| Bar chart | Comparing discrete categories | Does not show relationship between two variables |
| Histogram | Frequency distribution of grouped continuous data | Often confused with bar chart — x-axis must be continuous scale |
| Pie chart | Parts of a whole (proportions/percentages) | Hard to compare multiple sets; only shows proportions, not absolute values |
| Scatter graph | Correlation between two variables | Does not prove causation; line of best fit is a trend, not a prediction formula |
| Climate graph | Monthly temperature + precipitation together | Shows averages — hides year-to-year variation and extreme events |
| Population pyramid | Age-sex structure of a population | One point in time only; doesn't show change directly |
| Choropleth map | Spatial distribution of a variable across areas | Hides variation within areas; class boundaries create artificial "sharp" edges |
| Flow line map | Direction and volume of movement between places | Difficult to draw; arrows can overlap and obscure data |
🔍 The TACT Framework: How to Describe Any Graph
The most valuable skill in this whole unit is being able to describe a graph or chart systematically and precisely. The TACT framework gives you a reliable structure that works for any type of data presentation — line graph, bar chart, scatter graph, table, or map.
State the overall direction or pattern. Is the data increasing, decreasing, fluctuating (going up and down irregularly), or stable? Give the overall shape of the data before diving into detail. Quote the starting and ending values, or the minimum and maximum, to illustrate the trend. Use precise comparative language: "increased steadily", "declined sharply", "remained approximately constant", "showed a gradual upward trend".
Identify any data point or period that does not fit the general trend — a sudden dip, a spike, an outlier, or a country that doesn't fit the pattern. State its value and, if you can, suggest a geographical reason. Anomalies often carry high exam marks because they require interpretation, not just description.
Compare the highest and lowest values, or compare two groups, two time periods, or two places. Do NOT describe each one separately — make a direct comparison with a single statement. Use comparative language: "twice as high", "three times greater", "a difference of 24°C between the hottest and coldest months". If comparing two countries or regions, name both explicitly and state both values.
Quote specific data values with their units. "High temperatures" scores almost nothing. "A maximum temperature of 32°C in July" scores marks. Always include units (°C, mm, per 1,000, km²). Use precise geographical terminology rather than vague language: "annual temperature range" not "how much the temperature changes"; "precipitation" not "rain"; "cohort" not "age group".
TACT in Practice: Global Temperature Anomaly Graph (1880–2020)
Imagine a line graph showing how global average temperature has changed relative to the 1951–1980 average. Here is how TACT produces a Level 3 description:
Compare this to a typical Level 1 response: "The temperature has been getting higher over time. It was lower in 1880 and higher in 2020." Both students looked at the same graph. The first student applied TACT and deployed specific figures with units. The difference in marks is enormous.
The Figure Rule
You do not need to quote every figure on a graph. In fact, quoting too many figures without analysis is itself a Level 1 mistake (see Misconceptions below). Aim for three to five specific, well-chosen figures that illustrate your points. Think of figures as evidence in a court case — you select the most telling ones, you don't read out the entire case file.
Quick Check: A bar chart shows annual rainfall for five cities: London 601 mm, Paris 641 mm, Madrid 436 mm, Rome 523 mm, Berlin 571 mm. Write a two-sentence TACT description of this data.
TACT answer: "Annual rainfall across the five European cities ranges from 436 mm in Madrid to 641 mm in Paris, a difference of 205 mm (Comparison + Total with units). Madrid is notably lower than the four northern European cities, which cluster between 523 mm and 641 mm, suggesting a Mediterranean climate pattern with reduced precipitation — Madrid is an anomaly that reflects its semi-arid inland location and summer drought (Anomaly). The overall pattern shows higher rainfall in north-west Europe, consistent with prevailing westerly winds bringing Atlantic moisture (Trend)." Award yourself full marks if you: quoted at least two specific values with units, used the word 'range' or made a direct comparison, and identified Madrid as different from the others with a reason.
🌡️ Reading Climate Graphs in Detail
Climate graphs (also called climographs) appear in almost every geography exam — you need to be able to read them quickly and accurately. The structure is always the same: temperature shown as a line graph, precipitation shown as a bar chart, both on the same diagram with months along the x-axis. There are always two y-axes — check which scale is for temperature and which is for precipitation before you start reading values.
Step-by-Step Reading Method
Read the peak and trough of the temperature line. Note the month and the value in °C. Example: "The hottest month is July at 24°C; the coldest is January at 3°C."
Subtract the coldest month temperature from the hottest: Annual range = highest – lowest. Example: 24°C – 3°C = 21°C annual temperature range. A large range (over 20°C) indicates a continental interior climate. A small range (under 10°C) indicates an oceanic or tropical climate.
Read the tallest and shortest bars. Note month and value in mm. Example: "The wettest month is August at 85 mm; the driest is February at 25 mm."
Add up the 12 monthly bars, or estimate by averaging and multiplying. This does not need to be exact — an estimate with a clear method scores marks. Example: "Total annual precipitation is approximately 560 mm."
Use the overall shape to classify the climate. See the climate type guide below.
Climate Types — What to Look For
| Climate Type | Temperature Pattern | Precipitation Pattern | Example Location |
|---|---|---|---|
| Equatorial | High all year (25–30°C), small annual range (<5°C) | High all year (over 2,000 mm total), no dry season; often two rainfall peaks | Manaus, Brazil; Singapore |
| Tropical monsoon | High all year (20–30°C) | Extreme seasonal contrast — intense wet season (June–Sept) and dry season (Oct–May) | Mumbai, India; Dhaka, Bangladesh |
| Mediterranean | Hot dry summer (25–30°C), mild wet winter (8–12°C) | Low in summer (<20 mm/month), higher in winter; total 300–600 mm | Barcelona, Spain; Athens, Greece |
| Maritime temperate | Mild all year; small range (10–15°C); no extreme cold or heat | Moderate, fairly evenly spread throughout year (600–900 mm total) | London, UK; Dublin, Ireland |
| Semi-arid / desert | Very high summer, cool winter; very large annual range (30°C+) | Extremely low all year (<250 mm total); near-zero monthly values | Cairo, Egypt; Phoenix, Arizona |
| Tundra / polar | Very cold all year; most months below 0°C; short warm summer (5–10°C) | Very low all year (100–300 mm), often as snow; no month exceeds 50 mm | Fairbanks, Alaska; Tromsø, Norway |
Worked Example: London Climate Graph
Using the five-step method:
- Hottest month: July at approximately 18°C. Coldest month: January at approximately 4°C.
- Annual temperature range: 18 – 4 = 14°C — a moderate range indicating an oceanic/maritime climate.
- Wettest month: October/November at approximately 65 mm. Driest: April at approximately 37 mm.
- Total annual precipitation: approximately 600 mm, fairly evenly distributed with no pronounced dry season.
- Climate type: Maritime temperate. Evidence: small annual temperature range (14°C), mild winters (above 0°C), moderate precipitation year-round with no strong seasonal concentration.
Examination answer using TACT: "London's climate graph shows a trend of moderately warm summers and mild winters, with a temperature line rising from approximately 4°C in January to 18°C in July — an annual range of 14°C. This small range is an anomaly compared to continental European cities at similar latitudes (Berlin has a range of over 20°C), reflecting London's proximity to the Atlantic Ocean and the moderating influence of the North Atlantic Drift. Comparing the wettest and driest months, October (65 mm) receives approximately 75% more precipitation than April (37 mm), though this variation is relatively modest. The total annual precipitation of approximately 600 mm, evenly spread, classifies London as a maritime temperate climate."
Worked Example: Cairo Climate Graph
Cairo's graph looks dramatically different:
- Hottest month: July at approximately 35°C. Coldest: January at approximately 13°C.
- Annual temperature range: 35 – 13 = 22°C — a large range reflecting an inland, desert location.
- Wettest month: December at approximately 8 mm. Most months show near-zero rainfall.
- Total annual precipitation: approximately 25 mm — extremely arid. Most months record 0–2 mm.
- Climate type: Hot desert / semi-arid. Evidence: near-zero precipitation throughout the year, extreme summer temperatures, large annual temperature range.
Quick Check: A climate graph shows: hottest month July (29°C), coldest January (10°C), wettest month January (90 mm), driest month July (5 mm), total annual precipitation approximately 620 mm. What climate type is this, and give two pieces of evidence from the graph to support your answer.
Mediterranean climate. Evidence 1: Hot, dry summer — July is the hottest month (29°C) but also the driest (5 mm), which is the defining characteristic of a Mediterranean climate where high pressure dominates in summer, suppressing rainfall. Evidence 2: Wet, mild winter — January is cool (10°C) but the wettest month (90 mm), reflecting the influence of low-pressure weather systems in winter when the jet stream shifts southward. The annual temperature range of 19°C (29 – 10°C) and total precipitation of 620 mm are consistent with Mediterranean locations such as Rome, Athens, or parts of California and South Africa. A maritime temperate climate has its precipitation spread more evenly throughout the year — the clear wet winter / dry summer contrast confirms Mediterranean. Award marks for: naming the climate type (1 mark); quoting temperature values with units as evidence (1 mark); quoting precipitation values with units as evidence (1 mark); linking the pattern to a geographical explanation such as pressure systems or distance from sea (1 mark for explanation).
👥 Reading Population Pyramids and Scatter Graphs
Population Pyramids in Detail
A population pyramid is one of the most information-rich graph types in geography. In a single diagram it shows birth rate, death rate, life expectancy, age structure, sex ratio, and — by comparison over time — population policy effects, migration, and demographic transitions. Mastering how to read one is worth significant marks.
The Basic Structure
Each horizontal bar represents a five-year age cohort (0–4, 5–9, 10–14... up to 80+). The length of the bar shows either the number of people (raw figures) or the percentage of the total population in that cohort. Males on the left, females on the right. The y-axis (vertical) shows age groups with the youngest at the bottom and oldest at the top.
LIC Pattern (Low-Income Country) — Triangular Shape
HIC Pattern (High-Income Country) — Columnar / Barrel Shape
Reading Special Features
| Feature | What It Looks Like | What It Means |
|---|---|---|
| Baby boom bulge | A cohort noticeably wider than those above and below it | A period of unusually high birth rate (e.g., UK post-WW2 baby boom 1946–1964 = large cohort now aged 60–78) |
| Pinched middle (inward dent) | Working-age cohorts narrower than the cohorts above them | Emigration (young adults leaving); or a past period of low birth rates (e.g., China after one-child policy) |
| Outward bulge in working age | Working-age cohorts larger than expected given the base | Immigration — young adults arriving for work (UAE, Qatar have extreme versions of this) |
| Asymmetry (males ≠ females) | One side noticeably wider than the other at a specific age | War casualties (more males missing from affected cohort); female life expectancy advantage at older ages |
| Very broad top with more females | Bars over 75 wider on the female side | Female life expectancy exceeds male by 4–6 years (typical in HICs) |
Worked Example: Writing About a Population Pyramid
Question: "Describe the population pyramid for Country A. It has a very wide base, rapidly narrowing bars from age 15 upward, and very few people over 60."
Level 1: "Country A has a lot of young people and not many old people."
Level 2: "Country A has a wide base, indicating a high birth rate, and the pyramid narrows rapidly above age 15. There are very few people over 60, suggesting a low life expectancy."
Level 3: "Country A displays a classic LIC population pyramid with a wide base — the 0–4 cohort likely represents around 18–20% of the total population, indicating a high birth rate of approximately 40 per 1,000. The rapid tapering above age 15 reflects high childhood and early adult death rates, consistent with limited access to healthcare and high rates of preventable disease. Less than 3% of the population appears to be over 65, suggesting life expectancy of approximately 55–60 years. This triangular structure is characteristic of Stage 2 or 3 of the Demographic Transition Model. The large proportion of young dependants — those under 15 — creates a high youth dependency ratio, placing substantial pressure on schools, healthcare and food supply, and limiting the proportion of the population that is economically productive."
Interpreting Scatter Graphs and Correlation
A scatter graph plots individual data points (each representing one country, city, or observation) to show whether two variables are related. Understanding what the pattern of dots tells you — and what it does not — is crucial.
| Pattern | What It Shows | Example in Geography |
|---|---|---|
| Positive correlation | As X increases, Y increases. Points cluster around a line rising from bottom-left to top-right. | As GDP per capita increases, life expectancy increases |
| Negative correlation | As X increases, Y decreases. Points cluster around a line falling from top-left to bottom-right. | As GDP per capita increases, birth rate decreases |
| No correlation | Points are scattered randomly with no pattern. A line of best fit would be nearly horizontal. | Shoe size and exam score — no relationship |
| Strong correlation | Points cluster tightly around the line of best fit. | Strong positive correlation between HDI and female literacy rate |
| Weak correlation | Points are loosely scattered around the line of best fit. | Weak positive correlation between rainfall and crop yield (many other factors involved) |
The Causation Rule — CRITICAL
This is the most important rule in data interpretation and one of the most frequently tested misconceptions:
Correlation does NOT equal causation. Two variables can move together without either causing the other. There may be a third variable — an underlying factor — driving both.
Classic example: Countries with higher chocolate consumption have more Nobel Prize winners per capita. This is a genuine positive correlation. But chocolate does not cause scientific achievement — the underlying variable is national wealth. Rich countries both buy more chocolate AND invest more in education and research.
Geography example: There is a strong positive correlation between GDP per capita and life expectancy. But GDP does not directly cause people to live longer. The mechanism is indirect: higher GDP → more government revenue → better-funded healthcare → lower infant mortality, effective treatment of diseases → longer life expectancy. There are also confounding variables: Cuba has life expectancy of approximately 79 years despite a GDP per capita of only around $9,000 — this is an anomaly on the scatter graph that suggests other factors (healthcare prioritisation, diet, social equality) can compensate for lower income.
Identifying and Explaining Anomalies on Scatter Graphs
An anomaly on a scatter graph is a point that lies noticeably far from the line of best fit. Identifying anomalies and explaining them is a high-value skill because it requires genuine geographical understanding.
Example: On a scatter graph of GDP per capita (x-axis) against birth rate (y-axis), most points cluster around a downward-sloping line (as GDP increases, birth rate falls). Saudi Arabia and the UAE may appear as anomalies — they have high GDP per capita but moderately high birth rates. This might reflect: the young age structure of the migrant worker population; cultural and religious factors supporting larger families; government policies and high marriage rates; or the way national GDP data can be skewed by oil wealth that doesn't translate equally into education and healthcare access for all citizens.
Quick Check: A scatter graph shows GDP per capita on the x-axis and birth rate on the y-axis. Country X has a GDP per capita of $45,000 but a birth rate of 32 per 1,000 — much higher than other countries at this income level. Identify what type of anomaly this is and suggest two possible geographical reasons for it.
Country X is a positive anomaly (higher birth rate than expected given its GDP level) — it lies above the line of best fit. Two possible reasons: (1) Cultural or religious factors — in some high-income countries in the Middle East, cultural and religious norms encourage larger families regardless of income level; women may have lower workforce participation, which is strongly correlated with birth rate. (2) Population structure / immigration — if the country has a large young adult immigrant population (e.g., a Gulf state like Qatar or UAE), the birth rate will be elevated by this young demographic structure even if GDP per capita is very high. Always explain why GDP alone doesn't determine birth rate — multiple variables are involved. Award marks for: identifying the anomaly type (1 mark); explaining that it lies above the line of best fit (1 mark); giving two plausible geographical reasons with explanation (1 mark each).
🧮 Statistical Skills: Mean, Median, Mode, Range and IQR
Geography at GCSE level requires you to calculate and interpret basic statistical measures. More importantly, you need to know WHICH measure to use and WHY — because choosing the wrong average can give a misleading picture of a dataset.
The Three Averages
| Measure | How to Calculate | When to Use It | Weakness |
|---|---|---|---|
| Mean | Add all values, divide by number of values | When data is roughly evenly distributed with no extreme outliers | Distorted by outliers — a single very large or very small value pulls the mean towards it |
| Median | Order values from smallest to largest; find the middle value (or average of two middle values if even count) | When data has outliers; income and wealth data; house prices | Ignores the actual values at the extremes — misses the magnitude of inequality |
| Mode | Most frequently occurring value | Non-numerical categories (e.g., most common land use type); when the "typical" response matters more than the average | A dataset can have no mode (all values different) or multiple modes; tells you nothing about spread |
Why Median Beats Mean for Income Data
This distinction carries marks in exams and illustrates a genuinely important principle. Consider a street with 9 households earning £25,000 per year and one household earning £1,000,000 per year.
- Mean income: (9 × £25,000 + £1,000,000) ÷ 10 = £1,225,000 ÷ 10 = £122,500
- Median income: The 5th value in the ordered list is £25,000
The mean says the "average" person on this street earns £122,500. The median says they earn £25,000. The median is far more representative of the actual experience of people on this street — the mean has been pulled upward by a single outlier. This is why economists and geographers use median household income, not mean household income, when analysing inequality and living standards.
Measures of Spread: Range and Interquartile Range
Averages tell you about the centre of a dataset. Measures of spread tell you how much variation there is around that centre — how consistent or variable the data is.
Range = Highest value − Lowest value. Simple to calculate, but vulnerable to the same problem as the mean: a single extreme outlier massively increases the range even if most values cluster tightly together.
Interquartile Range (IQR) = Q3 − Q1, where Q1 is the lower quartile (25th percentile) and Q3 is the upper quartile (75th percentile). The IQR measures the spread of the middle 50% of data, ignoring the extreme top and bottom 25%. It is much more resistant to outliers than the range and gives a better picture of how spread out the "typical" data values are.
Worked Example: River Discharge Data
Monthly mean discharge (m³/s) for a river: 12, 15, 22, 31, 45, 58, 62, 55, 38, 28, 19, 14
- Mean: (12+15+22+31+45+58+62+55+38+28+19+14) ÷ 12 = 399 ÷ 12 = 33.25 m³/s
- Median: Order the values: 12, 14, 15, 19, 22, 28, 31, 38, 45, 55, 58, 62. The two middle values (6th and 7th) are 28 and 31. Median = (28+31) ÷ 2 = 29.5 m³/s
- Range: 62 − 12 = 50 m³/s
- Q1 (lower quartile — median of lower half): median of 12, 14, 15, 19, 22, 28 = (15+19)/2 = 17 m³/s
- Q3 (upper quartile — median of upper half): median of 31, 38, 45, 55, 58, 62 = (45+55)/2 = 50 m³/s
- IQR: 50 − 17 = 33 m³/s
Interpretation: The river's discharge has a mean of 33.25 m³/s but peaks at 62 m³/s in summer (July) and falls to 12 m³/s in winter (January). The high range of 50 m³/s indicates significant seasonal variability — this has implications for flood risk management and water supply planning. The IQR of 33 m³/s shows that even the "middle 50%" of monthly discharges vary by 33 m³/s, confirming that this is not a stable, predictable river system.
Percentage Change
Percentage change allows fair comparison between values of different original sizes — it standardises the change as a proportion of the starting point.
Percentage change formula: ((New value − Original value) ÷ Original value) × 100
Example: A city's population increases from 2.4 million to 3.1 million over 20 years.
Percentage change = ((3.1 − 2.4) ÷ 2.4) × 100 = (0.7 ÷ 2.4) × 100 = +29.2%
Why percentage change matters: If City A grows from 100,000 to 150,000 people (+50,000) and City B grows from 2,000,000 to 2,050,000 (+50,000), City A has grown by 50% while City B has grown by only 2.5%. The absolute change is the same, but the relative change is dramatically different. Percentage change captures this difference; raw change does not.
📖 Key Terms
Trend — The overall direction of change in a dataset over time or across locations. Expressed qualitatively (increasing, decreasing, fluctuating, stable) and supported by specific data values. Example: "Global average temperature shows an upward trend from −0.2°C anomaly in 1880 to +1.2°C anomaly in 2020."
Anomaly — A single data point, period, or place that does not fit the general trend. Anomalies should be identified, described with a precise value, and explained with reference to geographical factors. A well-explained anomaly demonstrates interpretation skills and is worth significant marks.
Correlation — A statistical relationship between two variables where a change in one is associated with a change in the other. Positive correlation: both increase together. Negative correlation: one increases as the other decreases. Correlation does NOT prove causation.
Annual temperature range — The difference between the hottest and coldest monthly average temperatures in a climate graph (highest – lowest, in °C). A large range (over 20°C) typically indicates a continental location far from the sea. A small range (under 10°C) indicates a maritime or tropical location.
Birth rate — The number of live births per 1,000 people per year. A birth rate above 30 per 1,000 is considered high (typically LICs); below 15 per 1,000 is low (typical of HICs and ageing populations). Birth rate is shown on population pyramids by the width of the youngest cohorts.
Death rate — The number of deaths per 1,000 people per year. High death rates reflect limited healthcare access and high rates of disease; low death rates reflect effective medical systems and good living standards. On a population pyramid, rapid tapering above young cohorts indicates a high death rate.
Life expectancy — The average number of years a person born in a country can expect to live. Reflects healthcare quality, nutrition, and living conditions. On a population pyramid, shown by how far the bars extend upward — wide bars above age 70 indicate high life expectancy.
Dependency ratio — The ratio of "dependants" (those aged under 15 or over 65, who are not typically in paid work) to the working-age population (15–64). A high youth dependency ratio (as in LIC population pyramids) places financial pressure on a small working-age base to support schools, healthcare, and food supply for large young populations.
Median — The middle value in an ordered dataset. More representative than the mean when data contains outliers (extreme values). Used for income, house prices, and other economic data where a small number of very high values would distort the mean upward.
Interquartile range (IQR) — The range of the middle 50% of data values (Q3 − Q1). Less distorted by outliers than the full range, making it a more reliable measure of spread for datasets with extreme values.
Positive correlation — A relationship where as one variable increases, the other also tends to increase. On a scatter graph, points cluster around a line of best fit rising from bottom-left to top-right. Example: GDP per capita and life expectancy.
Negative correlation — A relationship where as one variable increases, the other tends to decrease. On a scatter graph, points cluster around a line of best fit falling from top-left to bottom-right. Example: GDP per capita and birth rate.
⚠️ Common Misconceptions
Misconception 1: "Correlation means causation — if the graph shows a relationship, one thing must be causing the other"
This is the single most important misconception in data skills and appears across geography, science, and statistics. A scatter graph showing a positive correlation between GDP and life expectancy does NOT prove that GDP causes longer life. It means they tend to increase together — but the mechanism is indirect (higher GDP → more government revenue → better-funded healthcare and sanitation → lower disease burden → longer life). There are also confounding variables: Cuba has a life expectancy of approximately 79 years despite a GDP per capita of only $9,000 — similar to some countries with life expectancy in the 60s. This anomaly suggests factors other than GDP (healthcare prioritisation, diet, social systems) can produce long life expectancy independently of income. Always write "is associated with" or "there is a positive correlation between" rather than "causes". Examiners actively look for and penalise causal language where it is not justified.
Misconception 2: "Good description means quoting as many figures from the graph as possible"
Transcribing numbers is not description — it is copying. A student who writes: "In January rainfall is 45 mm, in February it is 38 mm, in March it is 31 mm, in April it is 22 mm..." for each of 12 months has not described anything. They have reproduced data without identifying any pattern. The examiner already has the graph — they don't need the numbers repeated. What they want is: the overall trend, a comparison of highest and lowest, any anomaly, and a well-chosen figure or two as evidence. TACT (Trend, Anomaly, Comparison, Total/terminology) gives you the structure to do this. Three to five carefully chosen figures beat twelve passively listed ones every time.
Misconception 3: "The mean is always the best measure of average"
The mean is the most commonly used average, but it is also the most vulnerable to distortion by outliers. When data contains extreme values — income, house prices, rainfall totals, population data — the mean can be deeply misleading. A country with nine million people earning $5,000 per year and one billion-dollar fortune means the "average" income looks much higher than the typical person's experience. The median filters out this distortion by finding the genuine middle value. In GCSE geography questions that ask you to choose or justify a measure of average, always consider whether the data has outliers or skew — if it does, argue for the median.
Misconception 4: "A histogram and a bar chart are the same thing"
This is a common source of lost marks in graph-choice questions. A bar chart shows discrete categories with gaps between bars (the categories are separate). A histogram shows a continuous numerical scale divided into class intervals, with bars touching (because the data is continuous — there is no gap between "10–20 mm rainfall" and "20–30 mm rainfall"). The x-axis of a histogram is always a continuous numerical scale; the x-axis of a bar chart shows category names. Confusing these two graph types, or recommending the wrong one in an answer, costs marks. If the question asks you to recommend a graph type for showing "how many days had rainfall between different ranges", the answer is a histogram, not a bar chart.
🧠 The TACT Mnemonic — Never Forget How to Describe Data
The single most useful memory aid in this topic is the TACT framework. Every description of a graph, chart, or table should include all four elements:
| Letter | Stands For | What to Write |
|---|---|---|
| T | Trend | Overall direction — increasing / decreasing / fluctuating / stable. Quote start and end values. |
| A | Anomaly | Any point that doesn't fit the trend. Quote its value and suggest a reason. |
| C | Comparison | Directly compare highest to lowest, or two groups/places. Quote both values with units. |
| T | Total / Terminology / figures | Quote specific figures with units. Use precise vocabulary (annual temperature range, birth rate, etc.). |
Climate Graph Memory Aid
For reading a climate graph, use the initials HCWC (Hottest, Coldest, Wettest, Coldest-rainfall):
- Hottest month — identify the peak of the temperature line, note month and °C value
- Coldest month — identify the trough of the temperature line, note month and °C value, calculate annual range = H − C
- Wettest month — identify the tallest bar, note month and mm value
- Calculate total annual precipitation — add bars or estimate, note mm value and seasonal pattern
Population Pyramid Quick Key
- WIDE base = HIGH birth rate
- NARROW base = LOW birth rate
- TAPERS fast upward = HIGH death rate / LOW life expectancy (LIC)
- STAYS wide upward = LOW death rate / HIGH life expectancy (HIC)
- BULGE in working age = immigration or baby boom cohort
- PINCH in working age = emigration or one-child policy effect
Average Selection Rule
Remember this with "MILO":
- Mean — for normal, evenly spread data with no extreme values
- Income / inequality data — use Median, not Mean (outliers distort)
- List all values if asked for mode — it is the most frequently repeated number
- Outliers present? → Choose Median or IQR over Mean or Range
🎯 Exam Connection
Frequency: Graph and data skills questions appear in every geography paper from every exam board. AQA Paper 3 (Geographical Applications) is particularly data-heavy, with a dedicated "Issue Evaluation" section. OCR B includes data interpretation throughout all components. Edexcel uses climate graphs, population data and scatter graphs across all three papers.
Typical question stems:
- "Describe the trend shown in Figure X." (2–4 marks)
- "Compare the population pyramids for Country A and Country B." (4 marks)
- "Suggest reasons for the pattern shown in the climate graph." (4–6 marks)
- "What type of correlation is shown in Figure X? Describe the relationship." (2–3 marks)
- "Calculate the annual temperature range for Location X." (1–2 marks)
- "Describe and explain the population pyramid for Country A." (6 marks)
- "Using Figure X, describe the distribution of GDP per capita." (3 marks)
- "Assess the reliability of this data presentation." (4–6 marks)
Level progression for "Describe the population pyramid for Country A" (LIC, triangular shape):
- Level 1 (1–2 marks): "The country has lots of young people and not many old people. The pyramid is wide at the bottom and narrow at the top." — No data values, no specific terminology, no explanation of what this means.
- Level 2 (3–4 marks): "Country A has a wide base, suggesting a high birth rate, and the bars narrow rapidly above age 20. Life expectancy appears to be approximately 55–60 years as the bars become very narrow above 60. This triangular shape indicates a high birth rate and high death rate." — Some terminology, some approximate data values, identifies the characteristic shape.
- Level 3 (5–6 marks): "Country A's population pyramid has a wide base with the 0–4 cohort representing approximately 18–20% of the population, indicating a high birth rate of approximately 40 per 1,000. The pyramid narrows sharply above age 15, with each successive cohort noticeably smaller — this reflects a high death rate particularly in childhood, consistent with limited access to healthcare and high rates of preventable disease. The bars become very narrow above age 55, with less than 3% of the population over 65, suggesting a life expectancy of 55–60 years. This triangular pattern is characteristic of Stage 2 of the Demographic Transition Model and creates a high youth dependency ratio — the large proportion of under-15s places substantial pressure on educational infrastructure and food supply." — Specific percentage estimates, named DTM stage, dependency ratio concept, causal explanation, geographical implication.
Level progression for "Describe the climate graph for Location X" (Mediterranean climate — hot dry summer, wet mild winter):
- Level 1: "It is hot in summer and rainy in winter." — No data, no units, no terminology.
- Level 2: "The hottest month is July at about 28°C and the coldest is January at about 10°C, giving an annual temperature range of 18°C. Rainfall is concentrated in winter, with January being the wettest month at about 85 mm, and summer months receiving little rain." — Correct values with units, annual range calculated, seasonal pattern identified.
- Level 3: "The climate graph shows a Mediterranean pattern with a pronounced seasonal contrast. The temperature line peaks in July at approximately 28°C and falls to a minimum of 10°C in January, giving an annual temperature range of 18°C. The defining characteristic is the inverse relationship between temperature and precipitation — the hottest months (June–September) are also the driest, each receiving under 15 mm, while the coolest months (December–February) receive 70–90 mm. Total annual precipitation is approximately 580 mm. This pattern results from high-pressure anticyclone dominance in summer suppressing convective rainfall, and the northward migration of mid-latitude depressions bringing frontal rainfall in winter. The anomaly is the slight secondary precipitation peak in spring, which may reflect transitional weather systems between the winter westerlies and summer anticyclone." — Annual range calculated, seasonal pattern with mechanism, total annual precipitation, anticyclone/depression explanation, anomaly identified and explained.
Command word precision:
- "Describe": What you see. Use TACT. Quote figures with units. Do NOT explain why.
- "Explain": State the pattern AND explain the geographical process that causes it. What and why.
- "Suggest reasons for": The causes are not obvious from the data alone — use your geographical knowledge to interpret the pattern.
- "Analyse": Apply TACT AND identify relationships between variables AND evaluate what the data shows and what it does not show.
- "Assess the usefulness/reliability": What does this data presentation show clearly? What does it hide or misrepresent? What would a more reliable presentation look like?
💡 Exam Tips for Graph and Data Skills
🎯 The Three Rules for Top Marks
- Always quote figures with units. "High temperature" = Level 1. "Maximum temperature of 32°C in July" = Level 2–3. Units are compulsory: write °C, mm, per 1,000, km², not just numbers.
- Use TACT as your structure, not as a checklist. Don't write four separate sentences headed "Trend:", "Anomaly:", "Comparison:", "Total:". Weave the elements together into coherent description. The examiner rewards analysis, not headings.
- If asked to compare, compare directly. "Country A has a birth rate of 38 per 1,000 whereas Country B has 12 per 1,000 — more than three times higher" is a direct comparison. Writing about each country separately and hoping the examiner notices the difference is not comparison — it is description.
📝 Common Command Word Traps
- "Describe" does NOT mean "explain". If a question says "Describe the pattern shown", do not explain why the pattern exists. You will not get extra marks for explaining when the question asks for description — and you waste time that could be spent on a later question.
- "Suggest reasons" means you must give a geographical explanation. "Suggest reasons for the anomaly shown" is asking you to apply geographical knowledge, not just identify the anomaly.
- "Analyse" requires both description AND interpretation. It is NOT enough to describe the data. You must also explain what it means, what relationship it shows, and ideally what the data does NOT show or what its limitations are.
⚠️ Graph Choice Questions — The Most Missed Marks
- Never recommend a pie chart for time-series data. Pie charts cannot show change over time. If data has a time dimension, the answer is a line graph (if continuous) or a bar chart (if discrete time periods).
- Never recommend a bar chart for frequency distribution. If the question asks about distributing data into class intervals (e.g., how many days had wind speeds in each range), the answer is a histogram.
- When recommending a graph, always justify your choice. "A scatter graph would be most appropriate because it shows the relationship between two continuous variables — GDP per capita and life expectancy — and allows a line of best fit to be drawn to show the direction and strength of the correlation." Naming the graph type alone scores 1 mark; justifying it scores 2–3.
🧮 Quick Calculation Checks
- Annual temperature range: Always Highest month − Lowest month in °C. Easy mark — always do this calculation explicitly when reading a climate graph.
- Percentage change: ((New − Old) ÷ Old) × 100. If the result is positive, it's an increase; if negative, it's a decrease. Include the % sign.
- Mean check: If your calculated mean looks implausibly high or low, you have probably made an arithmetic error or included an outlier without checking. Always sense-check against the data — the mean should fall somewhere within the middle range of values.
- Median with even-numbered datasets: Average the two middle values. For 12 monthly values, the median is the average of the 6th and 7th ordered values.
Quick Check: A student is given data on annual rainfall for 10 towns (in mm): 450, 520, 480, 510, 490, 530, 475, 2,100, 505, 495. Calculate: (a) the mean, (b) the median, and (c) explain which average better represents the "typical" rainfall for these towns and why.
(a) Mean: Add all values: 450+520+480+510+490+530+475+2100+505+495 = 7,555. Divide by 10 = 755.5 mm. (b) Median: Order values: 450, 475, 480, 490, 495, 505, 510, 520, 530, 2100. The 5th and 6th values are 495 and 505. Median = (495+505)÷2 = 500 mm. (c) The median (500 mm) better represents the typical rainfall. The mean (755.5 mm) has been pulled upward by the single outlier value of 2,100 mm — likely a coastal or upland town with exceptionally high orographic rainfall, or possibly a measurement error. If you told someone the "average" rainfall was 755.5 mm, they would have a completely misleading picture of rainfall across these towns, since 9 out of 10 towns receive between 450 and 530 mm. The median of 500 mm accurately represents what a typical town in this dataset experiences. Award: (a) 755.5 mm (1 mark); (b) 500 mm (1 mark); (c) identifies median as more appropriate (1 mark) + explains that the mean is distorted by the outlier (2,100 mm) (1 mark) + states this gives a misleading picture of typical rainfall (1 mark).
Make this stick
Signing in gives access to all PrepWise features.