Why is "Naming 'the internet' or 'a government website' instead of a specific agency" a common mistake in Data Sources?

Why it happens: Students don't know the specific bodies. How to avoid it: Memorise the key sources: UK rivers → Environment Agency; UK Census → ONS; earthquakes → USGS; development → World Bank + UNDP; weather → Met Office. Naming is what earns marks. Source: Pearson 4GE1 Examiner Reports — Paper 2 Section B

Why is "Treating primary fieldwork as a complete enquiry, without secondary data" a common mistake in Data Sources?

Why it happens: Focus on fieldwork; secondary feels like 'cheating'. How to avoid it: Pearson 4GE1 spec EXPLICITLY requires BOTH. Always cite at least 2-3 secondary sources alongside primary fieldwork. Triangulation is the strongest method. Source: Pearson 4GE1 Examiner Reports — Paper 2 Section B

Why is "Using secondary data without evaluating its reliability" a common mistake in Data Sources?

Why it happens: Government data feels authoritative. How to avoid it: Every secondary source has strengths AND weaknesses. State both. 'ONS Census 2021 is high-quality but decadal — 2026 data is 5 years out of date.'

Why is "Using the wrong source for the question (e.g. UK Environment Agency for global development)" a common mistake in Data Sources?

Why it happens: Students apply familiar sources rather than the right source. How to avoid it: Match source to question: UK rivers → EA; global development → World Bank + UNDP; UK Census → ONS; earthquakes → USGS. Use the SPECIALIST for each topic.

Why is "Treating newspaper reports as factual + technical sources" a common mistake in Data Sources?

Why it happens: Newspapers feel like they're 'reporting the news'. How to avoid it: Newspapers carry narrative bias + early-figure errors (Haiti 2010 death tolls: ~100k early vs ~316k final — a 3× range). Use newspapers for QUALITATIVE context; verify TECHNICAL claims against scientific sources (USGS, UN OCHA). Source: Pearson 4GE1 Examiner Reports — Paper 2 Section B

Why is "Relying on a single data source for the conclusion" a common mistake in Data Sources?

Why it happens: Easier; familiar. How to avoid it: Always TRIANGULATE: combine 1-2 primary + 2-3 secondary sources. When all agree, conclusion is robust. When they disagree, that's also informative. Source: Pearson 4GE1 Examiner Reports — Paper 2 Section B

Edexcel IGCSE Geography (4GE1)

Data Sources

0% Complete

Detailed Study Notes

Detailed notes on Development and human welfare for Edexcel IGCSE Geography, covering key concepts, explanations, examples, and exam-focused revision points.

Data Sources (Primary + Secondary) — Pearson Edexcel International GCSE Geography (4GE1) Study Notes

Choosing + citing data sources for geographical enquiry. Primary = student-collected. Secondary = collected by others. The strongest enquiries TRIANGULATE both. Covers named secondary sources for each 4GE1 topic: UK rivers (Environment Agency), UK Census (ONS), urban (ONS + council data), coastal (Met Office, Channel Coastal Observatory, EA), hazards (USGS, UN OCHA), development (World Bank Open Data, UNDP HDR, OECD ODA, WHO). Citing conventions + bias awareness + reliability evaluation. Pearson examiners reward students who NAME specific agencies + JUSTIFY each choice.

At a glance

Primary = student-collected (your measurements / surveys / interviews).
Secondary = collected by others (government, agencies, academics, news).
Triangulation = use multiple sources whose biases differ — agreement is robust, disagreement is informative.
Key UK sources: ONS (Census + urban), Environment Agency (rivers + coast + flooding), Met Office (weather), OS (maps), BGS (geology).
Key global sources: USGS (earthquakes), NOAA (cyclones), UN OCHA (humanitarian), World Bank (development), UNDP HDR (HDI), WHO (health), FAO (food + agriculture).
Name specific agencies — not 'a government website' — to score marks.
Evaluate reliability: who collected, when, how, what sample, what bias.
Pearson examiners reward students who triangulate + name + evaluate.

What you’ll learn

Mapped to the Edexcel IGCSE 4GE1 syllabus (2026 onwards).

Distinguish primary from secondary data and give specific examples of each.
Name the appropriate specialist source for each 4GE1 enquiry theme (rivers, urban, hazards, coastal, development).
Evaluate strengths + weaknesses + reliability of secondary sources.
Combine primary + secondary data through triangulation in a fieldwork enquiry.
Cite specific sources accurately in fieldwork write-up.

Primary vs secondary

Primary = your own data; secondary = others'. Use BOTH for the strongest enquiry.

Primary data is data the researcher has collected themselves through original fieldwork.

Examples:

Pedestrian + traffic counts at chosen sites.
Beach profile measurements with ranging poles + clinometer.
River velocity by float method.
EQS scores at each site.
Interviews + questionnaires the student designs + carries out.
Field sketches + fixed-point photographs.

Strengths. Directly relevant to your hypothesis + sites; you know exactly how it was collected; you OWN the data; can target the specific question.

Weaknesses. Time-consuming; limited to one day; small sample size; observer skill matters; cannot capture long-term trends.

Secondary data is data collected by someone else (government, agency, academic, NGO, company, news) made available for re-use.

Examples:

UK Census 2021 (ONS).
Environment Agency river-flow gauging records.
Met Office rainfall + weather records.
OS topographic maps + historic OS maps from 1850.
BGS geology maps.
USGS earthquake catalogue.
World Bank GNI data.
UNDP HDI rankings.
WHO health indicators.
News + photographic archives.
Aerial photographs.
Google Earth imagery + Street View.
GIS layers.

Strengths. Massive scale; long time periods (decades, centuries); collected by experts with calibrated instruments; usually free; produces context primary cannot.

Weaknesses. Not specific to your hypothesis OR site; AGGREGATED at unit level that may hide variation; may be DATED; methodology may be unknown or differ between time periods.

Best practice — TRIANGULATE.

Primary: 'Beach width at Mappleton on 12 May 2026 = 38 m.'

Secondary (UK Environment Agency): 'This is consistent with the post-1991 sediment trapping pattern documented in the Holderness historic record.'

Combined: 'My measurement confirms the long-term EA pattern.'

The two together are stronger than either alone. Pearson 4GE1 mark schemes explicitly credit combining + naming.

Primary = your own data (counts, measurements, surveys, interviews).
Secondary = others' data (government, agencies, academics, news, satellites).
Primary strengths: relevant + owned. Weaknesses: snapshot + small.
Secondary strengths: scale + multi-year + expert. Weaknesses: generic + dated.
ALWAYS use BOTH — triangulation is the strongest enquiry method.

Named secondary sources by theme

The right source for the right question. Pearson examiners reward naming.

Pearson 4GE1 mark schemes explicitly credit students who NAME specialist sources. Vague references ('the internet', 'a government website') score 0; NAMED references ('UK Environment Agency Catchment Data Explorer', 'UNDP Human Development Report 2024') score the mark.

For UK rivers + flooding:

Source	What it provides
UK Environment Agency	River-flow gauging data (m³/s) at named stations going back decades; flood-risk maps; Catchment Data Explorer; water-quality monitoring
Met Office	Antecedent rainfall + storm records — vital for explaining river flow + flood events
OS topographic + historic maps	Channel position + flood-plain land-use change (1850-present)
BGS geology	Underlying geology + soils control infiltration vs runoff

For UK urban fieldwork:

Source	What it provides
UK ONS (Census 2021)	Population, housing, ethnicity, deprivation, transport, by output area + ward + district
Council planning portals	Development applications + decisions; rezoning
UK Index of Multiple Deprivation (IMD)	Multi-dimensional deprivation by lower-layer super output area
Department for Transport	Traffic data; public-transport patronage
OS + Google Earth + Street View	Building density + land-use mapping; visual comparison over time

For coastal fieldwork (UK):

Source	What it provides
UK Environment Agency	Coastal-erosion records; Shoreline Management Plans
Met Office	Wind, wave, storm data; tide forecasts
Channel Coastal Observatory	Beach surveys, photographs, aerial surveys — UK south coast specialist
OS + historic OS	Coastline change over decades

For hazards (worldwide):

Source	What it provides
USGS (US Geological Survey)	Global earthquake catalogue; ShakeMaps; volcano monitoring
NOAA National Hurricane Center	Tropical-cyclone tracks, intensity, forecasts
UN OCHA ReliefWeb	Humanitarian impact data — deaths, displaced, response, funding
WHO	Health impacts (disease outbreaks, injury, mental health)
Insurance + reinsurance reports (e.g. Munich Re)	Economic damage

For development indicators:

Source	What it provides
World Bank Open Data	GNI per capita, GDP, education, health, internet by country + year (1960-present)
UNDP Human Development Report	HDI rankings — composite of income + education + health
OECD ODA data	Official development assistance flows by donor + recipient + sector
UN Population Division (UNDESA)	Population pyramids, fertility, urbanisation by country + year
WHO + UNICEF	Health indicators (infant + maternal mortality, malnutrition, immunisation)
FAO	Food + agriculture data

Citing conventions. In a fieldwork write-up, cite as: Author / Agency, Year, Title, URL (if online). Example:

UK Environment Agency (2024) Catchment Data Explorer — River Severn at Bewdley. https://environment.data.gov.uk/

For news / photos cite the publication + date + photographer.

Examiner tip. Cite at least 2-3 SPECIFIC SECONDARY sources in any fieldwork write-up. Pearson 4GE1 mark schemes reward this consistently.

Rivers / flooding → Environment Agency + Met Office + OS.
Urban → ONS + council + IMD + Department for Transport.
Coastal → Environment Agency + Met Office + Channel Coastal Observatory + OS.
Hazards → USGS + NOAA + UN OCHA + WHO.
Development → World Bank + UNDP HDR + OECD ODA + WHO + FAO.
Cite agency name + year + URL where possible.

Pros + cons of secondary sources

Every source carries strengths AND weaknesses. Reliability depends on who, when, how, and bias.

Secondary sources are NOT all equal. Reliability depends on:

1. WHO collected it.

Government agencies + UN bodies (ONS, USGS, World Bank, WHO) — most reliable; transparent methodology; subject to audit.
Peer-reviewed academic journals — high reliability; methodology + uncertainty published.
NGOs + advocacy groups — useful but may carry advocacy bias.
News media — useful for human context; weak for technical detail.
Commercial sources (e.g. company reports) — may have commercial bias.
Wikipedia / random websites — variable; verify against primary academic + government sources.

2. WHEN it was collected.

Recent data is generally better for CURRENT analysis.
Historic data is unique + irreplaceable for CHANGE analysis (OS maps 1850-present; long Environment Agency records).
BEWARE: data more than 5-10 years old may be misleading for fast-changing topics (urban, migration, technology, climate-affected systems).
UK Census is decadal — 2021 data in 2026 is 5 years old; 2031 Census won't be published until ~2034.

3. HOW it was collected.

Census = near-total survey (almost all UK residents) — high reliability.
Sample survey = depends on sampling design; if biased, results are biased.
Continuous monitoring (river gauges) = high reliability if station is maintained.
One-off study = depends on methodology.
Crowdsourced (e.g. some climate observations) = variable quality.

4. WHAT BIAS it carries.

Coverage bias — some places + populations are under-measured (poorest countries, marginal urban areas, informal sectors).
Definition bias — what counts as 'urban' vs 'rural' differs by country; what counts as 'unemployed' affects rates.
Reporting bias — countries may inflate or suppress data for political reasons.
Methodological bias — methodology changes between time periods make comparisons tricky.
Narrative bias — news media over-reports drama, under-reports recovery.

5. WHAT IT MEASURES vs WHAT IT MISSES.

GDP per capita measures formal economy; misses informal sector + household + care work.
HDI measures income + education + health; misses environment, freedom, happiness.
Census measures usual residence; misses homeless, undocumented migrants.
River-gauging data measures flow at the station; misses ungauged tributaries.

Reliability verdict format. When citing a secondary source, evaluate it explicitly:

'UK Environment Agency river-flow data at Bewdley gauging station is highly reliable for the period 1980-present (consistent methodology + continuous monitoring), but earlier data has gaps + the station was upgraded in 1992 so pre-1992 readings should be treated with caution.'

Pearson 4GE1 mark schemes credit students who EVALUATE — not just cite — their sources.

WHO: gov + UN + academic = high; advocacy + news + Wikipedia = variable.
WHEN: recent for current; historic for change. Census decadal.
HOW: census = near-total; sample = depends on design; monitoring = depends on station.
BIAS: coverage, definition, reporting, methodological, narrative.
WHAT IT MEASURES vs MISSES: every measure is partial.
Always state reliability verdict — not just citation.

Triangulating data sources

Use multiple sources whose biases differ. When they agree, conclusion is robust. When they disagree, that's informative.

Triangulation is the use of multiple data sources to investigate the same question.

Why it works. Different sources carry different biases. When multiple sources whose biases are INDEPENDENT all point to the same conclusion, the conclusion is far more likely to be reliable than if it came from any single source.

Example — testing 'pedestrian footfall is higher in the CBD than the inner suburb'.

Primary data:

Pedestrian counts at 6 sites along a CBD → suburb transect, 10-minute slots, 3 time slots × 2 days.

Secondary data:

UK ONS Census 2021 — population density + journey-to-work data for the area.
UK Department for Transport — traffic data on the main road.
Council retail footfall data (where available).
Google Earth + Street View for spatial context.

When the primary count + ONS Census + DfT + Google Earth ALL show the CBD-to-suburb gradient, the conclusion is multi-source robust.

When they disagree — perhaps your primary count caught an unusual day, or the Census data is 5 years old + the area has changed — the disagreement is informative + drives further investigation.

Example — testing 'beach width is greater updrift of groynes than downdrift' (Holderness).

Primary data:

Beach width + profile + sediment measurements at 5 sites either side of Mappleton groynes.

Secondary data:

Environment Agency historic beach surveys for the Holderness coast.
Channel Coastal Observatory aerial photos (UK south coast specialist — for comparable methodology).
Historic OS maps (1850-present) showing coastline change.
Academic literature on Holderness coastal management.
News + photographic record of Mappleton-Withernsea downdrift starvation.

When all sources tell the same story — that Mappleton groynes have built up beach updrift + starved Withernsea downdrift — the conclusion is robust + uncontroversial.

Example — testing 'global development gaps persist between MEDCs + LEDCs'.

Secondary sources:

World Bank Open Data — GNI per capita, comparable across 200+ countries 1960-present.
UNDP HDI — composite measure.
WHO + UNICEF — health indicators.
UN Population Division — demographic structure.
FAO — food + agriculture.
OECD ODA — aid flows.

All point to a persistent gap (whose magnitude varies by measure but whose existence is robust). Triangulation makes the conclusion defensible.

When sources disagree — investigate why.

Different scales (national vs ward).
Different methodologies.
Different time periods.
Different biases.
Different aspects of the same reality.

The DISAGREEMENT often reveals the most interesting geography. Pearson 4GE1 mark schemes specifically credit students who triangulate + investigate disagreement.

Examiner tip. A conclusion based on PRIMARY + at least 2-3 SECONDARY sources cited by name is significantly stronger than one based on a single source. Always TRIANGULATE.

Multiple sources whose biases differ → agreement = robust conclusion.
Disagreement is informative + drives further investigation.
Primary + 2-3 named secondary sources is the standard for top-band answers.
Triangulation reveals when conclusions are over-claimed.

Citing conventions + bias awareness

Cite properly. Evaluate explicitly. Acknowledge what you don't know.

Citing conventions in fieldwork write-up.

For any secondary source, cite in this format:

Author / Agency (Year). Title. Publisher / URL.

Examples:

UK Environment Agency (2024). Catchment Data Explorer — River Severn at Bewdley. https://environment.data.gov.uk/
UK ONS (2022). Census 2021 — Topic Summaries. https://www.ons.gov.uk/census
USGS (2024). Earthquake Catalog. https://earthquake.usgs.gov/earthquakes/
UNDP (2024). Human Development Report 2024. New York: UNDP.
World Bank (2024). World Development Indicators. https://data.worldbank.org/

For news + photos: Publication, Date, Photographer (if known).

Include a SOURCES section at the end of your write-up listing all secondary sources used.

Bias awareness. Every source carries bias. The MATURE response is not to despair of objectivity but to ACKNOWLEDGE it explicitly.

Five questions to ask of every source:

WHO collected it? (gov / UN / academic / NGO / commercial / news — affects framing)
WHEN? (recent or historic — affects relevance)
WHY? (commercial / political / academic motivation — may shape framing)
WHAT did it capture + miss? (every measure is partial)
WHAT BIASES does it carry? (coverage, definition, reporting, narrative)

Honest acknowledgement. State explicitly:

'World Bank GNI per capita data captures formal-economy activity well; informal sector (subsistence agriculture, household labour, care work) is largely invisible. My conclusion that LEDCs have lower 'development' than MEDCs is reliable as a measure of formal-economy productivity but does not capture all dimensions of human well-being.'

The Pearson 4GE1 mark schemes consistently credit students who:

NAME specific sources.
JUSTIFY their choice.
EVALUATE reliability + bias.
TRIANGULATE with primary + multiple secondary sources.
ACKNOWLEDGE what each source misses.

These five disciplines together separate a credible enquiry from an over-claimed one.

Examiner tip. In any fieldwork question, mention AT LEAST ONE specific source by full name + give one strength + one weakness. This earns several marks at little time cost.

Cite as Author / Agency (Year). Title. URL.
Include a Sources section at end of write-up.
Ask the 5 source-evaluation questions: who / when / why / what captured / what missed.
ACKNOWLEDGE bias explicitly — over-claiming loses marks.
NAME + JUSTIFY + EVALUATE + TRIANGULATE + ACKNOWLEDGE = top-band.

Quick recap

Primary = your data; Secondary = others' data; use BOTH.
Triangulate primary + 2-3 named secondary sources.
Name specific agencies for each topic (EA, ONS, USGS, World Bank, UNDP, WHO, OS, BGS, Met Office, UN OCHA).
Evaluate reliability: who, when, how, what bias, what missed.
Cite properly: Author/Agency (Year), Title, URL.
Pearson examiners reward NAMING + JUSTIFICATION + EVALUATION.
When sources agree → robust conclusion. When they disagree → informative.
Honest acknowledgement of what each source misses lifts a good enquiry to A*.

Memorise this

Verbatim phrases and definitions Edexcel mark schemes credit.

Primary = student-collected; Secondary = others.
Triangulation = multiple sources whose biases differ.
UK rivers → Environment Agency. UK Census → ONS. Earthquakes → USGS. Development → World Bank + UNDP HDR. Weather → Met Office. Geology → BGS. Hazards humanitarian → UN OCHA. Health → WHO.
5 source-evaluation questions: who / when / why / what captured / what missed.
Cite as Author/Agency (Year). Title. URL.

How it’s examined

Spec Step 3 (Data sources + primary/secondary) is tested in Paper 2 Section B fieldwork (out of 70 marks).

Typical Paper 2 Section B question styles:

Define (1-2 marks) — define primary / secondary / triangulation.
Identify (1-3 marks) — name secondary sources for a given enquiry theme.
Compare (4 marks) — pros + cons of primary vs secondary.
Justify (4-6 marks) — design a primary + secondary data package for a hypothesis + justify each choice.
Evaluate (4-6 marks) — evaluate reliability of a given source OR set of sources.

Mark-scheme keywords examiners credit: primary data, secondary data, triangulation, ONS, Census, UK Environment Agency, Met Office, OS, BGS, USGS, NOAA, UN OCHA, ReliefWeb, World Bank Open Data, UNDP HDR, HDI, WHO, UNICEF, FAO, OECD ODA, citing, bias, reliability, methodology, sample, coverage, recent / dated, replicability.

Pearson 4GE1 Appendix 5 + Appendix 7 explicitly require students to use BOTH primary AND secondary data. Always cite at least 2-3 named secondary sources alongside your fieldwork.

Sources: Pearson Edexcel International GCSE Geography (4GE1) Specification — Issue 4, February 2026, Appendix 5 + Appendix 7; UK Office for National Statistics (ONS) — UK Census 2021; UK Environment Agency — Catchment Data Explorer + Flood Map; USGS Earthquake Hazards Program; World Bank Open Data + UNDP Human Development Reports; UN OCHA ReliefWeb; Pearson 4GE1 Examiner Reports — Paper 2, Section B. Last reviewed 2026-06-03.

Take this whole topic with you

Step-by-step worked examples — Data Sources

Step-by-step solutions to past-paper-style questions on data sources, written exactly the way a tutor would explain them at the board.

1Primary vs secondary
Getting started• AO1, AO4
Question
Define PRIMARY and SECONDARY data and state one PRO + one CON of each. [4]
Step-by-step solution
1. Step 1
  Primary (1 mark). Data the researcher has collected themselves through original fieldwork. Pro: directly relevant to the hypothesis + you know how it was collected. Con: time-consuming + limited in scale.
2. Step 2
  Secondary (1 mark). Data collected by someone else (books, agencies, government, online). Pro: large scale + long time periods + uses expensive equipment you don't have. Con: not specific to your site + can be out of date or biased.
3. Step 3
  Best practice (1 mark — synthesis). Use BOTH: primary for the specific finding, secondary for the wider context.
4. Step 4
  Example (1 mark). Primary = pedestrian count outside a CBD store. Secondary = UK ONS Census 2021 footfall + retail data for the area.
Answer
Primary = student-collected (relevant + owned but small-scale); Secondary = others' data (large + long-term but generic). Use BOTH for the strongest enquiry. Example: primary pedestrian count + secondary ONS Census 2021.
Examiner tip
Pearson 4GE1 mark schemes credit defs + named secondary source (e.g. 'ONS', 'Environment Agency'). 'Online sources' alone is too vague.
2Naming specific secondary sources
Getting started• AO1
Question
Match each ENQUIRY THEME to a NAMED secondary data source: (a) UK rivers; (b) UK Census + urban; (c) Earthquakes worldwide; (d) Development indicators; (e) Tropical-cyclone hazards. [5]
Step-by-step solution
1. Step 1
  (a) UK rivers (1 mark) — UK Environment Agency. River-flow gauging stations + flood-risk maps + Catchment Data Explorer.
2. Step 2
  (b) UK Census + urban (1 mark) — ONS (Office for National Statistics). UK Census 2021 — population, housing, deprivation by ward.
3. Step 3
  (c) Earthquakes worldwide (1 mark) — USGS (US Geological Survey). Earthquake catalogue, magnitudes, ShakeMaps.
4. Step 4
  (d) Development indicators (1 mark) — World Bank Open Data OR UNDP Human Development Report. GNI per capita, HDI, education, health by country + year.
5. Step 5
  (e) Tropical-cyclone hazards (1 mark) — NOAA National Hurricane Center OR UN OCHA ReliefWeb. Storm tracks, casualties, humanitarian impacts.
Answer
(a) UK Environment Agency. (b) UK ONS Census. (c) USGS. (d) World Bank Open Data / UNDP HDR. (e) NOAA / UN OCHA.
Examiner tip
Pearson 4GE1 explicitly rewards students who NAME agencies. 'A government website' = no mark; 'UK Environment Agency Catchment Data Explorer' = full mark.
3Pros + cons of primary + secondary
Building confidence• AO3, AO4
Question
Compare the STRENGTHS + WEAKNESSES of primary vs secondary data for a coastal-management enquiry. [4]
Step-by-step solution
1. Step 1
  Primary strengths (1 mark). Directly relevant to your hypothesis + site (e.g. beach width at YOUR study sites on YOUR field day); you know exactly how it was collected; you OWN the data.
2. Step 2
  Primary weaknesses (1 mark). Time-consuming + limited to one day / one season; small sample; affected by your weather + observer skill; cannot capture long-term trends.
3. Step 3
  Secondary strengths (1 mark). Massive scale + long time periods (Environment Agency historic records back decades; OS maps from 1850s); collected by experts with expensive equipment; cheap or free.
4. Step 4
  Secondary weaknesses (1 mark). Not specific to your hypothesis OR site; may be DATED or aggregated at the wrong scale (e.g. county-level when you need ward); collection method may be unknown OR biased.
Answer
Primary = relevant + owned BUT small-scale + snapshot. Secondary = scale + expertise + multi-year BUT generic + dated. Combine both.
Examiner tip
Top-band 4-mark answer needs TWO contrasts (strengths + weaknesses of EACH). Naming SPECIFIC sources lifts to top of mark band.
4Triangulating primary + secondary
Building confidence• AO3, AO4
Question
Explain how a student investigating river flooding on the River Severn (UK) could TRIANGULATE primary + secondary data sources. [6]
Step-by-step solution
1. Step 1
  Primary data collection (1 mark). Student measures river width, depth, velocity (float method × 0.85) at 5 sites along ~5 km of the River Severn. Photographs of flood debris lines. Annotated sketches of channel cross-sections.
2. Step 2
  Secondary 1 — UK Environment Agency (1 mark). River-flow gauging-station data — discharge in m³/s at named sites going back decades. Compare your field day's discharge to long-term mean + recent extremes. Flood-risk maps online (Environment Agency Flood Map).
3. Step 3
  Secondary 2 — Met Office (1 mark). Rainfall records for the preceding week + month + recent storm events. Severn floods often track antecedent rainfall in Wales.
4. Step 4
  Secondary 3 — OS maps + historic OS maps (1 mark). Show channel position + flood-plain land use; historic maps show how the floodplain has been built over.
5. Step 5
  Secondary 4 — News + photographic archives (1 mark). Photos of historic Severn floods (e.g. 2000, 2007, 2020) — visual evidence of past extents.
6. Step 6
  Triangulation logic (1 mark). When primary data + Environment Agency long-term records + Met Office rainfall + OS land use ALL agree, conclusion is robust. When they DISAGREE, investigate why — often reveals the most interesting geography.
Answer
Primary: width / depth / velocity + photos + sketches at 5 sites. Secondary: (1) Environment Agency river-flow + flood-risk maps; (2) Met Office antecedent rainfall; (3) OS + historic OS for floodplain land use; (4) news + photo archives for past extremes. Triangulation: when all agree, conclusion is robust.
Examiner tip
Top-band 6-mark answer needs PRIMARY + multiple SECONDARY sources + an explicit triangulation point. Pearson 4GE1 mark schemes credit NAMED agencies (Environment Agency, Met Office, OS).
5Evaluating bias in secondary sources
Stretch• AO3
Question
A student is using NEWSPAPER REPORTS as a secondary source for an enquiry into the 2010 Haiti earthquake. EVALUATE the reliability of newspaper reports + suggest alternative or complementary secondary sources. [6]
Step-by-step solution
1. Step 1
  Strengths of newspapers (1 mark). Eyewitness accounts; immediate human impacts; photos; named individuals; gives 'on-the-ground' qualitative texture missing from technical sources.
2. Step 2
  Weakness 1 — early-stage uncertainty (1 mark). Haiti 2010 official death tolls varied between ~100,000 (early reports) and ~316,000 (final Haitian government figure). Newspaper reports based on early estimates are systematically WRONG by a factor of 3.
3. Step 3
  Weakness 2 — narrative bias (1 mark). Newspapers select dramatic stories + photos. They UNDER-REPORT recovery + over-report disaster. The narrative shaped global aid response but may distort historical understanding.
4. Step 4
  Weakness 3 — national perspective (1 mark). US + UK papers emphasise their own response (US military aid; UK aid). Haitian + Caribbean papers offer different perspectives. A student using only Western papers gets a culturally narrow view.
5. Step 5
  Complementary sources (2 marks). (a) USGS — earthquake magnitude (M7.0), epicentre, ShakeMap, technical detail. (b) UN OCHA ReliefWeb — official humanitarian response data + casualties + displaced. (c) Haitian government final reports — final official death toll + reconstruction data. (d) World Bank + IMF reports — economic impact + reconstruction financing. (e) Academic literature — peer-reviewed analyses of the disaster.
Answer
Newspapers give human texture BUT early death tolls vary 100k-316k; narrative bias + national perspective. Complement with USGS (technical), UN OCHA ReliefWeb (humanitarian), Haitian government final reports, World Bank reconstruction data + peer-reviewed literature. Triangulation reveals bias.
Examiner tip
Top-band 6-mark answer needs MULTIPLE WEAKNESSES + SPECIFIC alternatives. The Haiti death-toll range (100k vs 316k) is the killer example showing why source choice matters.
6Designing a data-sources package
Stretch• AO3, AO4
Question
A student is investigating 'GLOBAL DEVELOPMENT GAPS' between MEDCs + LEDCs. Design a SECONDARY-DATA package + justify each source. [6]
Step-by-step solution
1. Step 1
  Source 1 — World Bank Open Data (1 mark). GNI per capita, life expectancy, school enrolment, internet penetration by country + year. JUSTIFY: largest free open dataset; data going back 60+ years; comparable across countries.
2. Step 2
  Source 2 — UNDP Human Development Report (1 mark). HDI scores + rankings; composite index of income + education + health. JUSTIFY: HDI is the single most-cited development metric; published annually; allows country comparisons.
3. Step 3
  Source 3 — OECD ODA data (1 mark). Official Development Assistance flows — who gives + receives + how much. JUSTIFY: shows aid as a development response; reveals donor priorities.
4. Step 4
  Source 4 — UN Population Division (1 mark). Population pyramids, fertility, urbanisation by country + year. JUSTIFY: demographic context for development; reveals the 'demographic dividend' or burden facing each country.
5. Step 5
  Source 5 — WHO / UNICEF (1 mark). Health indicators (infant mortality, maternal mortality, malnutrition, immunisation). JUSTIFY: SDG-aligned + globally comparable; reveals health-development links.
6. Step 6
  Source 6 — Country statistical bureaux + regional bodies (e.g. ASEAN, AU) (1 mark). National data may differ from international estimates. JUSTIFY: triangulating against international data reveals potential national bias.
Answer
(1) World Bank Open Data — GNI, education, health. (2) UNDP HDR — HDI scores. (3) OECD ODA — aid flows. (4) UN Population Division — demographics. (5) WHO / UNICEF — health. (6) National statistical bureaux — triangulation. Each justified by its specific role + comparability.
Examiner tip
Top-band 6-mark answer needs MULTIPLE NAMED sources + JUSTIFICATION for each. Pearson 4GE1 mark schemes specifically credit students who use the right development source for the right question.

Model Answers — Data Sources

High-scoring sample answers for data sources on the Cambridge IGCSE paper, with examiner-style notes mapping each response to the mark scheme and assessment objectives.

Question 1
2 marks
[EASY] Define 'secondary data' and give ONE named example used in coastal fieldwork. [2]
Model answer
Secondary data is data that has been collected by someone else (a government agency, university, company, NGO or other researcher) and made available for re-use. It contrasts with primary data, which is collected by the student themselves.

Example for coastal fieldwork: the UK Environment Agency's cliff-retreat monitoring records for the Holderness coast, which go back decades and would be impossible for a student to collect in one field day.
Why this scores
Mark scheme: 1 mark for 'collected by someone else and re-used'; 1 mark for a NAMED specific source (Environment Agency, OS, BGS, Met Office, etc.) — not 'the internet'.
Question 2
3 marks
[EASY] What does it mean to TRIANGULATE data sources, and why is it important? [3]
Model answer
Triangulation is the use of MULTIPLE different data sources (primary + secondary, quantitative + qualitative, different agencies) to investigate the same question.

Why important.

If all sources AGREE, the conclusion is robust + trustworthy.

If sources DISAGREE, the disagreement itself is informative — it often points to bias in one source, a real spatial / temporal mismatch, or an interesting underlying complexity.

Triangulation reduces the risk that the conclusion depends on the limitations of any SINGLE source.
Why this scores
Mark scheme: 1 mark for using multiple sources; 1 mark for 'agreement = robust'; 1 mark for 'disagreement = informative / reveals bias'.
Question 3
4 marks
[MEDIUM] Name FOUR named secondary data sources, stating what each provides + a possible weakness. [4]
Model answer
1. UK Environment Agency. River-flow gauging data, flood-risk maps, water-quality monitoring, coastal-erosion records. Weakness: UK-only; data quality varies by gauging station.

2. UK ONS (Office for National Statistics). UK Census 2021 — population, housing, ethnicity, deprivation, transport, by ward / output area. Weakness: Census is decadal — data ages quickly; aggregated to unit level (output area) so may hide within-unit variation.

3. USGS (US Geological Survey). Global earthquake catalogue, magnitudes, locations, depths; ShakeMaps for major events. Weakness: good for technical data but does not capture humanitarian impacts; earthquake catalogue completeness varies by region (better in monitored areas).

4. Met Office. UK weather records (rainfall, temperature, wind, storm tracks) + global tropical-cyclone data. Weakness: weather stations have spatial gaps; long-term records exist but are not equally spread.

Other commonly-cited sources: OS (maps), BGS (geology), UNDP (HDI), World Bank (development indicators), UN OCHA ReliefWeb (humanitarian disasters), WHO (health), FAO (agriculture), NOAA (climate + hurricanes), Google Earth (visual change), Channel Coastal Observatory (UK coastal data).
Why this scores
Mark scheme: 1 mark per source named + content + weakness (4 needed). Naming with the FULL agency name is what scores — 'a government website' = 0 marks.
Question 4
4 marks
[MEDIUM] Compare the strengths + weaknesses of PRIMARY and SECONDARY data for a fieldwork enquiry. [4]
Model answer
Primary data strengths: directly relevant to YOUR hypothesis + YOUR sites + YOUR variables; you know exactly how it was collected (no hidden methodology); you OWN the dataset; can target specific questions secondary data does not answer.

Primary data weaknesses: time-consuming + expensive in effort; limited to one field day + few sites; affected by your weather + observer skill; small sample size; cannot capture long-term trends.

Secondary data strengths: MASSIVE scale (national, global); LONG time periods (decades, centuries — historic OS maps go back to 1850s); collected by experts with expensive instruments students do not have (calibrated river gauges, satellites); cheap or free (most government data is open).

Secondary data weaknesses: not specific to your hypothesis or site; AGGREGATED at unit level that may hide variation (e.g. census ward when you need street-level); collection method may be unknown OR biased; may be DATED (Census is decadal); methodology may have changed between time periods making comparisons hard.

Best practice — triangulation. A reliable enquiry uses BOTH:

Primary for the specific finding ('beach width at Mappleton on 12 May 2026 = 38 m').

Secondary for the wider context ('UK Environment Agency historic data shows this is typical of post-1991 protected coast').

The two TOGETHER are stronger than either alone. Pearson 4GE1 mark schemes credit candidates who combine both with named sources.
Why this scores
Mark scheme: 1 mark for primary strength + weakness; 1 mark for secondary strength + weakness; 1 mark for explicit triangulation point; 1 mark for naming at least one specific source (e.g. Environment Agency, ONS, USGS).
Question 5
4GE1 Paper 2 Section B style (AO3/AO4)6 marks
[HARD] A student investigates 'river flooding on the River Severn (UK)'. Explain how they would COMBINE primary + secondary data sources, justifying each choice. [6]
Model answer
Primary data (collected on the field day).

River width, depth, velocity at 5 sites along ~5 km of the Severn. Velocity by float method × 0.85 → mean velocity → discharge.

Photographs of channel + recent flood debris (high-water marks on trees, on buildings).

Annotated sketches of cross-sections at each site.

Pebble survey for sediment size.

Justification: primary data captures the river's state on YOUR day at YOUR sites — direct measurement of the variable you're testing. Photographs of debris lines are powerful direct evidence of recent flood extent.

Secondary data sources.

1. UK Environment Agency (river-flow gauging data). Continuous discharge records (m³/s) for named Severn gauging stations (e.g. Bewdley, Worcester) going back decades. JUSTIFY: lets you compare your field-day discharge to long-term mean + recent extremes. Reveals whether your day was 'high', 'low' or 'normal' flow. Environment Agency Flood Map shows the modelled flood extent for 1-in-100-year + 1-in-1000-year events — crucial context.

2. Met Office (rainfall records). Antecedent rainfall — how much rain fell in the catchment in the preceding week / month / season. JUSTIFY: Severn floods correlate strongly with antecedent Welsh rainfall; without this context, you cannot explain the river's state.

3. Ordnance Survey maps + historic OS maps (1850-present). Show channel position + floodplain land use. Historic maps reveal how the floodplain has been built over (housing, roads, businesses) — explains why flood IMPACTS have grown even when discharge has not.

4. News + photographic archives. Photos + reports of historic Severn floods (e.g. 2000, 2007, 2020, 2024). Provides VISUAL EVIDENCE of past flood extents + IMPACTS (evacuated towns, flooded roads). Reveals the human dimension that gauging data misses.

5. BGS + soil maps. Underlying geology + soils control infiltration vs runoff in the catchment. Justifies why some sub-catchments respond faster than others.

6. Insurance / EA flood-damage reports. Document economic + social impacts of historic events — useful for the 'impacts' section of any flooding enquiry.

Triangulation.

When primary (your discharge measurement) + Environment Agency long-term data + Met Office rainfall + OS land use + photographic record ALL TELL THE SAME STORY — for example, 'your field day caught a relatively low-flow day in a catchment that has built over its floodplain such that 1-in-100-year flooding now affects 15,000 homes' — your conclusion is multi-source robust. When sources disagree, that's also informative — it reveals where to investigate further.

Why this matters. A primary-data-only enquiry on a Severn field day is a SNAPSHOT. A primary + secondary triangulated enquiry is a defensible study of the SYSTEM. Pearson 4GE1 mark schemes credit students who NAME specific sources + justify each + integrate them in the conclusion.
Why this scores
Mark scheme: 1 mark for primary methods named; 1 mark for Environment Agency with discharge / flood-risk context; 1 mark for Met Office antecedent rainfall; 1 mark for OS / historic OS / floodplain land use; 1 mark for triangulation point; 1 mark for one additional source (BGS, news archives, insurance) OR for justification quality.
Question 6
6 marks
[HARD] Evaluate the RELIABILITY of secondary data sources used in geographical enquiries. Reference at least TWO specific source types. [6]
Model answer
Source 1 — UK ONS Census.

Strengths: collected to a strict methodology + audited; coverage is near-total; data is comparable across England, Wales + Northern Ireland; available at fine spatial scales (output area).

Weaknesses: collected every 10 years — so by year 5+ post-Census, data is increasingly out of date; some questions are voluntary so self-selection bias can creep in (e.g. religion, sexual orientation); aggregation to output area level can MASK within-unit variation; some populations (homeless, undocumented migrants) are under-counted.

Reliability verdict: HIGH for what it measures, but USE WITH CAUTION for recent change and for under-represented populations.

Source 2 — Newspaper reports + news media.

Strengths: capture human impacts + immediate context that technical sources miss; widely accessible + free; provide photographic evidence; quote eyewitnesses.

Weaknesses: dramatic + emotive narrative bias — disasters over-reported, recovery under-reported; early figures often WRONG (e.g. Haiti 2010 earthquake death tolls varied from ~100k early reports to ~316k final Haitian government figure — a 3× range); national perspective biases (Western papers under-report non-Western experiences); inaccurate technical detail (magnitudes, distances) common; can repeat claims uncritically.

Reliability verdict: USEFUL for qualitative + human dimension, but technical claims must be cross-checked against scientific sources.

Source 3 — USGS earthquake catalogue.

Strengths: scientifically rigorous; consistent methodology; global coverage; events catalogued within hours; methodology + uncertainty estimates published.

Weaknesses: catalogue completeness varies by region (better in monitored areas like Japan + California than in remote regions); magnitudes for historic events less certain; does NOT capture humanitarian impacts.

Reliability verdict: HIGH for technical data; complement with UN OCHA / WHO for humanitarian impacts.

General principles for evaluating secondary-source reliability.

WHO collected it? Government agencies + UN bodies + peer-reviewed academic sources are most reliable; commercial sources + advocacy groups carry more bias.

WHEN was it collected? Recent data is generally better, but historic data is unique + irreplaceable.

HOW was it collected? Transparent + replicable methodology = more reliable.

WHAT IS THE SAMPLE? Census = ~total; sample surveys depend on sampling design.

WHO FUNDED IT? Funding source can shape design + reporting.

DOES IT TRIANGULATE WITH OTHER SOURCES? Multiple-source agreement = increased reliability.

The mature geographer treats secondary data as PROVISIONAL + CROSS-CHECKED — not as gospel. Triangulating multiple sources is the only reliable way to draw a defensible conclusion. Pearson 4GE1 mark schemes credit students who explicitly evaluate source reliability.
Why this scores
Mark scheme: 1 mark per source named with strengths + weaknesses (2 needed); 1 mark for explicit reliability verdict per source; 1 mark for general evaluation principles (who, when, how, sample, funding); 1 mark for triangulation point. The Haiti 100k-316k death-toll example is the killer illustration of narrative bias.
Question 7
4GE1 Paper 2 Section B style (AO3/AO4)8 marks
[CHALLENGE] Design a SECONDARY-DATA package for an enquiry into 'why the global development gap persists between MEDCs + LEDCs'. Justify the choice of EACH source + explain how they would be combined. [8]
Model answer
Package of secondary sources for global-development enquiry.

1. World Bank Open Data. GNI per capita, GDP growth, life expectancy, school enrolment, internet penetration by country + year (1960-present).

Justification: largest free open-source development dataset. Comparable across 200+ countries over decades. Allows direct testing of the development-gap hypothesis with quantitative measures.

Limitations: focuses on economic + social indicators; aggregated to national level (hides within-country inequality, especially urban-rural); some self-reported data may be inflated by national governments.

2. UNDP Human Development Report + HDI. HDI = composite index of income (GNI per capita) + education (mean + expected schooling) + health (life expectancy). Annual report; country rankings.

Justification: HDI is the single most-cited development metric. Multi-dimensional — captures more than just income. Allows direct comparison of MEDCs + LEDCs on a SINGLE score. Recent reports include inequality-adjusted HDI + gender development index.

Limitations: composite — masks the differences between income / health / education; doesn't capture political freedom, environmental sustainability, happiness.

3. OECD ODA (Official Development Assistance) data. Who gives + receives + how much aid; by donor country + recipient + sector + year.

Justification: development is shaped by aid flows + their patterns. Reveals donor priorities + recipient dependencies. Useful for testing whether aid reaches the poorest or follows political alliances.

Limitations: ODA is one channel — private + remittance + FDI flows often dwarf it; aid 'tied' to donor country exports underestimates real benefit.

4. UN Population Division (UNDESA). Population pyramids, fertility, urbanisation, dependency ratio by country + year. UN World Population Prospects.

Justification: demographic structure is foundational to development. The 'demographic dividend' (large working-age population) drives many emerging economies (China, India, Vietnam); aging populations face different challenges (Japan, much of Europe).

Limitations: Census quality varies by country; projections are uncertain.

5. WHO + UNICEF data. Health indicators — infant + maternal mortality, malnutrition, immunisation rates, access to safe water + sanitation.

Justification: development isn't just GDP — it's whether children survive + are nourished + receive school. SDG-aligned. Globally comparable.

Limitations: data quality is highly variable in low-income countries (where it matters most); often based on Demographic + Health Surveys with sampling limitations.

6. FAO (Food + Agriculture Organisation). Food production + nutrition + agricultural employment + food security.

Justification: developing-economy fundamentals — most LEDCs are agricultural; nutritional + production trends reveal whether development is reaching rural areas.

Limitations: subsistence farming often missed by formal statistics.

7. National statistical bureaux + regional bodies (ASEAN, AU, EU, LAC). National-level data — may differ from international estimates.

Justification: Triangulating national + international data reveals potential bias. UN estimates for China + India have differed from national figures in important ways.

Limitations: politically sensitive data sometimes suppressed; methodologies differ.

8. Academic literature (peer-reviewed development studies). Sen, Sachs, Easterly, Banerjee + Duflo etc. — empirical studies + theoretical framing.

Justification: Provides theoretical lens (e.g. capability approach, poverty traps, RCT evidence) that no raw dataset alone provides.

Limitations: access can be restricted; literature is large + contested.

9. Specific case-study sources. For each country / region of interest, identify the BEST single source. For example:

China — National Bureau of Statistics + World Bank China Quarterly.

India — Ministry of Statistics + UNDP India HDR.

Sub-Saharan Africa — African Union + African Development Bank reports.

Latin America — ECLAC, IDB.

How to combine the sources.

Map the gap quantitatively using World Bank + UNDP HDR data → produce comparison charts of MEDC vs LEDC indicators over time.

Explain the gap mechanisms using academic literature + WHO / UNICEF / FAO data → why does the gap persist?

Test specific hypotheses with the right sub-source: aid flows (OECD ODA), demographics (UN Population Division), health (WHO), agriculture (FAO).

Triangulate national vs international — does the World Bank story for India match India's own data?

Recognise data limitations — most reliable data comes from high-capacity states; the very poorest are under-represented (the 'statistical blackout' problem in fragile states).

Why this matters. Development is a multi-dimensional + politically contested topic. No single source captures it. The strongest enquiry uses the right source for the right question + acknowledges that quantitative indicators are PROXIES for human well-being, not the thing itself. Pearson 4GE1 mark schemes credit students who NAME the right specialist sources + integrate them critically.
Why this scores
Mark scheme: 1 mark per source named with justification + limitation (6+ sources needed for 6 marks); 1 mark for explicit combination / integration strategy; 1 mark for higher-level reflection (proxies for well-being; data quality varies by country; political contestation). Top-band 8-mark answer integrates the package critically rather than listing.
Question 8
4GE1 Paper 2 Section B style (AO3 evaluation)10 marks
[EXTENDED] 'The DATA SOURCES we choose SHAPE the conclusions we reach.' Discuss this statement with reference to geographical enquiries. [10]
Model answer
Thesis. The statement is BROADLY CORRECT + cuts to the heart of methodological transparency. Geographical enquiries are not neutral pictures of the world — they are PICTURES TAKEN BY SPECIFIC CAMERAS at specific angles in specific moments. The choice of data sources (primary vs secondary, which agency, which time period, what scale) shapes what the enquirer can see + therefore what conclusion can be reached. The mature geographer recognises this + makes the choice + its limits TRANSPARENT.

Five ways data-source choice shapes conclusions.

(1) SCALE shapes what patterns are visible. UK Census data aggregated to ward level reveals neighbourhood-scale deprivation patterns invisible at county level. Global development data at country level reveals between-country patterns but HIDES within-country inequality (which is often larger). World Bank country GNI per capita shows India + USA at different stages; an India-only enquiry using state-level data shows Kerala + Bihar are in different geographical worlds. The scale chosen DETERMINES which 'patterns' are visible.

(2) TIME shapes what trends + changes are visible. River-flow data over the last 5 years tells one story; over 50 years tells another (climate-change signal); over 200 years tells a third (Little Ice Age legacy). Census data is decadal; using 2011 data in 2026 misses 15 years of change. Newspaper archives back 50 years give one perspective; satellite imagery back 50 years gives another. Choosing the time range shapes what 'change' looks like.

(3) SOURCE PROVENANCE shapes what perspectives are heard. Western news sources for the Haiti 2010 earthquake gave one view; Haitian sources gave another; UN OCHA + USGS another. Each is partial + biased. Drawing a conclusion from Western sources only inevitably reproduces Western framings. Drawing from multiple-perspective sources gives a fuller picture. The same applies to development discourse — academic literature from MEDC institutions vs LEDC research institutions often differs in framing.

(4) METHODOLOGY shapes what is countable + therefore what counts. GDP per capita measures formal economic activity; informal sectors (subsistence agriculture, household labour, care work) are largely invisible. The HDI captures income + health + education; environment + political freedom + happiness are not directly counted. Choosing 'GDP' as the development measure produces one ranking; choosing 'HDI' produces another; 'GNH' (Gross National Happiness) produces a third. The choice of measure shapes who is 'developed'.

(5) WHICH DATA ARE COLLECTED + which are not. This is the most fundamental shaping. In rich countries with high state capacity, almost everything is measured; in fragile + low-income states, basic statistics may be unavailable. The very poorest are systematically UNDER-MEASURED. Climate change in the Pacific Small Island Developing States is documented less than in MEDCs. The 'global development gap' analysis can therefore systematically under-represent the most-affected places — the data simply is not there.

Examples from 4GE1 specification.

(a) Haiti 2010 earthquake. Newspaper sources gave early death tolls of ~100,000. Final Haitian government figure was ~316,000 — a 3× difference. A student using Western newspaper archives reaches a different conclusion than one using Haitian government final reports + UN OCHA data. Both can be honest enquiries; the source choice shapes the result.

(b) Urban-fieldwork pedestrian counts. A student counting pedestrians at 5 sites for 10 minutes each in May reaches one conclusion. Using ONS Census 2021 footfall data, the same area shows different patterns. Combining shows that the primary data is a snapshot of the secondary trend.

(c) Holderness coast. Primary beach width measurements on one day show the spatial pattern; UK Environment Agency historic records show that pattern is consistent with 30 years of post-1991 engineering. Without the secondary data, the primary measurement is a snapshot without context.

(d) Global development. World Bank ranking using GNI puts the USA + Germany + UK at the top. Using HDI, the rankings change slightly (Norway, Switzerland top). Using a 'happiness index', they change radically (Bhutan, Costa Rica do well). The 'development gap' looks different depending on the chosen measure.

Counter-argument — does data choice DETERMINE the conclusion, or just COLOUR it?

A naive reading of the statement might suggest that the data SHAPES the conclusion so much that 'truth' becomes impossible. This goes too far. There ARE underlying realities. The Holderness coast IS eroding regardless of which data we use to study it. India IS lower-income than the USA in any reasonable measure. The questions are: (a) which aspects of the reality are captured + which are missed; (b) what BIASES does the source carry; (c) what is the SOURCE-INDEPENDENT consensus where multiple-source triangulation agrees?

The skilled geographer's response — five disciplines.

CHOOSE deliberately. Match the source to the question. For UK rivers → Environment Agency. For US earthquakes → USGS. For global development → World Bank + UNDP + WHO.

TRIANGULATE. Use multiple sources whose biases differ. When they agree, conclusion is robust. When they disagree, investigate why — this is often the most interesting finding.

NAME sources. Pearson 4GE1 mark schemes specifically credit students who NAME (Environment Agency, ONS, USGS, World Bank) rather than vague 'online sources'. Naming forces source-reliability evaluation.

ACKNOWLEDGE limitations. State explicitly what your source captures + what it misses. 'World Bank GNI data captures formal economy; informal sector + household labour is not measured.' Honest acknowledgement EARNS marks; over-claiming loses them.

REVISIT. When sources change (new Census, updated UN report, recent academic literature), revisit conclusions. Geography is not static; data improves; reconsider.

Wider epistemological lesson.

This insight is not unique to geography. Medicine wrestles with which trial endpoint to measure (mortality vs symptom relief vs quality-adjusted life-years); economics with which inflation index to use; climate science with which baseline temperature to compare to. Every empirical discipline shapes its conclusions through measurement choices. The IPCC reports include explicit confidence levels precisely because the underlying climate data + models have known limitations. Honest geography follows the same epistemological discipline.

Judgement.

The statement is correct + important. Data sources DO shape conclusions — through scale, time, provenance, methodology + what is measured. This is not a problem to be solved (it cannot be); it is a CONDITION OF DOING GEOGRAPHY that must be acknowledged. The mature response is not to despair of objectivity, nor to over-claim it, but to (a) choose sources deliberately, (b) triangulate, (c) name + acknowledge, and (d) revisit as data improves. Pearson 4GE1 mark schemes consistently reward students who treat data sources critically — this is the methodological maturity that lifts a good enquiry to A*. The strongest enquiries are not those that use the 'best' source but those that use MULTIPLE sources transparently + honestly + comparatively. Truth-seeking geography is triangulated geography.
Why this scores
Mark scheme: 2 marks for clear thesis + framing; 2 marks for the five ways source choice shapes conclusions (scale, time, provenance, methodology, what is collected) with examples; 2 marks for 4GE1-specific examples (Haiti, Holderness, urban fieldwork, development); 1 mark for considering the counter-argument; 2 marks for the response disciplines (choose, triangulate, name, acknowledge, revisit); 1 mark for the wider epistemological framing + judgement. Top-band 10-mark answer connects fieldwork practice to scientific epistemology + uses real specification examples.

Key Definitions and Keywords — Data Sources

Definitions to memorise and the exact keywords mark schemes credit for data sources answers — sharpened from recent examiner reports for the 2026 Cambridge IGCSE sitting.

Primary data
Examiner keyword
Data the researcher has collected themselves through original fieldwork (measurements, surveys, interviews).
Secondary data
Examiner keyword
Data collected by someone else (government agencies, university researchers, companies, NGOs) and made available for re-use.
Triangulation
Examiner keyword
Using multiple data sources (primary + secondary, quantitative + qualitative, multiple agencies) to investigate the same question — agreement = robust, disagreement = informative.
Bias
Examiner keyword
A systematic distortion in how data was collected or what it captures — making the data point reliably in one direction.
ONS (Office for National Statistics)
The UK government body that runs the UK Census + publishes population, deprivation, transport + housing data.
UK Environment Agency
The UK government body responsible for river-flow monitoring, flood-risk maps, coastal-erosion monitoring + water quality.
Met Office
The UK national weather service — rainfall, wind, storm, climate records + global tropical-cyclone data.
USGS (US Geological Survey)
The US federal body publishing global earthquake catalogues, ShakeMaps, volcano + landslide data.
World Bank Open Data
Free open dataset of global development indicators (GNI per capita, life expectancy, education, internet) by country + year, 1960-present.
UNDP Human Development Report (HDI)
UN annual report + composite index of income + education + health — single most-cited development metric.
BGS (British Geological Survey)
UK government body publishing geology + soil maps + coastal monitoring data.
UN OCHA ReliefWeb
UN Office for the Coordination of Humanitarian Affairs — disaster casualty + displacement + response data.
Ordnance Survey (OS) maps
UK national mapping agency — topographic maps + historic OS maps (1850s-present) for tracking land-use change.
GIS (Geographic Information System)
Computer software for layering + analysing spatial data (e.g. Google Earth, ArcGIS, QGIS).

Common Mistakes and Misconceptions — Data Sources

The traps other students keep falling into on data sources questions — taken from recent Cambridge IGCSE examiner reports and mark schemes — and how to avoid them.

✕Naming 'the internet' or 'a government website' instead of a specific agency
Pearson 4GE1 Examiner Reports — Paper 2 Section B
Why it happens
Students don't know the specific bodies.
How to avoid it
Memorise the key sources: UK rivers → Environment Agency; UK Census → ONS; earthquakes → USGS; development → World Bank + UNDP; weather → Met Office. Naming is what earns marks.
✕Treating primary fieldwork as a complete enquiry, without secondary data
Pearson 4GE1 Examiner Reports — Paper 2 Section B
Why it happens
Focus on fieldwork; secondary feels like 'cheating'.
How to avoid it
Pearson 4GE1 spec EXPLICITLY requires BOTH. Always cite at least 2-3 secondary sources alongside primary fieldwork. Triangulation is the strongest method.
✕Using secondary data without evaluating its reliability
Why it happens
Government data feels authoritative.
How to avoid it
Every secondary source has strengths AND weaknesses. State both. 'ONS Census 2021 is high-quality but decadal — 2026 data is 5 years out of date.'
✕Using the wrong source for the question (e.g. UK Environment Agency for global development)
Why it happens
Students apply familiar sources rather than the right source.
How to avoid it
Match source to question: UK rivers → EA; global development → World Bank + UNDP; UK Census → ONS; earthquakes → USGS. Use the SPECIALIST for each topic.
✕Treating newspaper reports as factual + technical sources
Pearson 4GE1 Examiner Reports — Paper 2 Section B
Why it happens
Newspapers feel like they're 'reporting the news'.
How to avoid it
Newspapers carry narrative bias + early-figure errors (Haiti 2010 death tolls: ~100k early vs ~316k final — a 3× range). Use newspapers for QUALITATIVE context; verify TECHNICAL claims against scientific sources (USGS, UN OCHA).
✕Relying on a single data source for the conclusion
Pearson 4GE1 Examiner Reports — Paper 2 Section B
Why it happens
Easier; familiar.
How to avoid it
Always TRIANGULATE: combine 1-2 primary + 2-3 secondary sources. When all agree, conclusion is robust. When they disagree, that's also informative.

Data Sources (Primary + Secondary) — Pearson Edexcel International GCSE Geography (4GE1) Study Notes

Source

What it provides

UK Environment Agency

River-flow gauging data (m³/s) at named stations going back decades; flood-risk maps; Catchment Data Explorer; water-quality monitoring

Met Office

Antecedent rainfall + storm records — vital for explaining river flow + flood events

OS topographic + historic maps

Channel position + flood-plain land-use change (1850-present)

BGS geology

Underlying geology + soils control infiltration vs runoff

Source

What it provides

UK ONS (Census 2021)

Population, housing, ethnicity, deprivation, transport, by output area + ward + district

Council planning portals

Development applications + decisions; rezoning

UK Index of Multiple Deprivation (IMD)

Multi-dimensional deprivation by lower-layer super output area

Department for Transport

Traffic data; public-transport patronage

OS + Google Earth + Street View

Building density + land-use mapping; visual comparison over time

Source

What it provides

UK Environment Agency

Coastal-erosion records; Shoreline Management Plans

Met Office

Wind, wave, storm data; tide forecasts

Channel Coastal Observatory

Beach surveys, photographs, aerial surveys — UK south coast specialist

OS + historic OS

Coastline change over decades

Source

What it provides

USGS (US Geological Survey)

Global earthquake catalogue; ShakeMaps; volcano monitoring

NOAA National Hurricane Center

Tropical-cyclone tracks, intensity, forecasts

UN OCHA ReliefWeb

Humanitarian impact data — deaths, displaced, response, funding

WHO

Health impacts (disease outbreaks, injury, mental health)

Insurance + reinsurance reports (e.g. Munich Re)

Economic damage

Source

What it provides

World Bank Open Data

GNI per capita, GDP, education, health, internet by country + year (1960-present)

UNDP Human Development Report

HDI rankings — composite of income + education + health

OECD ODA data

Official development assistance flows by donor + recipient + sector

UN Population Division (UNDESA)

Population pyramids, fertility, urbanisation by country + year

WHO + UNICEF

Health indicators (infant + maternal mortality, malnutrition, immunisation)

FAO

Food + agriculture data

Thesis. The statement is BROADLY CORRECT + cuts to the heart of methodological transparency. Geographical enquiries are not neutral pictures of the world — they are PICTURES TAKEN BY SPECIFIC CAMERAS at specific angles in specific moments. The choice of data sources (primary vs secondary, which agency, which time period, what scale) shapes what the enquirer can see + therefore what conclusion can be reached. The mature geographer recognises this + makes the choice + its limits TRANSPARENT.

Five ways data-source choice shapes conclusions.

(1) SCALE shapes what patterns are visible. UK Census data aggregated to ward level reveals neighbourhood-scale deprivation patterns invisible at county level. Global development data at country level reveals between-country patterns but HIDES within-country inequality (which is often larger). World Bank country GNI per capita shows India + USA at different stages; an India-only enquiry using state-level data shows Kerala + Bihar are in different geographical worlds. The scale chosen DETERMINES which 'patterns' are visible.

(2) TIME shapes what trends + changes are visible. River-flow data over the last 5 years tells one story; over 50 years tells another (climate-change signal); over 200 years tells a third (Little Ice Age legacy). Census data is decadal; using 2011 data in 2026 misses 15 years of change. Newspaper archives back 50 years give one perspective; satellite imagery back 50 years gives another. Choosing the time range shapes what 'change' looks like.

(3) SOURCE PROVENANCE shapes what perspectives are heard. Western news sources for the Haiti 2010 earthquake gave one view; Haitian sources gave another; UN OCHA + USGS another. Each is partial + biased. Drawing a conclusion from Western sources only inevitably reproduces Western framings. Drawing from multiple-perspective sources gives a fuller picture. The same applies to development discourse — academic literature from MEDC institutions vs LEDC research institutions often differs in framing.

(4) METHODOLOGY shapes what is countable + therefore what counts. GDP per capita measures formal economic activity; informal sectors (subsistence agriculture, household labour, care work) are largely invisible. The HDI captures income + health + education; environment + political freedom + happiness are not directly counted. Choosing 'GDP' as the development measure produces one ranking; choosing 'HDI' produces another; 'GNH' (Gross National Happiness) produces a third. The choice of measure shapes who is 'developed'.

(5) WHICH DATA ARE COLLECTED + which are not. This is the most fundamental shaping. In rich countries with high state capacity, almost everything is measured; in fragile + low-income states, basic statistics may be unavailable. The very poorest are systematically UNDER-MEASURED. Climate change in the Pacific Small Island Developing States is documented less than in MEDCs. The 'global development gap' analysis can therefore systematically under-represent the most-affected places — the data simply is not there.

Examples from 4GE1 specification.

(a) Haiti 2010 earthquake. Newspaper sources gave early death tolls of ~100,000. Final Haitian government figure was ~316,000 — a 3× difference. A student using Western newspaper archives reaches a different conclusion than one using Haitian government final reports + UN OCHA data. Both can be honest enquiries; the source choice shapes the result.

(b) Urban-fieldwork pedestrian counts. A student counting pedestrians at 5 sites for 10 minutes each in May reaches one conclusion. Using ONS Census 2021 footfall data, the same area shows different patterns. Combining shows that the primary data is a snapshot of the secondary trend.

(c) Holderness coast. Primary beach width measurements on one day show the spatial pattern; UK Environment Agency historic records show that pattern is consistent with 30 years of post-1991 engineering. Without the secondary data, the primary measurement is a snapshot without context.

(d) Global development. World Bank ranking using GNI puts the USA + Germany + UK at the top. Using HDI, the rankings change slightly (Norway, Switzerland top). Using a 'happiness index', they change radically (Bhutan, Costa Rica do well). The 'development gap' looks different depending on the chosen measure.

Counter-argument — does data choice DETERMINE the conclusion, or just COLOUR it?

A naive reading of the statement might suggest that the data SHAPES the conclusion so much that 'truth' becomes impossible. This goes too far. There ARE underlying realities. The Holderness coast IS eroding regardless of which data we use to study it. India IS lower-income than the USA in any reasonable measure. The questions are: (a) which aspects of the reality are captured + which are missed; (b) what BIASES does the source carry; (c) what is the SOURCE-INDEPENDENT consensus where multiple-source triangulation agrees?

The skilled geographer's response — five disciplines.

CHOOSE deliberately. Match the source to the question. For UK rivers → Environment Agency. For US earthquakes → USGS. For global development → World Bank + UNDP + WHO.
TRIANGULATE. Use multiple sources whose biases differ. When they agree, conclusion is robust. When they disagree, investigate why — this is often the most interesting finding.
NAME sources. Pearson 4GE1 mark schemes specifically credit students who NAME (Environment Agency, ONS, USGS, World Bank) rather than vague 'online sources'. Naming forces source-reliability evaluation.
ACKNOWLEDGE limitations. State explicitly what your source captures + what it misses. 'World Bank GNI data captures formal economy; informal sector + household labour is not measured.' Honest acknowledgement EARNS marks; over-claiming loses them.
REVISIT. When sources change (new Census, updated UN report, recent academic literature), revisit conclusions. Geography is not static; data improves; reconsider.

Wider epistemological lesson.

This insight is not unique to geography. Medicine wrestles with which trial endpoint to measure (mortality vs symptom relief vs quality-adjusted life-years); economics with which inflation index to use; climate science with which baseline temperature to compare to. Every empirical discipline shapes its conclusions through measurement choices. The IPCC reports include explicit confidence levels precisely because the underlying climate data + models have known limitations. Honest geography follows the same epistemological discipline.

Judgement.

The statement is correct + important. Data sources DO shape conclusions — through scale, time, provenance, methodology + what is measured. This is not a problem to be solved (it cannot be); it is a CONDITION OF DOING GEOGRAPHY that must be acknowledged. The mature response is not to despair of objectivity, nor to over-claim it, but to (a) choose sources deliberately, (b) triangulate, (c) name + acknowledge, and (d) revisit as data improves. Pearson 4GE1 mark schemes consistently reward students who treat data sources critically — this is the methodological maturity that lifts a good enquiry to A*. The strongest enquiries are not those that use the 'best' source but those that use MULTIPLE sources transparently + honestly + comparatively. Truth-seeking geography is triangulated geography.

Data Sources

Detailed Study Notes

At a glance

What you’ll learn

Quick recap

Memorise this

How it’s examined

Take this whole topic with you

1Primary vs secondary

2Naming specific secondary sources

3Pros + cons of primary + secondary

4Triangulating primary + secondary

5Evaluating bias in secondary sources

6Designing a data-sources package

Primary data

Secondary data

Triangulation

Bias

ONS (Office for National Statistics)

UK Environment Agency

Met Office

USGS (US Geological Survey)

World Bank Open Data

UNDP Human Development Report (HDI)

BGS (British Geological Survey)

UN OCHA ReliefWeb

Ordnance Survey (OS) maps

GIS (Geographic Information System)

✕Naming 'the internet' or 'a government website' instead of a specific agency

✕Treating primary fieldwork as a complete enquiry, without secondary data

✕Using secondary data without evaluating its reliability

✕Using the wrong source for the question (e.g. UK Environment Agency for global development)

✕Treating newspaper reports as factual + technical sources

✕Relying on a single data source for the conclusion

Data Sources

Detailed Study Notes

At a glance

What you’ll learn

Quick recap

Memorise this

How it’s examined

Take this whole topic with you

1Primary vs secondary

2Naming specific secondary sources

3Pros + cons of primary + secondary

4Triangulating primary + secondary

5Evaluating bias in secondary sources

6Designing a data-sources package

Primary data

Secondary data

Triangulation

Bias

ONS (Office for National Statistics)

UK Environment Agency

Met Office

USGS (US Geological Survey)

World Bank Open Data

UNDP Human Development Report (HDI)

BGS (British Geological Survey)

UN OCHA ReliefWeb

Ordnance Survey (OS) maps

GIS (Geographic Information System)