G2 Geography Paper 1 Question 1 in the 2027 Singapore-Cambridge Secondary Education Certificate (SEC) is a compulsory 25-mark fieldwork question on Topic 1.3 Geographical Methods. The fieldwork context may or may not be drawn from one of the syllabus’s three main clusters, so students cannot prepare by memorising only one neighbourhood study or tourism location. They need to understand how to design an inquiry, collect data, identify patterns, present findings and evaluate the reliability and validity of conclusions.
This guide explains the question’s place in K226 standalone G2 Geography, how the school curriculum’s 10-week extended fieldwork relates to examination preparation, and how to handle unfamiliar research designs and datasets. It includes original practice exercises involving surveys, mental maps, sampling, percentages, scatterplots and evaluation. All example data are fictional educational material, not measurements from actual Singapore neighbourhoods or predicted examination tasks.
- Subject: G2 Geography, 2027 SEC syllabus K226, former reference 2246.
- Paper 1: 1 hour 45 minutes, 50 marks, 50% of overall G2 Geography.
- Question 1: 25 marks, compulsory, based on Geographical Methods and fieldwork.
- Question 2: 25 marks, compulsory, on Tourism, including a six-mark AO3 judgement task.
- Question format: Each structured question has no more than twelve subparts.
- Fieldwork learning: A 10-week extended fieldwork experience forms part of the taught syllabus.
- Important distinction: The published external assessment is two written papers, not a separately specified coursework component with its own SEC weighting.
- Primary objective: Ask a geographical question, gather suitable evidence, analyse it honestly and state a defensible conclusion.
Why G2 Geography Fieldwork Is Not Just Going Outdoors
A trip to a park, shopping centre or waterfront can be enjoyable, but a geographical investigation needs a specific question. Students must identify what they want to know and what evidence could answer it.
For example, “Is this neighbourhood good?” is too broad because different people value different things and “good” has not been defined. “How does reported satisfaction with access to bus services vary between two specified parts of the neighbourhood?” is more focused.
Once the question is clear, students can decide who to ask, which responses to collect and how to compare them.
The outdoor experience matters because it exposes the gap between an assumption and what happens in a real place. The examination tests whether students understand the thinking and methods behind that investigation.
The Official 2027 K226 Assessment Scheme
Paper 1 is worth 50 marks and lasts 105 minutes. Question 1 addresses Topic 1.3 Geographical Methods, while Question 2 addresses Cluster 2 Tourism. Both are compulsory and worth 25 marks.
SEAB states explicitly that Question 1 is on fieldwork and its context may or may not come from any of the syllabus clusters.
The six-mark generic-level-descriptor judgement question appears in Question 2, not automatically in Question 1. Question 1 still requires scientific and geographical reasoning, but students should not mislabel every fieldwork subpart as the special six-mark AO3 task.
Paper 2 has a different structure: Question 1 addresses Thinking Geographically and Sustainable Development, and Question 2 addresses Climate. Students must not confuse the different question placements.
How the Ten-Week Fieldwork Experience Fits
The 2027 K226 curriculum states that students undertake extended fieldwork over ten weeks, applying classroom learning to investigate a content area in the syllabus.
This is a learning experience, not a separately listed external Paper 3 or coursework percentage in the K226 examination scheme. Students should follow their school’s actual project schedule, teacher supervision and approved research procedures.
The value of ten weeks is that a meaningful study takes time: choosing a research question, testing a method, collecting evidence, analysing findings and evaluating limitations rarely happen well in one afternoon.
An effective learner should be able to explain what was changed during the investigation and why the adjustment improved the evidence.
The Four Geographical Methods Questions in the Syllabus
Topic 1.3 organises its learning around four big questions:
- How to design fieldwork? Formulate research questions and measurable hypotheses, select an inquiry approach and consider limits and risks.
- How to collect primary data? Use appropriate sampling, closed-ended questionnaires, scales, mental maps and semi-structured interviews.
- How to process and analyse data? Interpret counts, percentages, mean, mode, median, mental-map information, scatterplots and spatial patterns.
- How to present findings? Communicate through maps, suitable graphs, photographs and relevant textual representations.
These four questions form a logical chain. A weak research question produces unclear data. Unclear data produce uncertain patterns. A badly chosen graph can distort those patterns. A conclusion is only as credible as the earlier decisions that produced it.
Step 1: Frame a Research Question
A geographical research question identifies a topic, scope, place, relevant variables and relationship worth investigating.
Consider an original fictional school inquiry into how residents use two sheltered community spaces.
A broad question such as “Which place is better?” leaves too many meanings open. A clearer question could be: “Do surveyed users report different levels of comfort in Space A and Space B during comparable afternoon periods?”
The student now knows what will be compared, which people might be asked, what the main variable is and when observations should occur.
A question should be answerable with the available time, resources and ethical permissions. A very ambitious research design that cannot be carried out is not automatically stronger.
Step 2: Formulate a Testable Hypothesis
A hypothesis is a testable proposed relationship, not a fact the student has already proved.
For the fictional spaces, a hypothesis might be: “A higher proportion of respondents in Space A will report being comfortable than in Space B during comparable afternoons.”
The student should be prepared for evidence that supports, weakens or contradicts the hypothesis. It is unacceptable to alter responses or omit inconvenient results simply to make a predicted relationship appear correct.
The official K226 syllabus refers to a measurable hypothesis involving one or two variables. Students should practise keeping their hypothesis narrowly connected to the actual data.
A hypothesis does not have to be dramatic. Clarity is more useful than complicated wording.
Step 3: Decide Whether to Collect Quantitative or Qualitative Data First
The syllabus includes inquiry sequences that begin with quantitative data, then use qualitative methods to explore a pattern, and the reverse sequence in which qualitative observations are followed by quantitative checks.
Quantitative data are expressed through numbers, such as counts, ratings, distances or percentages. Qualitative data include descriptions, experiences and explanations that may not be represented adequately by a single number.
Suppose initial survey ratings suggest users are less satisfied in one space. A follow-up interview could explore what residents mean by “uncomfortable.”
Alternatively, interviews might reveal recurring concerns about seating, noise or access. A later structured questionnaire could then examine how widely those concerns appear among the sampled group.
The order should serve the question, not a belief that one data type is always superior.
Step 4: Understand Sampling Before Collecting Responses
Fieldwork surveys often collect information from some people, not every person who uses the area. That makes sampling important.
The K226 syllabus includes non-probability methods such as convenience and quota sampling, and probability approaches such as simple random and stratified random sampling.
A student should understand what makes these methods different, what each can achieve and where each may introduce limitations.
Convenience Sampling
This involves selecting people who are readily available. It can be practical, but the sample may not represent the wider population.
If students interview only people near a café entrance, they may miss users of quiet areas or people who pass through at other times.
The limitation is not merely “small sample”. It is that the selection method may favour particular types of users.
Quota Sampling
Quota sampling sets numbers for selected groups, then obtains responses to meet those targets. It can ensure that the sample includes specified categories, but the selection within each quota may still be non-random.
For example, the researcher might plan to speak with respondents from three age bands. Achieving the quotas does not automatically make the sample representative if interviewees are approached only at one convenient location.
The method should be evaluated in relation to the research question and available population information.
Simple Random Sampling
Under a suitable sampling frame, each eligible unit has a known, equal selection opportunity. Random selection can reduce certain researcher selection biases when correctly implemented.
It still requires an appropriate and complete frame. If the researcher has a list of only one building’s residents, choosing randomly from that list does not establish a representative sample of an entire town.
Randomness is not a magic remedy for a poorly defined population.
Stratified Random Sampling
This separates a population into meaningful subgroups before making random selections within those groups. It can help ensure representation of categories relevant to the inquiry.
Its quality depends on the definition of strata and the availability of appropriate information. Students should explain why the chosen categories matter, rather than merely use a sophisticated method name.
Worked Example 1: Two Spaces, Unequal Sample Sizes
Original fictional dataset: Twenty people in Space A answer a closed-ended comfort question; fifteen report feeling comfortable. In Space B, forty people answer, and twenty-four report feeling comfortable.
Space A’s reported comfort proportion is 15 ÷ 20 × 100 = 75%.
Space B’s reported comfort proportion is 24 ÷ 40 × 100 = 60%.
The absolute number of comfortable respondents is higher in B, but the percentage within the sample is higher in A.
This shows why comparing counts can be misleading when the sample sizes differ.
However, these percentages are based only on the fictional selected respondents. They do not establish that exactly 75% or 60% of all users of those spaces are comfortable.
Worked Example 2: What Would Make That Comparison Fairer?
Suppose the sample in A was collected on a quiet weekday morning while B was surveyed during a crowded weekend event. The settings differ in ways likely to affect reported comfort.
A better comparison might select similar time periods, use the same wording and response scale, and document other relevant conditions.
The proposed change should be linked to the weakness. Writing only “collect more data” does not explain why the two samples are currently hard to compare.
If the question concerns time-of-day differences, however, varying time deliberately may be the point. Fairness depends on the research question, not a rule that every condition must always remain identical.
Step 5: Write a Good Closed-Ended Questionnaire
A closed-ended question offers predefined response choices. It may ask for a number, category, rating or frequency.
The K226 syllabus refers to rating approaches including Likert scales, frequency scales and ranking scales.
A useful question might be: “How comfortable did you feel in this space during your visit?” with clear, balanced response choices. Students should avoid an emotionally loaded question such as “How excellent was the amazing new facility?”
A question should be understandable, neutral and related to the inquiry. The learner should also consider what a respondent may genuinely not know.
Why Leading Questions Weaken Evidence
Consider two fictional survey items.
Leading: “Do you agree that our comfortable, well-designed park is much better than the old one?”
More neutral: “How would you rate the comfort of the park during today’s visit?”
The first wording supplies a positive conclusion before the respondent answers. The second asks for an opinion with less pressure toward a particular judgement.
A neutral question does not guarantee perfect evidence. The sampling process, timing and interpretation can still introduce limitations.
The important skill is to recognise how the collection method could influence the answer.
Step 6: Use Mental Maps and Semi-Structured Interviews
A mental map helps represent how someone remembers, experiences or values a place. It is not necessarily a geographically exact map of the physical environment.
A respondent might mark a familiar meeting place prominently while omitting a nearby feature they rarely notice. That can reveal the person’s experience, not simply a cartographic mistake.
The K226 syllabus includes drawing features and labels on a study-area base map and using semi-structured interviews to explore them.
A student should distinguish what the map shows about a respondent’s perception from what has been independently measured about the physical area.
Worked Example 3: Interpreting an Imaginary Mental Map
Original fictional scenario: A respondent labels a bus stop, playground and food shop on a map, but omits a nearby footpath. During a follow-up interview, the respondent explains that the footpath feels poorly connected to their normal route.
One supported conclusion is that the respondent’s map emphasises places they use and regard as important. The omitted footpath may not form part of their habitual route.
It would be unsound to conclude that the footpath does not exist or that every resident avoids it.
A useful investigation might compare several mental maps and interview accounts, while recognising that each is shaped by individual experience.
This demonstrates a key difference between perceived space and measured physical geography.
Step 7: Process Counts, Percentages and Averages
The syllabus includes counts, percentages, mean, mode and median when analysing closed-ended survey responses.
The appropriate measure depends on the question and data type. A mean is meaningful for some numerical datasets. A mode can indicate the most common response. A median describes the middle value in an ordered distribution.
Not all responses should be averaged mechanically. A category such as “bus”, “train” or “walk” does not form a numerical scale merely because categories are coded 1, 2 and 3.
Students should check what each number represents before calculating.
Worked Example 4: Mean, Median and Mode
Original fictional dataset: Six respondents report spending 5, 5, 6, 6, 8 and 18 minutes travelling to a facility.
The mean is (5+5+6+6+8+18) ÷ 6 = 8 minutes.
The median is the mean of the two middle values, (6+6)/2 = 6 minutes.
The dataset has two modes, 5 and 6 minutes, since both occur twice.
The unusually large 18-minute value raises the mean above the median. If the question asks about a typical central value less influenced by the unusually long journey, the median may be more informative.
A strong interpretation identifies what the summary measure tells us and what the small dataset cannot establish about all users.
Step 8: Interpret Scatterplots and Correlations
The K226 syllabus explicitly includes positive and negative correlations, scatterplots and best-fit lines.
A positive relationship means that, in the observed data, higher values of one variable tend to accompany higher values of another. A negative relationship means that higher values tend to accompany lower values.
A scatterplot does not by itself establish that one variable causes the other.
For example, larger neighbourhood spaces might show more visitors, but location, accessibility and other factors may also influence attendance.
A meaningful explanation distinguishes observed association from proved causation.
Worked Example 5: An Invented Positive Relationship
Original fictional data: Four school clubs report the number of publicity messages sent and the number of event registrations.
| Messages sent | Registrations |
|---|---|
| 2 | 15 |
| 4 | 22 |
| 6 | 30 |
| 8 | 36 |
The numbers show a positive association across these four invented observations: more messages are associated with more registrations.
However, the data do not prove that messages alone caused the difference. Club size, appeal, timing and other conditions may vary.
A suitable conclusion would identify the pattern and propose what further comparison might help evaluate the relationship.
The data are for statistical practice, not actual school records or an advertising recommendation.
Worked Example 6: A Negative Relationship
Original fictional data: A small study shows travel times of 5, 10, 15 and 20 minutes, paired with hypothetical attendance frequencies of 8, 7, 5 and 3 visits over a stated period.
There is a negative association: as travel time increases, reported visits decrease in these observations.
That does not establish that travelling farther always causes fewer visits. Personal preferences, transport availability and available time could influence the result.
A student should describe the direction of the pattern, note the small dataset and explain which further evidence might be useful.
A best-fit line can help illustrate the overall trend, but it should not be forced through every point or used to claim certainty beyond the observed range.
Step 9: Recognise Spatial Patterns
Geographical data may reveal clusters, linear arrangements, regular spacing or other shapes.
For example, fictional facilities placed along a transport corridor may form a linear pattern. Shops concentrated around a junction may appear clustered.
Students should describe where the pattern occurs, what it consists of and what evidence establishes the shape.
If asked why a pattern exists, the learner must consider relevant geographical processes rather than assume the visual appearance alone proves one cause.
This is where maps connect to explanations of accessibility, planning and human movement.
Step 10: Present Data Through Appropriate Maps
The K226 syllabus includes representing spatial information with dots, lines and polygons, and maps containing relevant title, date, orientation, scale, legend, author and sources.
A good map communicates what is being shown. Symbols and colours should have clear meanings, and the scale should be appropriate to the area.
A map with no legend may be difficult to interpret even if its symbols are neatly drawn. A map with no scale may make distances hard to judge.
Students should understand how each mapping element helps another person interpret the evidence, not simply memorise a checklist of cartographic words.
Step 11: Choose the Right Graph
The syllabus includes bar charts and pie charts for distributions and line graphs for trends and relationships between variables.
A bar chart may compare categories or counts. A pie chart may show proportions of a whole, assuming the categories and totals are suitable. A line graph can show change over time or other ordered relationships.
Students should be able to explain why a particular graph fits the dataset.
A pie chart is not appropriate when the values do not represent meaningful parts of the same whole. A line graph may imply continuity between unrelated categories.
The graph is part of the reasoning, not merely a decoration beside the answer.
Step 12: Present Qualitative Findings Honestly
The syllabus allows photographs, satellite and aerial imagery, colour-coded quotations and word clouds as ways of presenting geographical information.
Qualitative evidence needs careful handling. A quotation can illustrate a participant’s experience without proving that every participant thinks the same way.
A word cloud may make frequent terms visually prominent, but frequency is not necessarily the same as importance. Students should explain how the source material was selected and coded.
Photographs may reveal features present at a location and time, but they cannot automatically establish what happened before the photograph was taken.
Good presentation shows evidence without exaggerating its meaning.
Reliability and Validity: The Most Important Evaluation Distinction
Reliability concerns the consistency or dependability of findings under relevant repetition or comparison. Validity concerns whether the method and evidence actually address the intended question.
A questionnaire may produce highly consistent results but measure the wrong thing. For example, asking users whether a park looks attractive does not necessarily measure their sense of safety while walking home.
A reliable instrument can still give a consistently biased result. A method may be valid for a narrow group while failing to represent a wider population.
Students should name the specific weakness and suggest an improvement linked to that weakness. Writing “make it more reliable and valid” is not enough.
Worked Example 7: A Reliable but Invalid Measure
Original fictional task: Students wish to investigate whether residents can find the entrance to a community facility easily. They repeatedly measure the building’s height and obtain consistent results.
The height measurements may be reliable, but they do not directly assess how easily residents locate the entrance.
A more relevant method might involve carefully designed observations of wayfinding or a neutral survey about whether participants experienced difficulty finding the entrance, conducted under approved school conditions.
This example shows why simply collecting numbers is not research. The numbers must answer the intended question.
Worked Example 8: A Valid Question With an Unrepresentative Sample
Original fictional task: A survey accurately asks about satisfaction with bus access, but interviews only students who are already waiting at the bus stop.
The question itself may be relevant, but the sample excludes people who avoid the stop or use other transport because of access difficulties.
A broader sampling approach may be needed to understand the intended population. The right adjustment depends on how that population was defined.
Students should distinguish question design from sample selection. Both affect the usefulness of the conclusion.
Step 13: Suggest an Improvement That Addresses the Real Problem
The phrase “use a bigger sample” is not a universal answer.
A larger sample may help address random variation, but it does not automatically remove selection bias if the researcher continues to interview only one convenient group.
An improvement should follow this structure: identified problem → specific method change → why it improves the inquiry.
For example, if observations are made at one unusual time, repeat them at several comparable periods to evaluate whether that time affected the pattern.
If participants misunderstand a survey question, pilot or revise its wording rather than simply collect more misunderstood responses.
Step 14: Plan Fieldwork Safety and Ethical Practice
The K226 syllabus includes recognising limitations and risks, adjusting aims to available resources and avoiding harm to oneself, others and nature.
Students should follow school supervision, permissions and safety arrangements. Do not enter dangerous or restricted areas, obstruct public paths, collect unnecessary personal information or disturb wildlife for a project.
A geographical inquiry should respect respondents’ privacy and their right not to participate where applicable. A student does not need to photograph identifiable people without suitable consent simply to show that a survey occurred.
Good research does not place the researcher or community at unnecessary risk.
Step 15: Interpret the Answer the Question Actually Asks
A fieldwork question might ask students to identify the research hypothesis, describe a data pattern, calculate a percentage, choose a graph, evaluate a limitation or propose an improvement.
Each command requires a different kind of answer. State may need a brief fact. Describe needs what the data show. Explain needs a mechanism. Evaluate needs a judgement based on evidence and limitations.
A student who writes several paragraphs describing a graph when asked to propose an improvement has not completed the task, however accurate the description may be.
Read the command word before choosing the response structure.
A Full Fictional Fieldwork Scenario
Original educational scenario: A student team wants to investigate whether people who report shorter walking times to a study space also report visiting it more frequently.
They design a survey asking respondents for an approximate walking time and weekly number of visits. The team approaches people at one entrance during a Friday afternoon.
They receive 30 responses. A scatterplot suggests that respondents with shorter walking times tend to report more visits, although several exceptions occur.
A good answer to “Describe the pattern” notes the negative association between walking time and frequency in this invented dataset.
A good answer to “Evaluate the conclusion” points out the limited sample, the choice of one time and entrance, and the distinction between association and causation.
A good answer to “Suggest an improvement” proposes surveying across more relevant locations or periods, explaining which limitation the change reduces.
Why the Same Scenario Can Test Many Skills
The fictional study contains a research question, two variables, a sampling method, a questionnaire, a scatterplot, an inference and an evaluation.
That is why fieldwork preparation should be integrated. Memorising a separate definition of “sample” does not automatically help when a long question asks how a sample affects the conclusion.
Students should practise seeing the investigation as a connected sequence. The research question determines the method; the method generates the data; the data support or weaken the conclusion.
This is a more powerful mental model than treating every fieldwork term as a disconnected flashcard.
Common Mistake 1: The Hypothesis Is Too Vague
“People like the neighbourhood” does not define what to measure or compare.
Repair by choosing a specific group, place, variable and measurable relationship.
Common Mistake 2: The Sample Does Not Represent the Intended Group
A study of all residents interviews only one convenient group.
Repair by examining the sampling frame and explaining a more appropriate way to reach the intended population.
Common Mistake 3: A Leading Question Shapes Responses
The wording suggests the answer before participants respond.
Repair by removing praise or blame from the question and using neutral, clear response choices.
Common Mistake 4: The Wrong Statistic Is Used
The mean is calculated on category codes with no meaningful numerical order.
Repair by matching the measure to the data type and the inquiry.
Common Mistake 5: Counts Are Compared Across Unequal Samples
A raw count is treated as a direct comparison of rates.
Repair by identifying denominators and calculating appropriate proportions where relevant.
Common Mistake 6: Correlation Is Presented as Proof of Cause
A scatterplot shows an association, but the explanation assumes all other factors are irrelevant.
Repair by distinguishing the observed pattern from the proposed mechanism and discussing alternative influences.
Common Mistake 7: The Map Is Missing Essential Information
An unlabeled scale or legend makes the map difficult to interpret.
Repair by checking the map elements needed for the specific spatial comparison.
Common Mistake 8: A Generic Improvement Does Not Fix the Weakness
“Use more respondents” may not solve a biased sample or invalid survey question.
Repair by identifying the error first and connecting the suggested change to it.
Common Mistake 9: Safety Is Treated as an Afterthought
The plan requires an unsafe location or unnecessary personal data.
Repair by adapting the study design and following the school’s supervised, ethical procedures.
Common Mistake 10: Fieldwork Becomes a Memorised Location Story
A student learns every detail of one neighbourhood walk but cannot apply methods to an unfamiliar context.
Repair by practising the same research method with different fictional settings and data.
The Fencing Method for G2 Geography Fieldwork
The eduKateSG Fencing Method begins with one clear research question and one variable relationship.
First, ask students to identify a measurable question and a suitable data source. Then introduce a sampling problem. Later, provide a small table and ask for a graph and conclusion. Finally, add a limitation and require an improvement.
Each step has a specific learning purpose. Students gain control before the task becomes complex.
Once the framework is stable, present an unfamiliar context without announcing which method should be used. The learner should select the appropriate approach independently.
A Four-Week Fieldwork Question 1 Practice Cycle
Week 1: Research Questions and Methods
Use original short scenarios. Practise distinguishing broad themes from measurable research questions and deciding between quantitative and qualitative approaches.
Week 2: Sampling and Survey Quality
Compare convenience, quota, simple random and stratified random sampling. Rewrite leading questions and identify sample-selection limitations.
Week 3: Data Processing and Presentation
Work with percentages, means, medians, modes, scatterplots, maps and graph selection. Require interpretations appropriate to the data.
Week 4: Reliability, Validity and Full-Question Transfer
Complete a fictional 25-mark-style fieldwork sequence with unfamiliar context, then evaluate the evidence and suggest specific justified improvements.
This four-week cycle supports revision but does not replace the school’s 10-week fieldwork learning experience.
What Parents Can Do
Parents can ask the child why a survey question is appropriate, who was included in a sample, and whether a graph actually answers the inquiry.
They can also encourage noticing patterns in familiar public spaces without turning ordinary outings into unsafe fieldwork assignments.
If the student has completed school-supervised fieldwork, ask them to explain what was learned and what they would change next time. The goal is reflection, not a perfectly polished report that conceals genuine difficulties.
eduKateSG’s broader 3-pax small-group instructional model offers opportunities for learners to defend methods and interpretations, but additional tuition is unnecessary when school support and independent work are effective.
Frequently Asked Questions
Is the 2027 G2 Geography fieldwork question compulsory?
Yes. Paper 1 Question 1 is a compulsory 25-mark fieldwork question.
How long is Paper 1?
One hour and 45 minutes, worth 50 marks and 50% of K226.
Can the fieldwork question use an unfamiliar context?
Yes. SEAB says the context may or may not be based on a syllabus cluster. The student should practise transferable methods.
Is there a separate external coursework grade?
The 2027 K226 assessment scheme lists two written papers. It does not specify a third separately weighted coursework component, though extended fieldwork over ten weeks forms part of teaching and learning.
What kinds of sampling are in the syllabus?
Convenience, quota, simple random and stratified random sampling.
What questionnaire types are specified?
Closed-ended surveys including Likert, frequency and ranking scales, plus mental maps and semi-structured interviews for relevant qualitative study.
Which numerical measures appear?
Counts, percentages, mean, mode and median, and relationships such as positive and negative correlation using scatterplots.
What should a map include?
Relevant elements include title, date, orientation, scale, legend, author and sources.
Is the six-mark judgement question in fieldwork Question 1?
The special six-mark AO3 level-descriptor question in Paper 1 is specified for Question 2 on Tourism, not automatically Question 1.
Should students memorise a particular neighbourhood survey?
No. Understand the method and practise applying it to unfamiliar evidence and contexts.
What is the difference between reliability and validity?
Reliability concerns consistency or dependability; validity concerns whether the method and findings meaningfully address the intended question.
Does a larger sample always solve survey bias?
No. A large but badly selected sample may remain biased. The sampling method and the intended population matter.
Official Sources and Related Reading
- SEAB 2027 G2 Geography K226 syllabus — printed page 6 for the 25-mark fieldwork question; pages 7 and 12–13 for the 10-week learning experience and Geographical Methods.
- SEAB 2027 G2 syllabuses.
- How G2 Geography Works | SEC Guide — the complete two-paper syllabus overview.
- What Is a Pilot Study in G2 Geography Fieldwork? — a focused companion for testing methods.
- What Are Reliability and Validity in G2 Geography Fieldwork? — targeted evaluation practice.
- How G2 Mathematics Real-World Problem Solving Works | SEC Guide — useful for quantitative reasoning.
- eduKateSG’s immutable Clementi small-group learning reference — error diagnosis and independent transfer.
Final Takeaway: Good Fieldwork Is a Chain of Defensible Decisions
G2 Geography Paper 1 Question 1 assesses whether a student can understand why an investigation is designed in a particular way, how information is collected and what the findings justify. For 2027 K226, the question is compulsory, worth 25 marks and can involve unfamiliar settings.
The most effective preparation combines real school-supervised inquiry with practice in sample selection, questionnaire design, numerical and spatial analysis, and specific evaluation. Students should aim to explain their method and conclusion when the location or dataset changes.
Examination requirements checked against the official SEAB 2027 K226 syllabus on 13 November 2026. All fictional examples above are learning exercises, not actual fieldwork results or predicted examination tasks.