AS 3: Fieldwork Skills and Techniques in Geography
Welcome to your complete revision guide for Unit AS 3: Fieldwork Skills and Techniques in Geography (Unit code: SGG31). Whether you love getting out into the field with a flow meter or find calculating statistical tests a bit daunting, this guide will break down every single concept into clear, manageable steps.
Exam Structure Quick Summary:
• Duration & Marks: 1-hour written examination worth 60 marks total (20% of your AS Level; 8% of the full A Level).
• Question 1 (Section A): Directly tests your own primary fieldwork investigation. You will take your completed Fieldwork Summary Statement (title, aim, hypotheses, location, and context) and your Table of Data (raw primary data) into the exam room and attach them to your booklet.
• Question 2 (Section B): Tests your unseen geographical skills using maps, data tables, satellite photos, and GIS resources provided in the examination paper.
Section 1: The Fieldwork Inquiry Cycle (Question 1 Preparation)
Every successful geographical investigation follows a logical cycle. Let us walk through each stage step-by-step.
Stage 1: Aims, Context & Hypotheses
Before stepping outside, you must know what you are investigating and why.
• Geographical Aim: The overarching purpose of your study (e.g., "To investigate downstream changes in river channel characteristics along the River Shimna").
• Geographical Context: Linking your investigation to established geographical theories and models (such as the Bradshaw Model or Hjulström Curve for rivers, or Burgess Concentric Zone Model and Distance Decay for urban studies).
• Hypotheses: Clear, measurable statements predicting a relationship between two variables.
- Alternate Hypothesis (\(H_1\)): Predicts a significant relationship or difference (e.g., "There is a significant negative correlation between distance downstream and bedload particle size").
- Null Hypothesis (\(H_0\)): Predicts no relationship (e.g., "There is no significant relationship between distance downstream and bedload particle size; any observed pattern is due to chance alone").
Stage 2: Risk Assessment and Safety Management
Examiners frequently ask you to explain how you managed risk. A good risk assessment identifies the hazard (what could cause harm), evaluates the risk severity and likelihood, and implements actionable mitigation strategies (precautions).
Common Fieldwork Hazards & Actionable Mitigations:
• Fast-flowing water / deep pools in rivers: Pre-fieldwork site checks, wearing waders/buoyancy aids, working in buddy pairs, never wading into water deeper than knee level, and avoiding fieldwork during/after heavy rainfall events.
• Slippery, uneven riverbeds or coastal rocks: Wearing sturdy footwear with high-traction rubber soles, moving cautiously, and carrying a basic first aid kit.
• Traffic hazards in urban studies: Wearing high-visibility reflective vests, standing away from road curbs, conducting pedestrian counts from wide footpaths, and using designated pedestrian crossings.
• Adverse weather conditions (hypothermia or heat exhaustion): Checking weather forecasts in advance, wearing waterproof and windproof clothing, carrying warm layers or sun protection and water.
Stage 3: Sampling Strategies
You cannot measure every single pebble on a beach or interview every person in a city. You must choose an appropriate sampling method:
1. Random Sampling
• How it works: Every point or individual in the study area has an equal chance of being selected. Locations are selected using a random number generator matched to grid coordinates.
• Strength: Eliminates researcher bias.
• Limitation: Can leave large spatial gaps across your study area, missing important features.
2. Systematic Sampling
• How it works: Data is collected at regular, equal spatial or temporal intervals (e.g., measuring river depth every \(0.5\text{ m}\) across a channel transect, sampling pebbles every \(5\text{ m}\) along a coastal transect, or surveying every \(10\text{th}\) person passing a cordon line).
• Strength: Ensures even, comprehensive spatial coverage across an environmental gradient.
• Limitation: May accidentally hit a periodic pattern (bias) in the environment.
3. Stratified Sampling
• How it works: The population or study area is split into distinct homogeneous sub-groups (strata), such as upper/middle/lower river reaches, age groups, or socioeconomic zones. Samples are taken proportionally from each sub-group.
• Strength: Ensures all distinct sub-environments or demographics are represented fairly.
• Limitation: Requires prior secondary data or knowledge to define the strata accurately.
4. Pragmatic / Opportunity Sampling
• How it works: Locations or participants are chosen based on accessibility, physical convenience, and safety constraints.
• Strength: Highly practical when time is restricted or riverbanks/cliffs are inaccessible.
• Limitation: High risk of bias and unrepresentative data.
Stage 4: Primary Data Collection Methods and Equipment
A. Fluvial (River) Data Collection:
• Velocity: Measured using an impeller / digital flow meter placed at \(0.6\) of the river depth from the surface (where mean velocity occurs). Alternatively, use the float method over a measured distance (e.g., \(10\text{ m}\)). Calculate velocity as:
\(\text{Velocity } (\text{m/s}) = \frac{\text{Distance } (\text{m})}{\text{Time } (\text{s})} \times 0.85\)
(Note: We multiply by a calibration coefficient of \(0.85\) to account for surface water moving faster than the average channel flow due to friction at the bed and banks).
• Channel Cross-Section: Stretch a taut tape measure across the wetted perimeter from bank to bank. Lower a graduated metre ruler vertically to the bed at regular intervals (systematic sampling) to record depth.
• Bedload Sediment Characteristics: Pick up pebbles systematically across the transect. Measure the \(a\)-axis (longest length), \(b\)-axis (intermediate width), and \(c\)-axis (shortest thickness) using a calliper or ruler. Assess particle shape using Powers' Index of Roundness (a visual visual chart scoring from \(1 = \text{Very Angular}\) to \(6 = \text{Well Rounded}\)) or Zingg's Shape Classification.
B. Coastal Data Collection:
• Beach Profile / Gradient: Place two ranging poles vertically at break-of-slope intervals. Sight through a clinometer from a set height on one pole to the identical height on the second pole to measure slope angle in degrees.
• Shingle & Sand Sampling: Use a quadrat or point-frame along a transect to sample pebble sizes and vegetation cover systematically across the shore.
C. Urban / Human Data Collection:
• Environmental Quality Assessment (EQA): Uses a bipolar semantic scale (typically rated \(-3\) to \(+3\)) across criteria such as noise levels, litter, building maintenance, and green space.
• Pedestrian / Traffic Footfall: Tally counts recorded over fixed time windows (e.g., \(5\) minutes) at selected survey points.
• Questionnaire Surveys: Administer structured questionnaires combining closed questions (for quantitative data) and open questions (for qualitative insights).
Key Takeaway for Section 1: Always justify your sampling methods and equipment by explaining how they reduce bias, increase accuracy, and keep you safe.
---Section 2: Data Presentation Techniques
Selecting the right graphical technique is essential for displaying data accurately:
• Scattergraphs: Ideal for showing the relationship between two continuous variables (e.g., distance downstream vs. pebble roundness). Allows you to plot a line of best fit and spot anomalies.
• Radial / Radar Graphs: Excellent for multi-variable scores at a single location, such as comparing different categories of an Environmental Quality Assessment (EQA).
• Triangular Graphs: Used when data consists of three components that sum to \(100\%\) (e.g., percentage composition of sand, silt, and clay in sediment; or agricultural, industrial, and service employment).
• Compound, Divided & Percentage Bar Charts: Used for discrete categories where sub-components are stacked to compare totals and relative proportions.
• Proportional Symbols: Circles or squares drawn proportional to the data value at specific locations on a map (e.g., pedestrian count volume).
• Choropleth Maps: Spatial areas shaded in darker tones to represent higher data densities or averages (e.g., census ward income levels).
• Isoline Maps: Continuous lines connecting points of equal value (e.g., contours for elevation, isovels for river velocity).
Key Takeaway for Section 2: Never choose a graph simply because it looks pretty. Match continuous data to scattergraphs/line graphs and 3-component percentage data to triangular graphs.
---Section 3: Statistical Analysis and Interpretation
Measures of Central Tendency & Dispersion
• Mean (\(\bar{x}\)): The arithmetic average calculated by summing all values and dividing by the total count (\(n\)):
\(\bar{x} = \frac{\sum x}{n}\)
• Median: The middle value when data is arranged in ascending order (resilient to extreme outliers).
• Mode: The most frequently occurring data value.
• Interquartile Range (\(IQR\)): The spread of the middle \(50\%\) of data values, calculated as the upper quartile minus the lower quartile:
\(IQR = Q_3 - Q_1\)
• Standard Deviation (\(\sigma\)): Measures the dispersion of data points relative to the mean. A low standard deviation shows data clustered tightly around the mean; a high value indicates wide variation:
\(\sigma = \sqrt{\frac{\sum (x - \bar{x})^2}{n}}\)
Spearman's Rank Correlation Coefficient (\(r_s\))
Spearman's Rank tests the strength and direction of a monotonic relationship between two sets of ranked continuous data.
The Formula:
\(r_s = 1 - \frac{6 \sum d^2}{n(n^2 - 1)}\)
Where:
• \(n = \text{number of matched data pairs}\)
• \(d = \text{difference between the ranks of the two variables for each pair}\)
• \(\sum d^2 = \text{the sum of all squared rank differences}\)
Step-by-Step Guide to Calculating \(r_s\):
Step 1: Set up a table with columns for: Variable 1, Variable 2, Rank 1, Rank 2, \(d\), and \(d^2\).
Step 2: Rank both variables separately from highest (\(1\)) to lowest (\(n\)).
Handling Tied Ranks: If two or more values share the same number, assign them the average of the rank positions they occupy. For example, if values occupy rank positions \(3\) and \(4\), both receive a rank of \(\frac{3 + 4}{2} = 3.5\). The next value gets rank \(5\).
Step 3: Calculate the difference in ranks (\(d = \text{Rank 1} - \text{Rank 2}\)) for each row.
Step 4: Square each difference (\(d^2\)). (All \(d^2\) numbers must be positive!)
Step 5: Sum all the squared differences to find \(\sum d^2\).
Step 6: Substitute \(\sum d^2\) and \(n\) into the formula and solve.
Interpreting the \(r_s\) Value:
• The calculated \(r_s\) value always falls between \(-1.0\) and \(+1.0\).
• \(+1.0 = \text{Perfect positive correlation}\)
• \(0.0 = \text{No correlation}\)
• \(-1.0 = \text{Perfect negative correlation}\)
Assessing Statistical Significance:
To find out if your result is genuine or just lucky chance, compare your calculated \(r_s\) value (ignoring any negative sign) to the critical values table for your sample size (\(n\)):
• \(p = 0.05\) (\(95\%\) confidence level): There is only a \(5\%\) probability that this correlation occurred by random chance.
• \(p = 0.01\) (\(99\%\) confidence level): There is only a \(1\%\) probability that this correlation occurred by random chance.
The Golden Rule of Exam Phrasing:
• If your calculated value exceeds the critical value at \(p = 0.05\):
"We reject the null hypothesis and accept the alternate hypothesis at the \(95\%\) significance level."
• If your calculated value is less than the critical value at \(p = 0.05\):
"We fail to reject the null hypothesis; there is no statistically significant relationship at the \(95\%\) level."
• NEVER write: "I accept the null hypothesis" or "This proves X caused Y". Correlation does not prove causation! Compounding variables may be at work.
Linking Results Back to Geographical Theory
In Question 1, do not stop at the numbers. You must explain the underlying physical or human processes behind the pattern:
• River Sediment Size Decline: Explain using attrition (pebbles colliding and breaking into smaller, rounder fragments) and hydraulic action.
• River Velocity Increase Downstream: Explain using the Bradshaw Model—increased discharge and reduced relative friction due to higher hydraulic radius despite lower channel slope.
• Urban Pedestrian Footfall: Explain using bid-rent theory and accessibility clustering in the Central Business District (CBD).
Key Takeaway for Section 3: Rank carefully, watch out for tied ranks, state whether you reject or fail to reject \(H_0\), and always explain the geographical process driving the numbers.
---Section 4: Unseen Geographical Skills (Question 2 Preparation)
Question 2 tests your ability to interpret unseen Ordnance Survey maps, satellite imagery, and spatial datasets.
Cartographic & Ordnance Survey (OS) Map Skills
1. Grid References:
• Memory Trick: Remember "Along the corridor, then up the stairs" (Eastings first along the bottom, then Northings up the side).
• 4-Figure Grid Reference: Identifies a \(1\text{ km} \times 1\text{ km}\) grid square (e.g., \(3462\)).
• 6-Figure Grid Reference: Identifies a precise \(100\text{ m} \times 100\text{ m}\) location by estimating tenths within the square (e.g., \(344628\)).
2. Scale & Distance:
• On a \(1:25,000\) map: \(4\text{ cm} = 1\text{ km}\) on the ground (each \(1\text{ km}\) grid square is \(4\text{ cm} \times 4\text{ cm}\)).
• On a \(1:50,000\) map: \(2\text{ cm} = 1\text{ km}\) on the ground (each \(1\text{ km}\) grid square is \(2\text{ cm} \times 2\text{ cm}\)).
3. Calculating Gradient:
Gradient measures the steepness of a slope between two points:
\(\text{Gradient} = \frac{\text{Vertical Interval (VI)}}{\text{Horizontal Equivalent (HE)}}\)
• Vertical Interval (VI): The difference in height between two points (calculated using contour lines in metres).
• Horizontal Equivalent (HE): The real-world horizontal distance between the two points measured along the map scale (converted to metres).
• Example: If height difference (\(\text{VI}\)) is \(100\text{ m}\) and map distance (\(\text{HE}\)) is \(1000\text{ m}\):
\(\text{Gradient} = \frac{100}{1000} = \frac{1}{10} \text{ or } 1:10\)
Remote Sensing & GIS (Geographic Information Systems)
• Aerial & Satellite Imagery: Look for textures, geometric vs. irregular shapes, tone, and linear features to identify changes in land use, informal settlement sprawl, rates of deforestation, or the impact of natural hazards (e.g., flood extent or coastal retreat).
• Geographical Information Systems (GIS): Digital systems that layer spatial datasets (e.g., overlaying a flood risk map, road networks, and population density layers). GIS uses buffering (creating zones of specific distance around features, like a \(500\text{ m}\) conservation buffer around a river) to support spatial decision-making.
Key Takeaway for Section 4: Always check the scale bar first, double-check your Eastings and Northings order, and ensure units match when calculating gradients.
---Section 5: Common Pitfalls and Examiner Advice
Avoid these frequent mistakes highlighted in CCEA examiner reports:
• Mistake 1: Vague sampling justification.
Wrong: "We used systematic sampling because it was quick and easy."
Right: "Systematic sampling was used because taking depth measurements at regular \(0.5\text{ m}\) intervals eliminated researcher bias and provided an even, objective profile across the entire river channel."
• Mistake 2: Ranking errors with Spearman's Rank.
Always check that the highest value is rank \(1\) and that tied ranks receive the mean position. Double check that your \(d\) values are calculated from the ranks, not the raw data values!
• Mistake 3: Stating correlation proves cause and effect.
A high correlation between pedestrian numbers and distance from CBD does not mean distance alone caused footfall; store quality, public transport hubs, and weather are compounding factors.
• Mistake 4: Disconnecting statistics from geography.
Always complete your data analysis answers by explicitly referencing relevant geographical concepts (e.g., load attrition, hydraulic efficiency, distance decay).
Final Review Checklist:
• Can you clearly define your Fieldwork Aim, \(H_1\), and \(H_0\)?
• Can you evaluate random, systematic, and stratified sampling with real examples?
• Can you calculate \(\bar{x}\), \(IQR\), \(\sigma\), and \(r_s\) step-by-step?
• Can you explain your statistical outcome using correct geographical terminology?
• Can you calculate gradient, 6-figure grid references, and interpret GIS layers?