Stage 3- Collecting and recording data appropriate to the geographical question or hypothesis

Primary data

• How was it collected? equipment, strategy and methods used- accuracy and reliability
• Questionnaire design- type of questions used (closed versus open) length, postal or face to face
• Sampling size and type- systematic, random, stratified, pragmatic
• Sampling unit- area, point, linear
• Map of sample sites, repeated sampling for accuarcy

Secondary data

• Source
• Data
• Reliability and accuracy

Recording

• Nature of tally sheet
• Any problems

Geographic Information Systems (GIS)

• Covers a wide range of ICT programs
• Primary data collection e.g. data logging temperatures at specific locations/times, measuring distances/areas
• Source of secondary data e.g. satellite photos, layers of data from maps
• Enables spatial data to be organised and investigated to show patterns and relationships

Stage 3: collecting and recording data

• Can be used in the field to measure and record as information can be linked to a specific point or location
• Can be used to measure distances in a straight line or areas or outlines more accurately than in the field or from a map
• Secondary data can be accessed from a number of sites such as census returns, aerial photographs and satellite images
• Recording data is the greatest advantage as data gets tied to an exact point on the earth's surface
• Data is usually usually stored in two forms- raster data= form of cells and their subsects or vector data= points, lines and outlines

• Area, line and point data can be displayed on the computer screen and manipulated
• When presenting mapping data, a high quality map can be created because symbols can be manipulated and colour coded to create an output map with ttitle, scale bar, north narrow etc.
• Data can be represented in a vast range of styles, colours and dimensions
• Data can be overlaid or integrated to enable comparisons to be seen or demonstrated
• Data can be linked, layered and shown in 3D to make patterns and trends visable and clearer

• Increases the ability to ask questions and allow data interrogation
• GIS enables selection of specific aspects and identification of features and their attributes to meet the particular criteria
• Distances can be measured between any nominated features or locations give exact measurements
• Raster analysis- GIS may also store geographic information as a grid, with each cell in the grid containing a subset of the geographic information

• Handles large amounts and varieties of data so it can be manipulated more easily
• Flexible over size of study area
• Gives precise data for precise locations
• Gives precise measurements of area, distance, direction etc
• Easy to keep up to date as it is dynamic and therefore current
• Faster and more efficient at handling data
• Can easily change the scale or sub-divisions of data
• Can overlay patterns, so suggesting possible links
• Produces striking visual impacts, including 3D images

Main ways of collecting primary data:

• Measuring and counting
• Mapping and plotting
• Making images, including sketches and photos
• Asking questions
• Using ICT- data logging

Main sources of secondary data:

• Graphical sources- maps, diagrams etc
• Images- photos, satellite images, internet
• Written sources- newspapers, books
• Oral sources- radio, tapes
• Statistics- census data

Advantages

• Accurate located detail
• Scale allows precise measurement
• Relative directions clear
• Can see over hills etc
• Get plan view
• Gives place names etc

Disadvantages

• Static picture- shows only static features so can't show seasonal changes
• Quickly dated
• Impersonal
• Sheer size

Advantages

• Captures an instant of time- shows ephemeral features e.g. weather
• Shows precise uses e.g. types of tree, shop etc
• Shows people
• Shows qualitative features

Disadvantages

• Instant of time- so quickly becomes out of date
• Scale problems
• Subjective- image changes with light, angle etc
• Can't show 'hidden' areas
• Doesn't show precise classification e.g. roads

Useful because:

• Show a context of the data collected
• Identify particular features
• Convey a subjective impression of a place

should always have

• A clear location
• An informative title
• A note of the date/time
• An indication of scale (vertical and horizontal) and direction
• Annotations to identify specific features

It is often effective to put a photograph and an annotated sketch of the same view on the same page or use the sketch as an overlay

Sampling is often more difficult in reality than when planning it reducing the reliability of samples

Questionnaires are difficult to ask randomly, systematically or in a stratified way because:

• Take time so only a limited number of people can be asked
• Many people are put off by questionnaires
• Certain groups find it difficult to respond e.g. parents with young children
• Certain groups may respond too readily to questionnaires e.g. pensioners

Therefore, questionnaires are usually pragmatic- and rarely answered by a genuine cross-section of society

The nature of the population being sampled will change according to: time of day, day of week, time of year and other factors like the weather

• A sample is a sub-set of items selected from the statistical population
• Most geographical investigations are based on samples rather than the statistical population. However, any investigation is only as good as the sample on which it is based
• A successful sample will enable researchers to make valid inferences about the characteristics of a population from the characteristics of the sampled area
• Sampling the entire statistical population is impractical and is too large

Qualties of a good sample

• Unbiased
• Precise
• Large enough to provide conclusive results in terms of statistical significance
• Can be collected easily with the minimum of resources

3 types- random, systematic and stratified

Random sampling

• Every item in a population has an equal chance of inclusion in the sample
• Use random number tables to select sample point
• Most appropriate if there is a listing of the population
• Advantages- statistically sound so opens up further analysis
• Disadvantages- same item could be picked more than once, easy to miss something

Systematic Sampling

• Items are sampled from a population at regular intervals
• Often used when sampling from an OS Map and wherever speed of sampling is needed
• Advantages- Easy to do, quick
• Disadvantages- interval may coincide with one in the data or location, can't make valid statistical references

Both random and systematic sampling methods are often combined with stratification

• Many statistical populations aren't homogeneous, consisting of many sub groups and some sub groups are relatively small they may be underrepresented by random and systematic sampling

There are two strategies that aim to represent the sub-groups:

• Stratified random and systematic methods select sample individuals objectively
• Quota sampling a researcher deliverately chooses individuals to fill the quota for each sub-group, although it is less scientific it is more practical due to constraints of time and money

Stratified sampling- takes account of underlying patterns and ensures all are sampled

• Used where there are distinctive differences in something
• Advantages- ensures no significant aspect is missed
• Disadvantages- data can be biased, can't make valid statistical inferences

Areal sampling- where the unit is a pre-defined area such as a quadrat

Linear sampling- where sampling is along a line (transect sampling occurs along a line)

Point sampling- where discrete or separate points or individuals are chosen as the sampling unit

Pragmatic sampling- method of sampling in which the samples are chosen from those that are easily and safely available

Random sampling- method of sampling in which all the elements of data have an equal chance of selection

Stratified sampling- method of sampling in which the sampling of the data from subsections of the whole is proportional to their share or size of the whole

Systematic sampling- method of sampling in which the sampling is at a set interval from the initial point, which is chosen at random

Used where location is an essential feature of the items being studied. samples are selected from points, areas and lines using random, systematic and stratified sampling strategies

Can be illustrated with reference to the 1:25000 land use map

Point sampling

• Random- can be generated using random number function on a calculator to generate grid references
  • This is time consuming and unless a large enough sample is taken, the distribution of random points on the map could give rather uneven coverage
• Systematic- relies on a grid of coordinates placed randomly over the map
• Randomly drawn lines of a transect- sample points can either be at regular intervals along a transect, or at locations dervived from random numbers
  • In order to collect a large enough sample several lines of a transect are needed
  • This may cause uneven coverage on the map

• Area sampling is based on quadrats which are either a square of standard size on a map or a metal frame, sub-divided into smaller squares used in the field
• Quadrats are located either randomly or systematically
• Within quadrats sampled items may be counted, recorded as absent or present, or their areal coverage estimated
• Outcomes from quadrat sampling are strongly influenced by quadrat size
• Quadrat size is also influenced by the geographical distributions of the population
  • Where populations are highly clustered, small quadrats randomly or systematically may fail to pick up enough samples
  • Small quadrats are much less likely to generate sufficient samples to s how influence on geographic distributions than larger ones

• Transects are particularly effective where they cross constrasting areas of relief, geology, land use, soils and so on
• In urban studies transects can follow the street pattern
• Sample data are collected at random and systematic intervals either at points along a line of transect
  • e.g. collecting a sample of scree particles on a talus slope- a base line is establised at the slope foot, transects are located randomly and at right angles to the base line to the top of the slopem then quadrats are located at systematic interavals on the slope
• Belt transects are a type of line sample
  • Often used when populations are thinly spread across an area and where a line transect would fail to generate a large enough sample
  • A belt is identified either side of a random line of transect and sample items are measured and recorded within this zone

• Some investigations may encounter errors if simple random or systematic sampling methods are used
• The solution is random stratified where the population is examined before the sample is taken enabling the combining of line, quadrat and stratified sampling if necessary

Reliability

• The extent to which sample data reflects the greater whole
• Refers to how dependable the data is and whether it reflects the whole population
• Reliability can be increased by the use of pilot studies and repeating the measurements or readings at different times
• An important part of making data reliable and as accurate as possible is the size, type and method of sampling used

• Is the level at which the data is exact and free from error
• Methods can be reliable but results can still be inaccurate
• This inaccuracy can occur from a number of sources- often the equipment isn't accurate enough, or different members of the group read the measurement differently
• Can be unexpected such as the weather
• Another source of inaccuracy is using other people's primary data when you don't know the exact conditions and circumstances it was collected
• Sometimes the raw data itself suggests inaccuracy for example anomalies which are data that does not fit into the common pattern or trend