Skills

Skills

1. Cartographical Skills

   • Sketch Map
     • Positives
       • Can pick out features you wish to identify and comment upon
     • Negatives
       • Helpful if you are a decent artist
       • Qualitative technique that is open to bias
     • Example
       • To draw a meander of the River Derwent. This helped me understand the asymmetrical channel shape they have and role of the thalweg in its formation
   • Proportional Symbols
     • Positives
       • Visual representation of data, can be places over the area under investigation
       • Good Spatial technique
       • Further info can be displayed within the symbol about that place
     • Negatives
       • Care needed to select scale so symbol fits
       • Symbols can overlap
       • Symbol may obscure location
     • Example
       • QGIS, plot river velocities over a base map of the River Derwent at our sites. Helped point out anomalies, shows how velocity increases as you go downstream
   • Flow/Desire/Trip Lines
     • Positives
       • Represent volume of movement from places
       • Can see patterns easily
     • Negatives
       • Desire lines are drawn directly from origin to destination without the route being taken into account
       • The selecting of scales
       • Overlap of lines sometimes
       • Lacks precise information unless a scale is drawn
     • Example
       • Course work, sphere of influence to show where people had come from to be in Sheffield that day. not many from west due to Peak District and Manchester.
   • Choropleth Map
     • Example
       • Map crime rates in calderdale, identified central Halifax as high in crime - lower in the east - Hipperholme
     • Positives
       • Easily constructed and visually effective
       • General patterns can be identified
     • Negatives
       • It assumes the whole area under one shade is the same, variations will exist
       • The boundaries decided upon the shades are open to personal interpretation
       • There is often an abrupt change in shade at a boundary, in reality there would be a gradual change
   • Isoline Map
     • Advantages
       • A distribution can easily be identified and colour can be added to show distribution more clearly
       • Appropriate for physical and human work
       • General trends in distribution are easy to see
     • Disadvantages
       • The joining of same values with a line is subjective and open to misinterpretation
       • Difficult to construct
     • Eg. Cranedale, plotted on a meander plots of equal water depth, linked river velocity to these positions to see if a relationship existed
   • Dot Map
     • Positives
       • Good visual impression of distribution
     • Negatives
       • Large numbers of dots are difficult to count
       • Less valuable if you need a precise idea of the values they represent
       • Areas can be missed out if they don't conform to the scale
     • Eg. Dot distributed american cities that have over 1/2 million people

2. Graphical Sources

   • Line Graph
     • Example
       • Can clearly see population explosion from 1950-2016
     • They show well absolute changes in data
     • 2 sets of data can be shown on the same graph, using both vertical axes - this data can then be connected perhaps
     • The scale can be difficult to set sometimes
     • Can only be used with continuous data
   • Pie Charts
     • Visually effective, size of circle and proportions within in
     • Informs reader of several pieces of information
     • Lots of segments can cause confusion and can be cumbersome to construct
     • No exact numerical data
     • Was plotted to show population of African countries, young, old, independent, dependent
   • Bar Graphs
     • Easy to construct, and show relative magnitudes effectively
     • Positive and negative data can be plotted
     • Too many bars can make it cluttered
     • Difficult to show a wide range of data
     • Show average rainfall and whether it is above/below average
   • Scatter Graph
     • Eg. Velocity vs hydraulic radius
     • When a best fit line is added, it shows well if relationships exist between two variables
     • Identified anomalies enable you to make further investigations
     • Not an accurate measure of degree of corrolation
     • Best fit lines are sometimes difficult to do
     • Too few data sets can produce misleading results
   • Triangular Graphs
     • Relative importance of each component can be seen
     • The dominant variable can be identified
     • Cluster of plots can be identified enabling conclusions to be drawn
     • Can show 3 variables
     • Can only be used for 3 components that equal 100%
     • Data must be a percentage
     • E.G percentage of population employed in services, industry and agriculture
   • Radical Graphs
     • Can display independent variables and useful for plotting the direction of a variable
     • Good for plotting information over a period of time
     • The scale around the edge must be of a continuous nature
     • Eg. Traffic counted cars at Ainley Top roundabout to see most popular route
   • Logarithmic Scale
     • Eg. Compare energy consumption between poor and rich countries, Hjulstrom curve
     • Useful for plotting data with a wide range to it
     • Correlations can be identified with the position of data plotted on it
     • Can be hard to plot accurate scale, 0 cannot be plotted
     • Negative and positive values cannot be plotted on the same graph
   • Kite Diagrams
     • Visually shows you changes over a distance
     • Differing species can be identified at different positions
     • Nightmare to construct, prone to human error
     • Sand dune succession as you moved inland on sand dunes, vegetation becomes more dense and varied

3. Statistical Test

   • Mean, Mode, Median
     • Crude measurements but allow further investigation using interquartile and standard deviation
     • They test the accuracy of central tendency for a set of data, S.D measures the mean, IQR measures the median
     • The answer can be used to plan and prepare for courses of action
   • The Mann-Whitney U Test
     • Test to see if there is a difference between the medians of two data sets, test assumes that the data is to normally distributed but that there is a smaller dispersion of both sets of data. The test begins with a a null hypothesis.
     • The data sets can have different sizes, eg one could have 10 values, the other 8
     • You can state is the relationship is significant or by chance, you can clearly see if there is a difference between the medians between the 2 sets
     • Very long to do and easy to make mistakes
     • It provides no explanation for your answer
     • Data has to be original
     • Can only be done for between 4 & 20 sets of data
     • Deciding when to use this test is difficult to judge