# Statistical Techniques

What is the purpose of statistical techniques? What are their practical applications? What is there value? What advantages/ disadvantages do they have?

• Created by: Bethany
• Created on: 08-04-14 19:26
• Statistical Techniques
• Measures of Central Tendancies (Mean, Mode, Median)
• Purpose
• Summarising data
• Comparing Data
• Application
• Comparing pebble sizes on two stretches of beach
• Enables us to summarise a data set, giving the middle value or the most frequently occurring.
• The MEAN is particularly useful if the data has a small range
• BUT if the range is large, the mean is likely to be heavily  influenced by extreme values which could give a distorted picture.
• The MODE is of no value if there are no repeating values
• There may be than one mode e.g. BI-MODAL
• The MEDIAN is not influenced by extreme values
• Range
• Purpose
• The dispersion and variability of data
• Allows to identify data in more depth
• Application
• Giving the difference between lowest and highest values in a data set e.g marks in a geography test.
• Comment
• Gives a basic idea of the spread of data.
• Affected by extreme values
• Anomolies can lead to a false picture
• Inter-Quartile Range
• Purpose
• Dispersion and Variability of Data
• In-depth analysis of data
• Application
• Shows middle 50% values
• Comment
• More useful than the range in indicating the spread of data as it excludes outliers
• Standard Deviation
• Purpose
• Analyse dispersion and reliability of data
• Measure of the degree of dispersion
• Application
• Working out the standard deviation of bedload across a stream.
• Comment
• Simply another way of examining the spread of data
• Lets you know reliabilty of mean
• Spearmans Rank
• Purpose
• Shows correlation of two sets of data and their statistical significance
• Test of strength of relationships
• Application
• Showing the relationship between distance inland and number of species in a psammosere
• Comment
• Good to analyse a scattergraph (not speculative, unlike a best fit line)
• Enables you to demostrate a clear relationship between two data sets
• Correlation does not mean causation
• Only really works with 10 - 30 data sets
• Chi-Square
• Purpose
• The degree to which there are differences between observed and expected data
• Application
• Investigating spatial disrabutions e.g. plants types at different stages of a succession.
• Comment
• Doesn't expalain patterns in distrabutions
• Can test significance/ test categories/ distrabutions
• Amount of observed data must be between 4 and 20