Statistics

• Standard deviation 


  Gives an indication of the spread of values around the mean

  range influenced by a single outlier, S.D isn’t

  low standard deviation allows us to test whether differences are significant

  if results overlap there is no significant difference
• Correlation test 


  Used to find significant correlation between two sets of measurements from same sample

  null hypothesis : no significant correlation between x and y

  Critical values : number of pairs of measurements
• T test 


  Used to find if significant difference between two means with the same variable

  null hypothesis : there is no significant difference between x and y

  Degrees of freedom - 2
• Chi squared 


  Difference between frequencies

  null hypothesis : there is no significant difference between observed and expected

  compare calculated value against critical value at 0.05

  If calculated value is more than the critical value p 0.05. Reject null hypothesis, difference is significant
• What do scatter diagrams show ?

  Relationships and potential correlations between two (discrete/ independent) variables
• Explain the difference between correlation and casual relationships ?

  Correlation: Relationship between two variables. ( there may be other named factors involved that produce a change in both Causal relationship: One variable directly influences the other.
• Why are order of magnitude calculations useful in biology

  To make approximate comparisons of size by rounding them to the nearest power of 10
• What is standard deviation

  A measure of the spread of values around the mean
• Whu is standard deviation a more susueful measure of the dispersion than the range for some data stets ?

  Standard deviation shows the spread of data around the mean where the range only shows the difference between the highest and lowest values

  so standard deviation reduced the effect of anomalies or outliers
• Define uncertainty 


  Potential amount of error in a reading or measurement

  indicating the range within which the true value is expected to lie
• When are why are logarithms useful when plotting graphs ?

  when there is a wide range of Values

  logarithmic scales compress wide ranges of data making individual data points easier to read