inferential statistics

• inferential statistics 


  allows you to infer if your hypothesis was correct (is the data significant?), and assess whether the data can be generalised to the target population
• distribution definition 


  how spread out our data is when plotted on a frequency graph
• negatively skewed distribution

  •the tail will be longer where majority of scores are smallest

  •measures of central tendency will increase in value

  
• normal distribution 


  •curve is symmetrical

  •no clear tail

  
• positively skewed distribution 


  •the tail will be longer where majority of the scores are larger

  •measures of central tendency will decrease in value

  
• probability definition 


  a numerical value that represents the likelihood of some events happening
• how is probability measured in

  on a scale from 0 to 1

  •0 means that it’ll never happen

  •1 means that it’ll definitely happen
• significance definition 


  for a result to be significant the probability for the result to have been down to luck/chance needs to be low
• what are the significance that can be used

  • P=0.05 (5%)

  • P=0.01 (1%)
• general rule regarding significance to stats tests

  completed statistical test:

  1. stats test finds a difference equal to 0.05 or less→results have happened for a reason→reject null hypothesis
  2. stats test finds a difference at more than 0.05→results have happened by chance→accept null hypothesis
• what are the two types of tests

  1. parametric
  2. non-parametric
• examples of non-parametric tests

  • Mann Whitney U Test
  • Chi Square
  • Binomial Sign Tests
  • Wilcoxon Signed Ranks Test
  • Correlations Spearman’s Rho